Karpovs Strategic "W'ins 2
The Prime Years
By
Tibor Karolyi
Quality Chess
www.qualitychess.co.uk
First edition 20 1 1 by Quality Chess UK Ltd
Copyright © 20 1 1 Tibor Karolyi
Karpov's Strategic Wins 2 The Prime Years
-
All rights reserved. No part of this publication may be reproduced, stored
in a retrieval system or transmitted in any form or by any means, electronic,
electrostatic, magnetic tape, photocopying, recording or otherwise, without prior
permission of the publisher.
Paperback ISBN 978- 1 -9065 52-42-8
Hardcover ISBN 978- 1 -9065 52-73-2
All sales or enquiries should be directed to Quality Chess UK Ltd,
20 Balvie Road, Milngavie, Glasgow G62 7TA, United Kingdom
Phone +44 1 4 1 227 677 1
e-mail: info@qualitychess.co.uk
website: www. qualitychess.co.uk
Distributed in US and Canada by SCB Distributors, Gardena, California, US
www. scbdistributors.com
Distributed in Rest of the World by Quality Chess UK Ltd through
Sunrise Handicrafts, Smyczkowa 4/98, 20-844 Lublin, Poland
Typeset by Jacob Aagaard
Proofreading by Colin McNab and John Shaw
Edited by Andrew Greet
Cover design by Adamson Design
Cover photo property of Caissa Italia
Printed in Estonia by Tallinna Raamatutriikikoja LLC
Contents
Key to Symbols used
Preface
Bibliography
1986
1987
1988
1989
4
5
6
7
33
61
97
1990
121
1992
175
1991
147
1993
219
1995
293
1994
1996
1997
265
34 3
375
1998
401
2000
433
1999
2001
2002
413
447
453
2003
477
2005
505
2004
491
2006
513
2008
533
2007
2009
523
541
2010
549
Epilogue
557
Classification
Game Index by Page Number
Game Index by Karpov's Opponents
Alphabetical Game Index - Non-Karpov Games
Name Index
Statistics
Tournament Record
555
559
565
567
569
575
576
Key to symbols used
±
+
+-+
�
iii
CD
?
??
!!
!?
?!
#
White is slightly better
Black is slightly better
White is better
Black is better
White has a decisive advantage
Black has a decisive advantage
equality
with compensation
with counterplay
unclear
a weak move
a blunder
a good move
an excellent move
a move worth considering
a move of doubtful value
mate
Preface
In this, the second volume, we will continue our investigation into Karpov's strategic victories, from
the time when he lost his World Championship title at the end of 1 985 until the end of the year
20 1 0. In the first volume we saw that during the ten-year period from 1 975- 1 98 5 , Karpov not only
held the highest title in chess, but also dominated the tournament circuit more convincingly than
any other player in chess history. And yet, despite being one of the greatest world champions of all
time, it was only after losing his title that Karpov played his very best chess. During his decade-long
reign as World Champion, Karpov worked hard but he already stood head and shoulders above the
other leading grandmasters. But when the young Garry Kasparov took over as the dominant force
in the chess world, everything changed.
Kasparov was able to do what no other player could. By defeating Karpov in their second
championship match (after their first match was aborted, as explained on page 4 1 7 of the first
volume) , he forced Karpov to work harder than he ever had before in an effort to reclaim his title.
Steel sharpens steel, and in the present book we will see how both of these titans were able to raise
their respective levels in an effort to outdo the other.
Despite his colossal efforts, Karpov was unable to reclaim the coveted crown from his nemesis.
Nevertheless, in terms of his overall skill as a chess player, Karpov's prime years occurred only after
his reign as World Champion had ended.
By the end of the 1 980s Karpov's level had begun to drop slightly, although he remained the
number two player in the world until well into the 1 990s. Indeed, his greatest tournament success,
which is widely considered the most impressive tournament performance of all time, occurred
in 1 994. AB the years went by, Karpov was gradually superseded by the next generation of super­
grandmasters, but he continued to achieve excellent results and produced a host of beautiful games
along the way.
Without further ado, I invite the reader to continue his voyage of discovery into the prime years
of Karpov's career.
Bibliography
Gelfand: My Most Memorable Games, alms 2005 .
Karolyi & Aplin: Endgame Virtuoso: Anatoly Karpov, New i n Chess 2007
Karpov: My Best Games, alms 2006
Kasparov: Kasparov vs Karpov, 1 986- 1 987, Everyman Chess 2009
Kasparov: Kasparov vs Karpov, 1 988-2009, Everyman Chess 20 1 0
Marin: Learn from the Legends (2nd edition) , Quality Chess 2006
Marin: Grandmaster Repertoire 3 - The English Opening Volume 1 , Quality Chess 2009
Periodicals
Skakbladet
1986
Rating 2700 (2 in the world)
For the first time in more than a decade, Anatoly Karpov began the year without holding the
title of World Champion. Having lost his crown, he must have spent some time contemplating
his future. Faced with the same situation, many players would probably have accepted that there
had been a changing of the guard, and settled down to enjoy life as a millionaire and a national
hero. Not so Anatoly Karpov. Luckily for the game of chess, he decided to focus all his efforts
on reclaiming his title. He must have known that it would take a phenomenal amount of work,
since Kasparov was not only a marvellous player, he was also only rwenry rwo years of age and
was only going to get better. Karpov was already thirry four, so if he was being realistic he would
have known he had roughly five or six years in which to achieve his goal.
One of the conditions of his last match with Kasparov was that, in the event of Karpov's defeat,
he would be guaranteed a rematch. It was scheduled for later in 1 986, but before that time
Karpov participated in a few other tournaments.
Karpov's first event of the year was the SWIFT tournament in Brussels. He started with a draw
against Torre, then defeated Timman after the Dutchman failed to find the best defence in a long
endgame. A fairly quick draw with Romanishin followed. In Round 4 Van der Wiel tried the Dely
Gambit, but Karpov had done some work since his match with Kasparov and he unveiled a strong
novelry which yielded a clear advantage. But Van der Wiel defended stubbornly and eventually
salvaged half a point after Karpov missed some chances. How Karpov must have wished he could
have found the right antidote to the gambit in time to use it against Kasparov!
In Round S Karpov drew quickly with Korchnoi, but then he switched into top gear and
amazingly won each and every one of his last six games! His win against Ljubojevic was mentioned
in the notes to Game 72 in the first volume. Then after beating the Belgian 1M Jadoul on the
black side of a Hedgehog, he met another local player, Luc Winants, who scored one point out of
six games against world champions.
I Gamel I
Luc Winants Anatoly Karpov
-
Brussels 1 986
l.d4 tiJf6 2.tiJa e6 3.e3 c5
Fifteen years earlier Karpov carried out a similar plan via a different move order: 3 . . . b6 4.�d3
�b7 S . tLl bd2 cS 6.0-0 cxd4 7.exd4 �e7 8.b3 0-0 9 .�b2 tLl c6 1 O.a3 Wc7 1 1 .Ei:e 1 Ei:ac8 1 2.c4 dS
The Prime Years
8
1 3 .:1:k 1 E!fd8 1 4.'1Mfe2 dxc4 1 5 .bxc4 Wf4 1 6.g3
Wh6 1 7.E!c2 Wh5 1 B .Wfl E!c7 1 9.ie2 Wf5
20.id3 Wh5 2 1 .ie2 Yz-Y2 Plato nov - Karpov,
Leningrad 1 97 1 .
b) Seventeen years later Karpov tested the
same line from the opposite side of the board:
1 1 .ttJe5 ttJxd2 1 2.Wxd2 ttJ d7 1 3 .f4 E!cB
1 4.E!ae 1 E!c7 1 5 .We2 g6 1 6.We3
4..id3 cx:d4
With this early exchange Karpov avoids a
symmetrical pawn structure.
5.exd4 b6 6.0-0 .ib7 7.c4 .ie7 8.ttJbd2 0-0
9.b3 d5
Karpov opts for a position with hanging
pawns; he has great experience with this
formation.
1O .ib2
a
b
c
d
e
f
g
h
•
1 6 . . . ttJ f6! The knight transfer to e4 equalizes
as it will be hard for White to build an attack.
1 7. f5 ? White embarks on an unsound attack.
1 7 . . . exf5 1 B .ixf5 gxf5 1 9.Wg5t mhB 20.E!xf5
icB 2 1 .E!f4 ttJ g4 Black defended successfully
and went on to win with his extra piece,
Lobron - Karpov, Frankfurt (rapid) 1 997.
8
7
6
5
4
1 1.'iMe2 'iMd6
3
This prepares an interesting plan based on
the transfer of the queen to the kingside.
2
1
a
b
c
d
e
f
g
h
10 ... ltJc6
Karpov chooses the main line. Interestingly,
he played the position after 1 0 . . . ttJ e4 with
both colours during his career:
a) 1 1 .E!e 1 Karpov used this move to defeat
Andersson twice, but then he abandoned it
after Portisch came up with the following
equalizing line: 1 1 . . .ttJxd2 1 2.ttJxd2 ttJ d7
1 3 . ttJ f3 E!cB 1 4.E!c 1 E!eB 1 5 .We2 id6 1 6.We3
dxc4 1 7.bxc4 Wc7 1 8 .g3 ttJ f6 1 9 .ifl WbB
20.ig2 ttJ d7 2 1 .Wd3 E!cdB 22.ttJg5 ttJ f6
23.d5 e5 24. ttJ e4 ttJxe4 2 5 .ixe4 g6= Karpov
- Portisch, Malta (01) 1 9BO.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
12J3adl
A more purposeful continuation was
1 2.E!ac 1 ! with the idea of a3 and later c5 . It
1 986
9
Luc Winants - Anatoly Karpov
looks like Winants was expecting Karpov to
exchange on c4, but the former champion
postponed it for a long time.
It is worth noting that White cannot play
1 2.ttJe5? due to 1 2 . . . ttJxd4! when Black wins
material.
queen in order to prepare h3 and g4, but he
never gets enough time to carry out the plan.
Correct was 1 5 .ttJe5 l'!fdS
( 1 6 ... ttJ e4!?) 1 7.ttJxe5 (If
l S . .ixc4 ttJ d5 Black is a bit
I S .l'!f1 reaching a balanced
chances for both sides.
1 6.ttJ df3 ttJ xe5
1 7.dxe5?! dxc4
better.) 1 7 . . . .ib4
position, rich in
12 .. J3ac8 13J3fel
1 3 .ttJe5 was possible, although after
1 3 . . . dxc4 ( 1 3 . . . ttJxd4?? does not work here, as
after 1 4 . .ixd4 dxc4 1 5 . ttJ dxc4 the bishop on
d4 is poisoned.) 1 4. ttJ dxc4 1.&d5 Black is not
worse in this IQP middlegame.
13" JWf4!?
The queen vacates the centre. She will help
to fortify the black kings ide, while making way
for the rooks to attack White's hanging pawn
centre.
14.g3
It was worth considering 1 4 .1.&e3!? 1.&xe3
1 5 .fXe3 when White's pawn centre has
been strengthened, although he has no real
advantage.
14".�h6
1 5".l'!fd8
Karpov calmly completes his development.
16.h3?!
White continues with his faulty plan.
With hindsight, it was worth considering
the prophylactic 1 6.a3. It would have been
interesting to see how Karpov would have
responded. There are several reasonable
options:
a) 1 6 . . . ttJ g4!? The threat of . . . dxc4 followed
by . . . ttJxd4 provokes a further pawn move
on the kingside. 1 7.cxd5 (another possibility
is 1 7.h3 dxc4 l S .bxc4 ttJ f6) 1 7 . . . l'!xd5 l S .h3
l'!h5 1 9.h4 l'!d5 Black has a good game, but he
has not achieved anything special.
b) 1 6 . . . g5!? Black wants to drive the knight
away from its defence of the d4-pawn. The
idea is interesting, but not at all in Karpov's
style. After 1 7.cxd5 ttJxd5 I S .ttJe5 the position
is about equal.
c) 1 6 . . . g6 1 7.h3 dxc4 (there is also 1 7 . . . 1.&fS ! ?
l S .c5 a 5 ) l S .bxc4 1.&fS 1 9.1'!a l ttJ eS Black
intends to increase the pressure on the d4pawn with . . . .if6.
a
b
c
d
e
f
g
h
15.1.&f1 ?!
This is too artificial. Winants retreats the
d) 16 ... l'!c7 This type of small improving move
was typical of Karpov's style. 1 7.h3 (If 1 7. ttJ e 5 ?
ttJxe5 l S .dxe5 ttJ g4 1 9.h3 ttJxf2! Black takes
over.) 1 7 . . . l'!dcS I S .ttJe5 The position remains
balanced after l S . . . g6 or l S . . . ttJxe5 1 9.dxe5
ttJ d7 20.cxd5 .ixd5 .
10
The Prime Years
6
Another idea was 1 9 .�c 1 ctJxd3 20.Ei:exd3 , but
after 20 . . . dxc4 2 1 .bxc4 Wfh5 22.g4 Wfa5 Black
dominates the light squares and is also well
placed to attack the hanging pawns.
5
19 ... llJxd3 20.gdxd3
4
After 20.Ei:exd3 ctJe4 2 1 .Ei:c2 f6 22.ctJf3 dxc4
(22 . . . Ei:d7 is also strong) 23.bxc4 b5! Black
secures his domination over the light squares.
24.Ei:b3 (24.c5 �c6) 24 . . . �d5 25.Ei:xb5 ctJd6
Black takes over.
8
7
3
2
1
a
b
c
d
e
f
g
h
16 ....ib4!
Positions with hanging pawns are similar
to those with an isolated pawn, in the sense
that the opposing side should generally try
to exchange minor pieces. Normally one
would be hesitant to exchange a bishop for an
enemy knight, but since White has weakened
his kingside with g3 , Karpov has devised a
powerful strategy based on playing on the light
squares.
17.ge3 hd2!
Removing a defensive piece makes the latent
power of the b 7 -bishop become real.
20 ... llJe4
Winants probably did not anticipate the
main point of this move.
2 1 .Ei:dl?!
It turns out that the rook is vulnerable on
this square.
The best chance for White to keep his position
together was 2 1 .�c 1 ! Wfh5 22.g4 Wfh4 23.Ei:d 1
ctJg5 24.Ei:ed3 f6 25 .�xg5 Wfxg5 26.ctJf3 Wff4
although White is under pressure here too.
8
7
6
lSJ3xd2?
Winants either missed Karpov's reply, or he
did not fully appreciate its power.
1 8 .ctJxd2 was essential, although following
1 8 . . . dxc4 ( I 8 . . . ctJb4 does not achieve much
after 1 9.�b l ) 1 9 .ctJxc4 Wfh5 20.a3 (20.�e2
Wff5) 2o . . . Wfd5 Black has the upper hand.
5
4
3
2
a
b
c
d
e
f
g
h
lS ... llJb4!
2 1 ...£6!
Suddenly White's position is on the brink of
collapse.
19.1lJe5
The knight was well placed on e5, where
it defended several vulnerable light squares.
Once it is driven away, White will not be able
to defend all his weaknesses.
White cannot save the bishop with 1 9 .�b I ?
as 1 9 . . . dxc4 20.bxc4 �xf3 wins a piece.
22.llJg4 'lWh5 23.llJh2?
1 98 6
Luc Winants - Anatoly Karpov
This loses quickly, but the game was already
beyond saving.
If 23.'it>h2 dxc4 24.bxc4 f5 25 .tLle5 tLlxf2
26J''k 1 tLl e4 Black wins.
Also after 23.f3 tLlxg3 24.\Wg2 Wfh4 25 .E:xe6
E:e8! 26.E:de 1 E:xe6 27.E:xe6 'it>f7! 28 .E:e 1
tLlh5! Black wins material.
23 dxc4 24.bxc4
..•
8
7
6
5
4
3
11
Karpov's next tournament was in Bugojno,
the scene of his 1 978 match versus Korchnoi.
He started by drawing with Spassky, then
won a fine game against Yusupov. He drew
comfortably with black against Timman, but
then suffered an unpleasant defeat on the black
side of a Zaitsev Ruy Lopez against Andrei
Sokolov. This turned out to be one of the
most beneficial losses of Karpov's career, as it
motivated him to learn the Caro-Kann, which
became a mainstay of his repertoire and served
him well for many years.
Karpov immediately bounced back with
a long endgame win over Ljubojevic, draws
with Portisch and Miles, then a fine win over
Spassky, which is referenced in the notes to
Game 1 4 (Karpov - Mikliaev) of the first
volume. Then he drew with Yusupov, won a
hard fought game against Timman, and drew
his final three games to win the tournament
with 8 Y21 1 4 , a point clear of Sokolov.
2
a
b
c
d
e
f
g
h
24 ... E:xc4!
Exploiting the loose rook on d 1 . White is
completely busted.
25.d5 E:xd5 26.llJg4 llJg5
0-1
Winants' level has never been world class,
but he is a good player and the way Karpov
dismantled him was highly impressive.
In the final three rounds Karpov defeated
Seirawan, Zapata and Miles. He finished with
an unbeaten 9/ 1 1 , winning first prize by a rwo
point margin ahead of Korchnoi. In many of
his earlier tournaments Karpov had settled for
draws in some games, provided they suited
his tournament strategy. But by this stage in
his career, it looks as though his rivalry with
Kasparov may have motivated him to strive for
an even higher level.
Third World Championship match
versus Kasparov
AI> was customary in those days, the defeated
champion was granted the right to a rematch.
Once again it was contested over rwenty four
games. The first rwelve games took place
in London and the remaining rwelve in
Leningrad. It was the first time that a world
championship match berween rwo Soviet
contenders did not take place solely within the
Soviet Union. It was probably symptomatic of
the fact that the superpower was in decline.
The overall weakening of the Soviet Union had
an impact on chess, as the state was unable to
offer the same level of economic support to its
most talented players.
On Kasparov's initiative, both players
generously agreed to donate the prize fund
from the London half of the match to help
the victims of the Chernobyl disaster, which
occurred in April of 1 986.
12
The Prime Years
Kasparov had not taken part in any tournaments
since winning the world title. He performed
a number of simultaneous exhibitions, and
played two training matches as a means of
sharpening his mind in preparation for his
next clash with Karpov. In December 1 985 he
won 4-2 against Timman, scoring three wins,
two draws and one loss. Six months later he
faced Miles, and destroyed the Englishman by
the score of 5 Y2-Y2. Miles never recovered from
that experience and was unable to maintain his
high world ranking.
The first few games of the match revealed
that both Kasparov and Karpov had worked
on their weaknesses. In Game 1 Kasparov
surprised Karpov with the Griinfeld, and the
latter took no risks and simplified to make a safe
draw. In the second game Karpov introduced
a novelty in the g3 Nimzo-Indian and
equalized comfortably. But later, in a queenless
symmetrical endgame, Kasparov masterfully
outplayed him although he later missed a win
and allowed Karpov to escape. The game was
a shining example of how both players made
remarkable improvements in areas where the
other was strong. Kasparov "taught" Karpov to
prepare better openings, and Karpov "taught"
Kasparov how to play endgames and simplified
positions.
In Game 3 Karpov tried a kingside fianchetto
against the Griinfeld. He got a symmetrical
position with a slight initiative, but Kasparov
drew comfortably. In Game 4 Karpov deviated
from Game 2 but soon found himselfin a worse
position, and this time Kasparov successfully
converted his advantage.
Here is the next game, in which the opening,
middlegame and endgame all merged into
one.
I Gamel I
Anatoly Karpov - Garry Kasparov
London/Leningrad (5) 1 986
l.d4!
Having failed to make any headway against
Kasparov's Scheveningen, Karpov gave up 1 .e4
completely against Kasparov, and only rarely
played it against other opponents as well. It
meant he was able to channel all his efforts into
his closed openings. His opening strategy with
the white pieces was a success in the present
match, although in their subsequent matches
it might not have been a bad idea to throw in
an occasional l .e4 to keep Kasparov guessing.
1. .. lLlf6 2.c4 g6 3.ttlc3 d5
Kasparov later revealed that if Karpov had
opened with 1 .d4 in the last game of their
previous match, he would have responded with
the Griinfeld. Karpov was a 1 .e4 player for
most of his career, so he still had only limited
experience against this opening.
a
b
c
d
e
f
g
h
4.i.f4
In the first two Griinfeld games Karpov was
not able to achieve much, but this time he
comes prepared with a different line. Later in
the match Kasparov was able to find a reliable
13
Anatoly Karpov - Garry Kasparov
1 9B6
antidote to this system, after which Karpov
never repeated it.
Towards the end of the match Karpov switched
to the following line:
4.ttJf3 j,g7 5 .Wb3
In the following game we see how well
Karpov was able to handle complications.
5 . . . dxc4 6.Wxc4 0-0 7.e4 ttJ a6 B .j,e2 c5 9.d5
e6 1 0.0-0 exd5 I l .exd5 j,f5 1 2.j,f4 Ei:eB
1 3 .Ei:ad l ttJ e4
a
b
c
d
e
f
g
h
27.j,xg5 j,xfl 28.�xfl ttJ d6 29.j,e7!
White obtained a winning position thanks to
his powerful passed pawn, Karpov - Kasparov,
London/Leningrad ( 1 9) 1 986.
4...i.g7 5.e3 c5 6.dxc5 %Va5 7Jk l ttJe4
In the next two Griinfeld games Kasparov
switched to the main line of 7 . . . dxc4 and drew
solidly.
a
b
c
d
e
f
g
h
l 4. ttJ b 5 ! ?
This was Karpov's novelty.
1 4 . . . Wf6 1 5 .j,d3
In a subsequent game White deviated:
1 5 .d6 j,d7 1 6.g3 g5 1 7.j,e3 h6 I B . ttJ c7
ttJxc7 1 9.dxc7 j,c6 20.ttJd2 Ei:acB 2 1 .ttJxe4
Ei:xe4 22.Wxc5 Ei:xc7 Black equalized and a
draw was soon agreed, Beliavsky - Kasparov,
Moscow 1 9B7.
1 5 . . . ttJ b4?
A year later in a training game Kasparov
unveiled an improvement:
l 5 . . . Ei:ad8!
1 6.Ei:de l Wxb2 1 7. ttJ c7 ttJxc7 I B .j,xc7 ttJ d2
1 9 .:gxeBt Ei:xeB 20.ttJxd2 Wxd2 2 1 .j,xf5
gxf5 22.g3 j,d4 23.d6 :ge l Black went
on to win in Janjgava - Kasparov, Baku
1 987.
1 6.ttJc7 ttJxd3 1 7.ttJxe8 Ei:xeB I B .Wxd3 Wxb2
1 9 .:gde l Wb4 20.ttJd2 Wa4 2 1 .Wc4 Wxc4
22.ttJxc4 j,c3 23.ttJd2 j,xd2 24.j,xd2 j,d7
25 .j,f4 j,b5 26.f3 g5
8.cxd5 lLlxc3 9.%Vdl %Vxa2 IO.bxc3
The exchange sacrifice 1 0.Ei:xc3! ? is possible,
but at the time of the present game it had
hardly ever been tried. Karpov played plenty
of novelties in his career, but not many of them
occurred in the early stages of the opening.
a
b
c
d
e
f
g
h
IO %Vxdlt
•.•
Kasparov deviates from the play of Fischer,
who preferred 1 O . . . Wa5 in the second game of
14
The Prime Years
his candidates final match against Petros ian in
Buenos Aires 1 97 1 . This game was especially
notable as it culminated in Fischer's first defeat
after twenty consecutive wins against top class
opponents.
The decision to enter an endgame at such
an early stage shows that Kasparov was fully
confident in the improvements he had made
since the previous two matches.
1 996.) 1 8 .lbf3 (White scores very well with
1 8 .f4 too.) 1 8 . . . b6 1 9.c5 bxc5 20.Ei:xc5 Ei:xc5
2 1 .ixc5 White got a clear advantage and went
on to win in Seirawan - Adorjan, New York
(rapid) 1 987.
1 5 ... Ei:fc8!? has hardly ever been tried. One
interesting reply looks to be 1 6.ie3!? ixe5
1 7.lbf3 ( 1 7.f4 ig7 1 8 .lbf3 e6) 1 7 . . . ig7 1 8 .Ei:b 1
Ei:c7 1 9 .id4 and White is somewhat better.
1 1 .@xd2 liJd7 12 ..ib5!
Karpov wants to eliminate the knight in
order to give back the c5-pawn under his own
terms.
16.e4 E:fe8
12 ... 0-0 13.i.xd7
After the greedy 1 3 .c6?! lbc5 1 4.\t>e2 bxc6
1 5 .ixc6 ia6t 1 6.c4 lbb3 Black is too active.
13 ...i.xd7 14.e4 £5
Kasparov
follows
the
theoretical
recommendation, although it carries a certain
risk as if he is not able to undermine his
opponent's centre, the powerful pawns could
suffocate him.
14 . . . Ei:ac8 fails to equalize after 1 5 .ie3 e6
1 6.lbe2 Ei:fd8 1 7.c4 if8 1 8 .lbf4 ixc5 1 9.ixc5
Ei:xc5 2o.lbd3 when White dominates the
centre.
15.e5!
Karpov gains space and blocks the g7-bishop.
The question is what Kasparov can do with the
time Karpov needs to consolidate his centre.
15 ...e6
This was Kasparov's novelty, but it fails to
impress.
The most common move is 1 5 . . . Ei:ac8 ,
when play may continue 1 6.c4 Ei:xc5 1 7.ie3
Ei:c7 (After 1 7 . . . Ei:a5 1 8 .lbf3 f4 1 9.id4 Ei:c8
20.Ei:a l White's powerful centre gives him a
clear advantage, Ostenstad - Elsness, Gausdal
a
b
c
d
e
f
g
h
17.e6!
The pawn was doomed anyway, so Karpov
takes the opportunity to alter Black's structure
in a way that fractures his queens ide pawns
and obstructs his rook and bishop.
17 ... bxe6 1 8.d6 e5?
This natural move is a mistake. In a few
subsequent games Black improved on
Kasparov's play with 1 8 . . . g5 ! , which is much
more challenging although it still does not
equalize: 1 9.ixg5 ixe5 20.c5 ig7!? (After
20 . . . Ei:cb8 2 1 .lbf3 ig7 22.E:he 1 White will be
able to block the a l -h8 diagonal with lbe5 .)
2 l .f4 Ei:cb8 22.lbf3 h6 23 .ih4 E:b2t 24.Ei:c2
E:b4 25 .Ei:hc l Ei:xf4 26.Ei:b 1 White obtained
excellent compensation for the pawn in
Timman - Popovic, Belgrade 1 989.
IS
Anatoly Karpov - Garry Kasparov
1 986
19.h4!
Karpov immediately prevents . . . gS .
19 ... h6
8
7
6
5
4
a
3
c
d
e
f
g
h
24.tlJd3
2
1
a b
20.tlJh3!!
b
c
d
e
f
g
h
Karpov embarks on a masterful plan to freeze
the Griinfeld bishop. Kasparov and his team
missed this idea in their home preparation. It
is possible that Karpov found the whole plan
at home.
20 ... a5
Kasparov won numerous great games by
utilizing his pawns on the edge of the board,
but this move is a sign of desperation. Black has
to take radical action of some sort, otherwise he
is helpless against White's plan of transferring
his knight to d3, followed by winning the cS­
pawn.
The other main candidate move is 20 . . .Ekb8,
but after 2 l .f3 E!b2t 22.E!c2 E!xc2t 23.c;t>xc2
ia4t 24.�d2 E!b8 2S .E!al E!b2t 26.�e l
White is winning according to Kasparov.
21.0 a4 22J�he l !
Before continuing his knight manoeuvre,
Karpov overprotects the e5-pawn in order to
take the sting out of . . . gS .
22 ... a3 23.tlJfl a2
Finally the knight reaches its ideal home.
From d3 it reinforces the eS-pawn, attacks the
cS-pawn, and defends the b2-square.
24.. J�a3 25.gal
Black's last hopes rest with the a2-pawn, so
Karpov intends to remove it.
25 ... g5?
This sacrifice achieves very little.
25 . . . E!b8!
This active move was the last real chance
for Black to create problems. Interestingly
Kasparov did not mention it in his first book,
which suggests that even he needed the help
of a computer to appreciate the potential
strength of Black's counterplay.
26.c;t>e2!
White should find this and the following
prophylactic moves to neutralize Black's
play.
26.E!ec l g5 ! 27.hxgS hxgS 28 .ih2 (28 .ixgS
E!bb3 29.ctJxc5 E!b2t Black became so active
he could sacrifice a piece.) 28 . . . f4 and Black
is in the game.
26 . . . ic6!
With this tricky move Black ensures that
his bishop will not be in the firing line after
ctJxcS .
16
The Prime Years
Kasparov mentions the line 26 . . . 'it>f7 27.i.e3
:gba8 28.ttJxc5, opining that the outcome is
not completely decided, but it seems to me
that the chances of surviving the endgame a
clear pawn down against Karpov would be
minimal.
28"J:�b8
a
a
b
c
d
b
c
d
e
f
g
29J3ecl!
e
f
27.:gec 1 !
White has to find this subtle move i n order
to maintain his advantage.
The alternatives 27.ttJxc5 ?! :gb2t and
27.:ged l i.a4 are worse.
27 . . . :gbb3 28.:gdl :ga5
After 28 . . . :ga4 29.ttJxc5 :gb2t 30.�e3 :ga3t
3 1 .�d4 :gxg2 32.ttJxe6 White wins.
29.:gd2 :gba3 30.:gb2
White will soon win a pawn, which should
be enough to decide the game.
26.hxg5 hxg5 27.i.xg5 <i!lf7?!
This is rather feeble. 27 . . . :gb8 was still the best
chance, despite having lost some of its power
in the present position. Once again White
should respond with 28 .'it>e2!, for instance:
28 . . .i.c6 (or 28 . . . :gbb3 29.ttJxc5 :gb2t 30.�fl
winning) 29 .i.c 1 :ga4 30.i.d2 Black's activity
is neutralized and he is losing.
28 ..if4!
Karpov does not lose sight of his main
priority, which is to maintain his solid defence
of the e5-pawn, thus ensuring that the g7bishop remains dead.
Karpov gets closer t o the a2-pawn and also
prepares to defend against a rook check on the
second rank, thus making ttJxc5 possible.
29".i.c6
29 . . . :gbb3 30.ttJxc5 :gb2t 3 1 .:gc2 wins.
30J3c3 ga5 3 1 .gc2 gba8 32.ttJcl
1-0
With the a2-pawn about to fall, Black's
position is hopeless. Karpov's convincing
refutation of Kasparov's opening novelty
showed how much the former had improved
his openings in the year since the previous
match. The manner in which he coordinated
his pieces while shutting out the g7-bishop was
exemplary.
It is amazing how the picture changed since
the first two matches. Kasparov now had the
ability to tortute Karpov from a dry technical
position with a tiny advantage, while Karpov
was able to destroy Kasparov's formidable
opening preparation.
The next two games were drawn, although
both contained plenty of action. In Game 8
Karpov grabbed a pawn on the queenside
but Kasparov launched a venomous attack
1 986
Anatoly Karpov - Garry Kasparov
on Karpov's king. The ensuing complications
were almost unfathomable, and for a long time
the game was balanced on a razor's edge, but
eventually Kasparov's incredible tactical ability
prevailed and Karpov lost on time in what was
by then a hopeless position.
In Game 9 Kasparov showed that he had
successfully repaired his Griinfeld, and he drew
comfortably. In the next game he once again
demonstrated his ability to squeeze Karpov in
a technical position, but did not quite make
the most of his chances and the game ended
in a draw. In Game 1 1 Karpov came back with
an ambitious new exchange sacrifice in the .tf4
Griinfeld. Wild complications ensued, and
both players missed chances before the smoke
cleared and a draw ensued. In the final game of
London's half of the match, Kasparov obtained
a slight plus but Karpov held the draw without
ever being in real trouble. Thus at the halfway
point Kasparov led by two wins to Karpov's
one, with nine draws.
17
that he was the master of wild complications.
With eight games remaining, Kasparov now
held a commanding 4- 1 lead.
Karpov desperately needed a win, and he
managed to get it in the very next game. He
prepared a strong improvement in the line
featured in the previous Griinfeld game (maybe
Kasparov overdid this opening) , and Kasparov
went down without much of a fight. Game 1 8
was a Nimzo-/Queen's Indian hybrid, in which
Kasparov seized the initiative with an energetic
pawn sacrifice. He built up a decisive attack,
but lost his way in the time scramble and
allowed Karpov to emerge with the advantage,
which he successfully converted in a complex
ending. Suddenly the momentum had returned
to the challenger. In Game 1 9 Kasparov tried a
different approach in the Griinfeld, but again
Karpov was ready and won a nice game, which
is shown in the note to White's fourth move
in Game 2. With five games remaining, the
match was all square!
From London to Leningrad
After the final London game, the players had
eight free days before the second half of the
match, which took place in Leningrad, Russia's
second largest city. Karpov reverted back to
the Fianchetto Variation against the Griinfeld,
but Kasparov held his own and even had the
better chances late in the middlegame, but
was unable to achieve more than a draw. In
Game 14 Kasparov switched to 1 .e4 for only
the second time in the match, and introduced
a novelty against Karpov's Zaitsev Ruy Lopez.
Karpov did not react in the best way and could
only reach a hopeless endgame which Kasparov
duly converted. In Game 1 5 Karpov tried
4 .lLlf3 .tg7 5 .�b3 against the Griinfeld, but
Kasparov equalized with the help of a pawn
sacrifice, and a draw was later agreed. In the
next game Karpov tried to improve his play in
the Zaitsev, but once again Kasparov showed
In Game 20 Kasparov sensibly opted for a quiet
Catalan, took no risks and drew comfortably.
In the next game he wisely abandoned the
Griinfeld, instead opting for a Queen's Indian,
and managed to neutralize Karpov's slight
pressure. It seems that over these two games
Kasparov regained his confidence and his
energy, and he won Game 22 in style. The last
games were drawn without major incident,
and thus Kasparov won the match by the score
of 1 2Y2- 1 1 liz.
It has been noted that, paradoxically, after
winning three consecutive games to draw
level in the match, Karpov seemed to play less
ambitiously. Perhaps fatigue played a role; it
had caused him to suffer a near collapse against
Korchnoi in 1 974, and now at the age of thirty
five, he probably lacked the energy reserves of
his young successor. Kasparov was renowned
for his intensive physical training regime, and
18
The Prime Years
it seems to have benefitted him in the late
stages of all three of the K-K matches that we
have covered thus far.
The London - Leningrad match highlighted
the considerable improvements that both
players had made, and in my opinion it
produced the highest quality chess of any
World Championship match in history.
Kasparov had made tremendous strides in
the way he handled endgames and simplified
positions, but it was primarily his tremendous
ability in complicated positions which won
him the match. Karpov had improved his
openings tremendously since the last match,
and no longer received more than he punched
in that department. Kasparov has since made
it known that he suspected one of his seconds,
Evgeny Vladimirov, of leaking his opening and
even adjournment analysis to Karpov's team.
If this was indeed the case then it would go
some way towards explaining Karpov's success
in the openings, but the allegations were never
substantiated. There are two sides to every
story, and since it is not the purpose of this
book to delve into such matters, I will say no
more about it here.
Overall Karpov scored one win and five
losses with the black pieces. His losses mainly
occurred in hugely complicated middlegames
rather than due to poor opening play, as
occurred in the previous match. With the white
pieces he showed a marked improvement,
scoring three wins and no defeats. Karpov was
seeded directly to the final of the next world
championship cycle, making another rematch
a likely prospect for the following year.
match, he took part in the strong Tilburg
tournament. He began with a quick draw
with Lj ubojevic, then outplayed Portisch
with a masterful endgame performance. He
drew his next seven games, achieving little in
the openings. Interestingly Timman was able
to improve strongly on one line of the Wi'b3
Griinfeld which Karpov had used to defeat
Kasparov in Game 1 7 of their recent match.
In Round 1 0 Karpov overcame Korchnoi,
but then he got into a difficult position against
Beliavsky and blundered badly. He drew his
final three games, to finish in third place with
a somewhat disappointing 7Y2/ 1 4. Although
Karpov's desire to put the Kasparov match
behind him was admirable, with hindsight it is
obvious that he needed a longer break to recover
his energy. Another problem was evident in
his opening play: it looks like he did so much
work preparing against Kasparov's repertoire
that he had no time or energy to prepare other
openings that he would need against lesser, but
still world world-class grandmasters.
A week and a half after Tilburg, Karpov took
part in the Dubai Olympiad. For the first time
in a long while, he played on the second board
behind Kasparov. He drew his first game with
the black pieces against Ftacnik, then met
Ribli. Since Karpov's win in Amsterdam 1 980
(see Game 5 8 of the first volume) , the two
players had drawn two relatively short games.
I Game] I
Anatoly Karpov Zoltan Ribli
-
Dubai Olympiad 1986
***
l .c4
Karpov has stated in interviews that he likes
to forget disappointment, and the best way of
doing it is to play another tournament. Just
eleven days after the end of the championship
Ribli has a relatively narrow repertoire. He
mainly plays the Najdorf, the Queen's Indian
and the Hedgehog against the English. Karpov
decides to test him in the last of the three.
1 986
1...c5 2.<�jf3 lLlf6 3.lLlc3 e6
Ribli deviates from the Amsterdam game, in
which he played 3 . . . lLl c6 4.g3 d5.
4.g3 b6 5.i.g2 i.b7 6.0-0 d6
Ribli deviates from the move order he used
against Karpov in Linares 1 98 1 , although the
game soon ends up in a similar position.
6 . . .ie7 7.d4 cxd4 8 .�xd4 d6
Let us review a few of Karpov's games
against the Hedgehog. From time to time he
also played the black side of this opening as
well. White has two main approaches from the
present position.
a) 9.e4
In an earlier game Karpov did without this
move and allowed an early queen exchange:
9.b3 0-0 1 O.:1l:d l lLl bd7 l 1 .ib2 a6 1 2.�e3
�b8 1 3 . ttJ d4 ixg2 1 4.i'xg2 �b7t 1 5 .�f3
�xf3t 1 6.ttJxf3 The position is equal
although White went on to win in Karpov Gheorghiu, Moscow 1 977. Despite Karpov's
eventual success, it seems he was not satisfied
with the outcome of the opening as he never
repeated this plan.
9 . . . 0-0 1 O.b3 ttJ bd7 1 1 .ib2 a6 1 2.:1l:fd 1 �c7
1 3 .�e3 :1l:fe8 1 4.:1l:ac l if8 1 5 .ttJ d4
In between this example and the main game,
Ribli chose 1 5 . . . :1l:ad8. On that occasion
Karpov did not really try to hunt the
hedgehog, and after a bit of manoeuvring
he settled for a draw: 1 6.:1l:e 1 (Later in the
same tournament Karpov reached the same
position with the black pieces and won a
fine game after 1 6.�e2 in Garcia Gonzales
- Karpov, Linares 1 98 1 ; see Game 59 in the
first volume.) 1 6 . . . �b8 1 7.�e2 g6 1 8 .:1l:cd 1
ig7 1 9.�c2 ttJ c5 20.c;t>h 1 �a8 2 1 .f3 �b8
22.�f2 ia8 23 .ia l Yz-Yz Karpov - Ribli,
Linares 1 98 1 .
1 6.h3 �b8 1 7.:1l:e l g6 1 8 .:1l:cd 1 ttJc5 1 9 .f3 ig7
20.ia 1 ttJ cd7 2 1 .f4 e5 22.ttJc2 exf4 23.�xf4
ttJe5 24.ttJ b4 h5 25 .:1l:n �c7 26.'it> h l :1l:e6
27.�d2 :1l:ce8 28.ttJcd5
Karpov obtained a small edge and went on to
win in Karpov - Gheorghiu, Leningrad 1 977.
b) Karpov also tested a radically different,
though still well-known plan in two games:
9.ig5 a6 1 O.ixf6
By making this exchange before the reply
. . . ttJxf6 is available, White hopes to disrupt
his opponent's coordination.
1 0 . . . ixf6 1 1 .�f4
a
a
b
c
d
e
19
Anatoly Karpov - Zoltan Ribli
f
g
h
1 5 . . . :1l:ac8
This was Gheorghiu's choice in a subsequent
Hedgehog game against Karpov in 1 977.
b
c
d
e
f
g
h
1 1 . . . ixf3
This was the choice of Walter Browne, two
years after another Karpov game, which
proceeded as follows: 1 1 . . . 0-0 1 2.:1l:fd l ie7
1 3 . ttJ e4 ixe4 1 4.�xe4 :1l:a7 1 5 .ttJ d4 �c8!
20
The Prime Years
1 6.b3 �e8 1 7.a4!? �c5 1 8 .�a2 i.f6 1 9.�ad2
�c7 20.�b l ! i.e7! 2 1 .b4 (2 1 .�c2 �b4!)
2 1 . .. �h5
1 6.�e3! �d7
If 16 . . . �e7 1 7.cxb5 axb5 1 8 .a3 White has a
nice positional advantage as the b5-pawn is
vulnerable.
1 7.cxb5 axb5
8
7
6
5
4
a
b
c
d
3
e
f
g
h
This was Karpov - Kasparov, Moscow 1 98 1 .
At this point White could have obtained
some advantage with 22.b5! (The game
continued 22.�c2?! �ec8 23 . b 5 ? axb5
24.axb5 �xc4! when the tactics favoured
Black, and Karpov had to struggle for a draw
a pawn down.) 22 . . . �xc4 (22 . . . a5 23 . lLl c6 is
better for White) 23.bxa6 �a5 (23 . . . �xa4
24. lLl c6!; 23 . . . lLlxa6 24.i.c6) 24.�b5 �xa4
25.a7 �xd4 26.�xe8t �xe8 27.a8=� White
can press.
1 2.i.xf3 �a7 1 3 .�fd l i.e5 1 4.�d2 0-0
1 5 .�ac 1
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
1 5 . . . b5 ?!
This thematic freeing move is premature.
Better is 1 5 . . . �c7 16.b3 lLl c6 when White's
advantage is small.
2
a
b
c
d
e
f
g
h
1 8 .b4!
Karpov fixes the weak b5-pawn.
1 8 . . . i.xc3 1 9 .�xc3 d5 20.�c5 �f6 2 1 .�dc 1
d4?
In a worse position Browne tries to compli­
cate the game with a pawn sacrifice, but this
only accelerates the end.
2 1 . . .�b7 was better, although after 22.�d6
�d8 23 .�xd8 �xd8 24.�c8 �f8 25 .e4 d4
26.e5 Black faces a difficult defence.
22.�d3 e5 23 .�xb5 �e6 24.�c5 �e7 25 .i.d5
Karpov obtained a winning position and
soon collected the full point in Karpov Browne, Bath 1 983.
21
Anatoly Karpov - Zoltan Ribli
1 986
7.d4
Nowadays 7.�e l is the main line, intending
e4 followed by d4, when White will be able to
recapture with the knight without allowing a
bishop exchange on g2.
7 ...cxd4 8.�xd4 a6 9.�dl �e7 IO.b3 .!iJ bd7
l l .e4
difficult endgame for Black; 1 4 . . . WcB!? may
be the best chance.) 1 5 .Wxc7 Ei:xc7 1 6.ic 1 !?
White is on top.
Before we return to the main game, let me
show you a beautiful example of winning
technique in a similar ending, courtesy of a
true chess legend.
By moving the e-pawn forwards at this stage,
Karpov retains the options of both ia3 and
ib2.
a
a
b
c
d
e
f
g
h
1 l ...�c8!
b
c
d
e
f
g
h
20. g4! a6 2 1 .g5 ttJe8 22. a4 E1a7 23.h4 E1b7?!
(23 . . . i>fB) 24.E1d3 E1c5 25.E1e l E1d7 26.E1ed l
g6 27.ttJe2 E1xd3t 28.E1xd3 b5 29 . cxb5 axb5
30.E1d7 i>f8 3 1 . a5 E1c6 32.E1b7 b4 33.E1b8!
�c5 34.ttJg3 1 -0 Bronstein - Rantanen,
Tallinn 1 97 5 .
Ribli introduces a strong novelty which
has since become established as the main line
in this position. Later Karpov played it with
Black, and overall the second player has a
plus score with it. The main idea is to avoid
the unpleasant endgames that can occur after
White's plan of ia3 and e5 followed by a
queen exchange. For instance:
1 1 . . .0-0 1 2 .ia3! lLl c5 1 3 .e5 dxe5 1 4 .'1Wxd8
�fxdB 1 5 .lLlxe5 ixg2 1 6.�xg2 White has
made a heavy plus score from this position.
I can recall Peter Szekely beating Ribli from
a similar endgame in a ten minute blitz
tournament in Hungary.
12.�b2
l 1 . ..'IWc7 1 2.ia3! lLl c5 1 3 .e5 dxe5 1 4.'1Wxe5
Ei:cB ( 1 4 . . . Wxe5 1 5 .lLlxe5 ixg2 1 6.'tt> xg2 is a
With the queen on cB instead of c7, 1 2.ia3
does not achieve much: 1 2 . . . lLl c5 1 3 .e5 ixf3
1 4.ixf3 dxe5 1 5 .Wxe5 lLl cd7 1 6.Ei:xd7 lLlxd7
a
b
c
d
e
f
g
h
The Prime Years
22
1 7.'\Wxg7 �f6 I B .ttJe4 �xg7 1 9. ttJ d6t c;t>dB
20.ttJxf7t <j;>eB 2 1 .ttJd6t c;t>dB The game ends
in an unusual perpetual.
12 ... 0-0 13.tlJd2
Karpov decides to deviate from the plan
involving '\We3 and ttJ d4, which he had used
previously against both Gheorghiu and Ribli.
With his chosen move he defends the c4-pawn
and prepares b4, and will decide later whether
to reposition his knight on the kingside or the
queenside.
More than a decade and a half later, Karpov
showed that he is equally adept on the black
side of the position: 1 3 .!"lac l !"leB 1 4.h3 '\Wc7
1 5 .'\We3 �fB 1 6.ttJd4 !"lacB 1 7.g4 h6 I B .'\Wg3
'\WbB 1 9.h4 ttJ c5 20.!"le l '\WaB 2 1 .'\Wf3 ttJ fd7
22.!"lcd l ttJe5 23 .'\Wg3
a
b
c
d
e
f
g
h
Ribli decides to wait patiently.
19.�e2
Karpov defends the c4-pawn, thus freeing
the knight on d2.
19 ... tlJe8 20.ltJfI i.b7 21 .i>h2 tlJef6 22.tlJd2
tlJe8 23.:8e l i.a8
8
7
6
5
a
b
c
d
4
e
f
g
h
23 . . . g5 !? 24.hxg5 hxg5 2 5 . ttJ f3?! ttJ ed3! 26.!"le2
ttJxb2 27.!"lxb2 �e7 Black's control over the
dark squares gave him somewhat better chances
in Van Wely - Karpov, Dubai (rapid) 2002.
13 ...�c7 14J�acl :8ac8 15.h3
With this and his next few moves, Karpov
creeps forwards and gains a little more space.
15 ... :8fe8 16.a3 �b8 17.b4 :8ed8 18.�e3
Karpov does not want his queen opposite
the enemy rook.
18 ...i.a8
3
2
1
a
b
c
d
e
f
g
h
24.tlJ b3!
After a few waiting moves Karpov brings
the knight to the queenside. Later the pawn
advance a4-a5 could prove unpleasant for
Black.
24 ... i.g5 25.:8c2 :8c7 26.f4 i.f6 27.:8ecl
:8dc8
23
Anatoly Karpov - Zoltan Ribli
1 986
Ribli ties Karpov's pieces to the defence of
the c4-pawn.
35.gdl!
Karpov focuses his heavy pieces against the
d6-pawn.
28.ltJdl!
Karpov is happy to exchange bishops,
as the d6-pawn will become slightly more
vulnerable.
28 ... i.b7
28 . . . �e7!? was worth considering.
35 ... tlJdf6 36.ged2 h6
Ribli decides to wait. Another idea
was 36 . . . �c6 intending to play for . . . e5.
Nevertheless after 37.VNe2 e5 38 .�f3 h6
(38 ... E1d7 39.g4) 39.c;t>g2 E1d7 40.f5 White is
clearly better, j ust like in the game.
29.VNd.3 i.xb2?!
Exchanging the bishop helps White. Ribli
mentions 29 . . . VNa8 as a good alternative.
30.ltJxb2 VNa8 3 1 .ltJdl a5?
This is a most uncharacteristic mistake for
Ribli, who was not only a fine positional player,
but also an expert on Hedgehog positions.
8
7
a
6
b
c
d
e
f
g
h
37.VNe2!
5
4
3
2
1
a
b
c
d
e
f
g
h
32.ltJd4!
Naturally Karpov wastes no time 10
transferring his knight to the newly created
outpost. Ribli had apparently forgotten that
this knight could find its way to the b5-square
as well.
32 J�d8 33.ltJb5 gee8 34.ltJdc3 VNb8
.•
It was worth considering 34 . . .t2Jb8!? to
transfer the knight to c6, which would at least
force White to pay attention to the b4-pawn.
Karpov continues to squeeze his opponent.
Now e5 becomes a threat.
37.e5?!
The immediate advance of the e-pawn is
tempting, but premature.
37 . . . dxe5?
This is the only option mentioned by Ribli
in his Chess Informant analysis, but it is not
the strongest.
Correct is 37 . . . �xg2! 38.exf6 �c6 39.fxg7
axb4 40.axb4 d5 4 1 .cxd5 �xb5 42.lLlxb5
E1xd5 43.VNe2 E1xd2 44.E1xd2 'lWb7 when
White will have a hard time making progress
due to his exposed king.
38 .'lWxd8 �xg2
Ribli ends his analysis here, evaluating the
position as unclear. It turns out White has a
wonderful resource lurking.
39.'lWe7 axb4
24
The Prime Years
Objectively Black should prefer 39 . . . iJ3! in
order to divert the rook away from the d-file,
although after 40J::!: f1 ia8 4 1 .fXe5 W'xe5
42.l::i:d 8 White remains dearly better thanks
to his extra exchange.
40.axb4 ia8
8
7
6
5
4
3
2
a
b
c
d
Black should have preferred 37 . . . e5! to gain
space, and after 38.if3 ic6 39.f5 we reach
something dose to the game continuation.
3S.c;t>gl ?
Karpov wants to put the time control
behind him before taking direct action. It is
understandable, although objectively he should
have preferred 38.bxa5 ! bxa5 39.l::i:b 2 when
Black is in serious trouble as tt:lxd6 is a huge
threat. Play might continue 39 . . . ib7 40.e5
dxe5 4 1 .fXe5 ixg2 42.tt:l d6 if3 (42 ... W'c7
43.tt:lxc8 wins.) 43.l::i:x b8 ixe2 44.tt:lxc8 and
White wins.
3S ....tc6 39.c;t>h2 e5
e
f
g
h
4 1 .f5!!
With the help of this beautiful pawn sacrifice
White keeps his king safe.
4 1 . . .exf5 42.l::i:d 8 f4 43.gxf4 exf4
43 . . . e4 44. tt:l e2 is safe enough.
44.l::i:xc8 W'xc8 4 5 .l::i:d 8 W'c6
Black is not helped by 45 . . . W'a6 46.W'a7.
46.tt:l d5
White is winning.
a
b
c
d
e
f
g
h
40.£5!
Karpov gains more space and prepares a
kingside pawn storm.
40...\WaS 41 .g4!
a
b
c
d
e
f
g
h
37 ....taS?
If 37 . . . ic6 38.e5 ixg2 39.exf6 ic6 40.fXg7
the d6-pawn is a big problem for Black.
There is no sense in pawn-grabbing with
4 1 .tt:lxd6 tt:lxd6 42.l::i:xd6 l::i:xd6 43.l::i:xd6,
due to 43 ... axb4 44.axb4 W'a3 45 .tt:ld5 ixd5
46.exd5 W'xb4 when Black is not worse.
41 ...\WbS 42.h4 \Wb7 43.i.f3
Karpov continues to strengthen his
position without giving his opponent any
counterchances. 43 .g5 would have been
1 986
25
Anatoly Karpov - Zoltan Ribli
premature due to 43 . . . hxg5 44.hxg5 'tJh7
45.g6 'tJhf6 when White can win the d6-pawn
but his position is somewhat loose.
43 JWe7
.•
48.g5
4B .'tJa7?! does not work with the king on g3 ,
due to 4B . . . 'tJd6! 49.'tJxcB 'tJdxe4t 50.i.xe4
'tJxe4t 5 1 .lfif3 'tJxd2t and Black gets back
into the game.
48 ... hxg5?
8
Ribli is a player with excellent defensive skills,
which your author has experienced firsthand.
However, on this occasion his sense of danger
lets him down. 4B . . . 'tJh7 was somewhat better,
though ultimately still insufficient: 49.gxh6
(49.'tJa7 is less convincing due to 49 . . . hxg5)
49 . . .'tJd6 (49 . . . gxh6 5 0.'tJa7 wins) 50J�b2
The blockade does not work and White is
winning.
7
6
5
4
3
2
a
b
c
d
e
f
g
h
44.�g3
Karpov purposefully builds his attack. He
is already thinking about a potential mating
attack on the h-file, so he opens the second
rank in order to facilitate the transfer of heavy
pieces.
44.lfig2!? Ribli assesses this move as winning
for White, although he does not give any
further analysis. He probably had in mind the
variation 44 . . . axb4 45 .axb4 d5! 46.cxd5 i.xb5
47.'tJxb5 'lMfxb4? 4B .'tJa7! when White wins an
exchange. However, Black can improve with
47 . . . 'tJd6! 48 .'tJxd6 'lMfxd6 when he has some
chances to hold, just like in the note to move
47 below.
44...axb4 45.axb4 d5 46.cxd5 i.xb5 47.lLlxb5
'lWxb4?
This move restores the material balance, but
leaves his kingside unprotected.
Ribli later recommended playing for a blockade
with 47 . . . 'tJd6!, when Black has fair chances to
keep his position together.
Black's most resilient defence was 4B . . . 1"k4! .
This does not solve all the problems, b u t i t
keeps him in the game: 49.gxf6 (49 .:gb2 'tJxe4t
50.i.xe4 [50.<;t>g2 'lMfa4] 50 . . . :gxe4 5 1 .:gxb4
:gxe2 52.d6 White has good compensation
for the pawn, but Black has decent chances
to hold.) 49 . . . 'lMfxb5 50.fxg7 'lMfc5 (50 . . . :gb4
5 1 .Wg2) 5 1 .<;t>h2 Ifixg7 52.Wg2t <;t>fB Black's
position remains difficult, but it is far from
hopeless as White's bishop is rather passive.
49.hxg5 lLlh7
8
7
6
5
4
3
2
a
50.d6!
b
c
d
e
f
g
h
The Prime Years
26
Karpov threatens to win a piece, while also
taking away the e7-square from the black king.
50 .. Jk5
After 50 . . . tDxg5 5 1 .d7 Elc6 52.dxe8=lMrt Elxe8
53 .�g4!? tDh7 54.tDd6 Eld8 5 5 .lMrb5 White
keeps the extra piece.
5 1 .l3b2!?
Karpov plays pragmatically. 5 1 .lMrh2 Elxb5
52.Elh 1 �f8 53.lMrxh7 Elxd6 54.f6! gxf6 5 5 .g6!
fxg6 56.lMrxg6 is mating, but why sacrifice a
rook to checkmate when one can achieve the
same result without it?
whole board: he activated his knights on the
queens ide, piled up his rooks on the d-file,
and eventually finished his opponent on the
kingside.
After this fine win Karpov suffered a setback
and lost to Ljubojevic. He drew his next game
with Black against Hjartarson, then faced
Haik of France. Apart from the present game,
Haik only faced one other world champion
in Spassky, with whom he drew on two
occasions.
I Game 4 I
Aldo Haik Anatoly Karpov
-
5 1 ...V9c4 52.V9h2!
Since move 40 Karpov has been playing for
mate on the kingside, and his strategy is about
to prevail.
Dubai Olympiad 1986
l .e4 c5
The two time French champion was not
usually a 1 .e4 player, so Karpov had at least
two good reasons to play a Sicilian: he wanted
to play for a win, while also avoiding his
opponent's preparation.
52 ...l3xb5
v!l�f3 e6
Karpov employed the Kan variation a few
times during his championship reign, but he
only seldom used it in the eighties.
3.<13
a
b
c
d
e
f
g
h
53.V9xh7t!
Karpov finishes the game in style, sacrificing
his queen to break through on the h-file.
This is a somewhat questionable choice
against one of the great masters of closed
positions. Furthermore, the King's Indian
Attack was sometimes used by Fischer, and
it is possible that Karpov prepared for it in
anticipation of the 1 975 world title match.
3 ttJc6 4.g3 d5 5.Vge2
••.
53 ... i>xh7 54.l3h2t i>g8 55.l3dhl
56.l3h8t
1-0
f6
Ribli decided not to allow Karpov to
checkmate with a pawn in the next move.
Karpov played excellently and utilized the
Haik probably played this move in order to
avoid theory. He may well have been aware of
the following game:
5 .tDbd2
Karpov faced this move twice.
5 . . . �d6 6.�g2 tDge7 7.0-0 0-0 8.Elel lMrc7
1 986
27
Aldo Haik - Anatoly Karpov
A year earlier Karpov used 8 . . . ic7 to defeat
Sznapik, but his play in the opening was not
convincing.
9.b3 id7 1 O.ib2 d4 I l . lLl c4 e5 1 2.a4 b6
1 3 .1.Wd2 f6 1 4.h4 Wb8 1 5 .ia3 ic7 1 6J''1 e b l
ie6 1 7.Wh2 Wc8 I B .We2 ig4 1 9.Wfl f5 !
20.tLlcd2
5 ... ttJge7
The other main line is 5 . . . lLl f6 intending
. . . ie7, but Karpov opts for a more dynamic
set-up involving a kingside fianchetto.
6.i.g2 g6 7.h4 h6!
Karpov remains flexible and defuses the
further advance of the h-pawn. Naturally he
avoids 7 . . . h5?, which would weaken his dark
squares and leave him static on the kingside.
8.e5
This move was first played by Torre, who
used it to beat Tukmakov. Despite this success,
the move is a bit premature and it is no longer
played by experts.
a
b
c
d
e
f
g
h
20 . . .f4!
Karpov wins additional space.
2 1 .ih3 h5 22.Wg2 tLl g6 23.tLlg5 id8 24.tLl gf3
fie7 25.!"lg l We6 26.!"lafl !"l£7 27.!"lh l !"lafB
2B.Wgl Wd6 29.Wh2 a6!
Karpov gets ready to open a second front on
the queenside. He never gets to carry out his
plan, as White makes a serious mistake and
collapses on the kingside.
30.Wgl !"lf6 3 1 .ixg4?? hxg4 32.tLlg5 f3
33.1.Wh2 tLlhB 34.!"lc l !"lh6 3 5 . tLl c4 Wc7
0- 1 Calvo Minguez - Karpov, Madrid
1 973.
8 ...i.g7
This is more natural and stronger than
Tukmakov's B . . . lLl d4.
9.c3 V!fc7 IO.i.f4 b6 1 l .a4
White can save his e5-pawn with I I .d4,
but this is unlikely to worry Black: 1 1 . . .cxd4
1 2. tLl a3 ( 1 2.cxd4 lLl b4! is good for Black)
1 2 . . . Wd7!? 1 3 .cxd4 ( 1 3.lLlb5 ia6) 1 3 . . . a5
1 4.lLlb5 ia6 1 5 .a4 0-0 Black has a good grip
on the queens ide.
I I . lLl a3 ! ? was also possible, although both
1 1 . . . ia6 and 1 1 . . . a6 give Black a full share of
the chances.
a
b
c
d
e
f
g
h
28
The Prime Years
1 1. .. gS!
Karpov exchanges his g-pawn for White's
e-pawn. This gives him the advantage of a
central majority, as well as one less pawn
island.
12.hxgS hxgS 13J3xh8t i.xh8 14.hgS
liJxeS IS.liJbdl
1 5 .lLl a3 should be met by 1 5 . . . lLJ5g6! (but
not 1 5 . . . a6? 1 6.lLJxe5 �xe5 1 7.lLJc4 when
Black has problems as the natural 1 7 . . . �g7
1 B .�xe7 wins a pawn) .
White probably wanted to simplifY with f4
and lLJe5 , but he does not have enough time
to do it.
The most logical continuation looks to be
2Uttd2 �d7 (2 1 . . .�a7 22.lLJe3) 22.lLJe3 lLJa5
(22 . . . lLJe5 ?! 23.c4) 23.�h 1 when White should
be safe.
8
7
6
5
IS ...i.a6 16.lLlxeS
4
8
3
7
2
6
1
5
a
4
b
c
d
e
f
g
h
2 1 ..J3a7!
3
The rook steps onto a defended square while
vacating the long diagonal.
2
a
b
c
d
e
f
g
h
16 .. .'�xeS
Karpov has to exchange queens, but that was
not usually against his will. 1 6 . . . �xe5 1 7.�xe7
�xe7 1 B . f4 �f6 1 9 .�xd5 �xd3 20.1Wxd3 �dB
2 1 . 0-0-0 reaches an equal position in which
Black has little to play for.
17.'�xeS heS
Black's central pawn majority enables him to
approach the endgame with optimism.
18.aS!
Haik takes the opportunity to activate his
rook.
18 ... liJc6
21.i.f6
19.axb6
axb6 20.lLlc4! i.c7
22.0-0-0
White could have considered sacrificing
a pawn with 22.\tId2!? �xc4 23.�xa7 lLJxa7
24.dxc4 dxc4 25 .�b7 when his drawing
chances are quite good, thanks to his two
bishops and Black's doubled c-pawns.
22 ... �d7 23.liJe3 i.eS!
Karpov offers a bishop exchange in order to
get closer to the d3-pawn.
24.heS?!
When playing for a draw, it can be hard
to resist the temptation to exchange pieces.
Nevertheless the text move allows Black's
remaining pieces to become more active.
Better was 24.�g5 �g7 25 .lLJg4 f5 26.lLJf6t
\tIcB 27.�c2 when it is hard for Black to make
progress.
1 986
29
Aldo Haik - Anatoly Karpov
24... tLlxe5 25.�c2 �d6!
Karpov improves his king a bit.
26.£4
26.d4 may have been a better bet: 26 . . . iLl c6
(After 26 . . . cxd4 27.:B:xd4 f6 Black is a fraction
better, but White should be safe.) 27.dxc5t ! ?
White can even think about playing for a
draw by sacrificing a piece for three pawns.
27 . . . bxc5 28.iLlxd5 exd5 29.:B:xd5 t White
should hold this ending.
cj;Je5 36.:B:xb6 iLl e3t 37.cj;Jd3 iLlxg2 38.:B:b5t
Black drops his last pawn so his extra knight
will not be enough to win the game.) 3 1 .J.h3
dxe3 32.:B:xe6t <Jid5 33.c4t J.xc4 34.dxc4t
cj;Jxc4 The position is drawish.
29 ... :B:a1 30J3bS?
Haik wants to be active, but
does not j ustifY it. A better idea
J.a4 t 3 1 . cj;Jd2 :B:a2 32. <Ji c 1 J.b3
still keeps some pressure, but is
from winning the game.
his position
was 30.:B:h l
when Black
a long way
8
8
7
7
6
6
5
5
4
4
3
3
2
1
a
b
c
d
e
f
g
h
26 ... tLlc6 27.:B:hl!
White frees the rook.
27 ...J.b5 2S.:B:h6
Threatening to take on d5. Alternatively
after 28 .:B:h7 :B:a1 29.:B:xf7 J.a4t 30.cj;Jd2 :B:b l
3 1 .:B:f6 :B:xb2t 32. <Jie 1 cj;Je7 33.:B:h6 Black is
better.
2S ... tLle7 29.:B:hS
This is not a bad move in itself, but it signifies
the beginning of a faulty plan.
2
1
a
b
c
d
e
f
g
h
30 ... �c7
Another idea is 30 . . . J.d7!? 3 1 .:B:f8 (if
3 1 .:B:xb6t?? cj;Jc7 traps the rook.) 3 1 . . .J.a4 t
32.cj;Jd2 :B:b l 33.:B:xf7 :B:xb2t 34.<Ji e 1 b5 and
Black is better thanks to his strong b-pawn.
31 .�ffi f5 32.�f7
32.J.f3 defends against the threat of . . . J.a4
and . . . :B:b l , but after 32 . . . :B:gI Black picks up
g3-pawn. But note that 32 . . . J.a4t 33.<Jid2
:B:b l does not work due to 34.J.d l ! when
White saves himself.
If 29.:B:h l d4 Black keeps a slight plus.
32 ... �d6 33.J.f3 J.a4t
29.f5!? Pawn exchanges will bring White
closer to a draw. 29 . . . d4 30.fxe6! fxe6 (30 . . . f6
3 1 .cxd4 cxd4 32.iLlc4t J.xc4 33.dxc4 iLl f5
[33 . . . cj;Jxe6 34.:B:h4!] 34.e7 :B:xe7 3 5 .:B:xf6t
Karpov nicely improves his pieces and
increases his control over the queens ide.
34.�d2
30
The Prime Years
If 39.ie2 Elf2 40.c;t>dl fxg4 4 1 .ixg4 ttJ d5
Black wins.
8
7
39 ... fxg4
6
Karpov wins a pawn, and his newly created
passed pawn is a runner.
5
4
40.ii.e4
3
If 40.Elg7 ttJ f5 ! wins.
2
1
a
b
c
d
e
f
g
h
34 ... b5!
By defending his bishop Karpov makes
. . . Elb 1 into a real threat, as the defence based
on id l will no longer work.
8
7
6
5
4
3
35.g4
2
3 5 .id l ? d4 wins.
1
8
a
7
b
c
d
e
f
g
h
40 ... tiJf5!
Karpov still plays it.
6
5
41 J�b7
If 4 1 .ixf5 Elxf4 42.ixe6 Elxf7 43 .ixf7 g3
Black's pawn promotes.
4
3
2
41 ..J3xf4 42Jhb5
1
This loses quickly, but with two extra pawns
and a powerful passed g-pawn, Black's victory
was already assured.
a
b
c
d
e
f
g
h
35 ... d4!
After a patient positional build-up, Karpov
takes full advantage of the tactical opportunity.
White is unable to keep his position together.
36.cxd4 cxd4 37.tlJc2
White cannot save the b-pawn with 37.ttJ d l
o n account o f 3 7 . . . Elb l 38 .g5 ttJ g6 winning.
37 ... .L:c2 38.�xc2 Elfl 39.ii.a8
42 ... g3 43.Elb8 E:g4
0-1
Not a sparkling win, nevertheless Karpov
played strongly and kept up the pressure, and
when the opponent faltered Karpov punished
him ruthlessly.
1 986
31
Karpov drew his next two games against Gheorghiu and Kiril Georgiev. Then in the penultimate
round he beat Braga of Brazil. Heading into the final round, the USA were leading by half a point
(the competition was decided by game points rather than match points) ahead of the USSR, with
England just half a point further behind in third place.
In the final round Karpov returned to l .e4 and crushed Sznapik in a Dragon, which helped the
USSR to secure a 4-0 victory over Poland. This enabled them to claim the gold medal as the USA
could only draw their final match with Bulgaria, although England also recorded a 4-0 victory so
the final result could hardly have been any closer.
Karpov's final score of 6/9 was good bur not spectacular. One gets the impression that he had
still not fully regained his energy, and indeed he was rested for five of the fourteen games, which
is more than one would normally expect for a player of his immense strength. Kasparov on the
other hand had rested for the full duration between the championship match and the Olympiad,
and this was reflected in his superb score of 8Yz/ l l , which brought him an individual gold medal
on board one.
Although the Olympiad finished on the first of December, Karpov incredibly played in yet
another tournament, the Vienna Open. It is unfortunate that nowadays the old-fashioned Elo
rating system virtually forces elite players to stay away from open tournaments. I would like to
see Anand at the Aeroflot Open and I wonder how Kasparov would have performed at the New
York Open.
Karpov began by outplaying the Austrian 1M Hoelzl from an equal rook ending with the black
pieces. For the rest of the tournament he took on a rather controversial strategy and agreed quick
draws in all four of his remaining black games, each lasting thirteen moves or less. Such a strategy
might work at elite events, bur it is not the way to win an open tournament. Fischer and Kasparov
never did anything like it. He tried to win his white games, and was successful twice, but drew
the other two. Karpov's final score of 6/9 was enough to share second place with Ftacnik, behind
Beliavsky, who finished ahead of Karpov for the second time in 1 986.
In 1 986 Karpov played great chess until the end of the Kasparov match. At times the two rivals
produced some of the most ferocious chess ever seen at this level, and Kasparov had to be at his
best to retain his title.
After the match, Karpov rather recklessly entered too many tournaments and his tiredness caused
his results to dip below their normal level. There is no question that Karpov's playing strength was
lower than that of Kasparov, bur the gap between them was smaller than a comparison of their
respective results towards the end of the year would suggest.
32
1986 Summary
Brussels ( l st place) : 9/ 1 1 (+7 =4 -0)
Bugojno ( l st place) : 8Y2/ 1 4 (+4 =9 - 1 )
World Championship match versus Kasparov, London/Leningrad: Lost 1 1 Y2- 1 2Y2 (+4 = 1 5 -5)
Tilburg (3rd place) : 7Y2/ 1 4 (+2 = 1 1 - 1 )
Dubai Olympiad (Board two) : 6/9 (+4 =4 - 1 )
Vienna (2nd-3rd place) : 6/9 (+3 =6 -0)
Tota1 59.9% (+24 =49 -8)
II Wins • Draws • Losses
1987
Rating 27 1 0 (2 in the world)
Karpov's first event of 1 987 was an important one: he faced his fellow Soviet Andrei Sokolov for
the right to challenge Kasparov again. Karpov had been seeded directly into the candidates final ,
thus ensuring that any new champion would have t o defeat both Karpov and Kasparov. Hitherto
the contestants had met twice before, and surprisingly it was Sokolov who had got the better of
their previous encounters, scoring a win and a draw.
Sokolov performed impressively in 1 986. In his candidates quarter-final match he beat Vaganian
convincingly. In the semi-final against Yusupov, he looked to be heading for defeat but then won
three games in a row to take the match by the score of 7Y2-6Y2.
The Karpov - Sokolov match took place in Linares in February and March, so Karpov was
finally able to take a proper rest after his hectic schedule in 1 986. Sokolov had a narrow opening
repertoire, which must have simplified Karpov's preparation considerably.
Sokolov played a total of thirty seven games against the world champions. He faced Smyslov,
Tal, Spassky, Karpov and Kasparov, but never Anand, Kramnik or Topalov. He scored a total of
three wins, twenty four draws and ten losses against them.
In the first game Karpov unveiled the Caro- Kann, which would in time become one of his most
successful openings. In this game he got a losing position against Sokolov's IQP set-up, but
eventually escaped with a draw. In Game 2, a Queen's Indian, Karpov was pressing and eventually
won after Sokolov erred. The next three games were drawn without major incident.
The sixth game was another Queens Indian. Sokolov showed his tactical awareness with a
spectacular double piece sacrifice, and the chief arbiter Medina even had to step in to calm the
live audience. Karpov kept his cool and transitioned to an endgame with two minor pieces versus
a rook, which he eventually converted. Three more draws ensued, then in Game 1 0 Karpov won
a fine game to take an almost unassailable lead of 3-0 with four games remaining. We will look
at the eleventh and final game of the match.
I Game � I
Andrei Sokolov - Anatoly Karpov
Linares ( I 1) 1987
l.e4
As has been mentioned previously, Sokolov had a very narrow repertoire for a top grandmaster.
In my opinion a player who is looking to challenge for the world title should be able to open the
game with more than just one move.
34
The Prime Years
1 ...c6 2.c4 d5 3.exd5 cxd5 4.cxd5 1lJf6
5.11Jc3
Here is another Karpov game from the end
of the previous decade: 5 .'lMfa4t tZl bd7 6.tZlc3
g6 7.d4 .ig7 8 .Wb3 0-0 9 . .ig5 tZl b6 1 O . .ixf6
.ixf6 I I . tZl f3 .ig7?! (This is unnecessary; it
was more accurate to play 1 1 . . .e6! at once.)
1 2 . .ie2
8
in an effort to avoid Kasparov's fearsome
preparation.
8
7
6
5
7
4
5
2
6
4
3
2
3
' ",��;;;:': �{�
�{�,ip;%jjr:({"" �'
1
a
b
c
d
f
e
g
h
8.h4!?
a
b
c
d
e
f
g
h
1 2 . . . e6! Karpov sacrifices a pawn to obtain
active piece play. 1 3 .dxe6 .ixe6 1 4 .Wb5 a6
1 5 .Wb4 tZl d5 Black's active pieces give him
enough compensation to maintain the balance.
1 6.Wxb7 :1%b8 1 7.Wxa6 :1%xb2 1 8 .tZlxd5
Wxd5 1 9. 0-0 :1%a8 20.Wd3 :1%axa2 Black held
the position fairly easily, Larsen - Karpov,
Montreal 1 979.
5 ... 11Jxd5 6.11Ja llJxc3!?
In the first game Karpov preferred 6 . . . e6 and
later got into trouble. Now he tries a different
approach.
7.bxc3 g6
The more conventional set-up with . . . e6
often leads to positions in which White has
chances to develop a kingside attack, so Karpov
decides to bolster his kingside with a fianchetto.
Hitherto it had only rarely been played, and
Black had scored poorly, but Karpov's home
analysis must have convinced him that it was
playable. Maybe he should have searched for
more of these relatively unknown possibilities
This is a highly ambitious and risky move.
It is still played occasionally, although 8.d4 is
unsurprisingly the main line.
Let me show you the first game in which
7 . . . g6 was played. It is rather amusing (and will
not take long!): 8.Wa4t .id7 9.Wd4 f6 1 O ..ic4
tZl c6 I l .tZl g5 e6 1 2 . .ixe6
� � �.� ��
� i.A.""'�. i
6 " %�1�
,�
"""%w;jj
�� ��1�
�� �r"�i � , �
�
�
wti��
� % '%�r� ��,����''0
���
��;JJlY," ,l!l� ��
� � � ��
8
7
5
4
�y" . .
3
2
a
b
c
d
e
f
g
h
Here the inventor of the line resigned in
Merkulov - Trapezonian, Moscow 1 9 56, but
it turns out that Black could have won! The
simple refutation is 12 . . . .ixe6 1 3 .tZlxe6 (Black
must have overlooked that after 1 3 .Wxd8t he
can play 1 3 . . . tZlxd8! when he keeps an extra
piece.) 1 3 . . . tZlxd4 1 4.tZlxd8 tZl c2t 1 5 .'it> d l
tZl x a l with a decisive material advantage.
1 987
8
7
� • .i.�.�
��
Ifi"� ""J_'i�� 'i
" H '� ��
�" H '� �
!
� ��
��
�
: . �. �. �.
6
35
Andrei Sokolov - Anatoly Karpov
tempo. The position was evaluated as clearly
better for Black by 1M Podgaets, who was a
long time helper of Karpov. This verdict seems
excessive, although there is no question that
the second player has a comfortable game.
��r%�� �� �""
�
� '"' �
••
��'�" "�W'l'l" "�� �
!
/�
a
b
�!
�if��� M
c
d
e
f
g
h
soo.ig?!
Karpov does not panic, but calmly continues
developing, Karpov may have prepared this
move at home, as the position before the
text move had occurred in a previous high
level game, which continued: 8 . . . h6 9.d4
ig7 1 O.ie2 ( l 0 .if4!?) 1 O . . . ltJ c6 1 1 .0-0 0-0
1 2.if4 ig4 1 3 .E1b 1 'lMfd7 1 4 .'lMfd2 <j;Jh7 1 5 .d5
ixf3 16.ixf3 ltJe5 1 7.ie2 E1fc8 1 8 .h5 White
had a small advantage and went on to win,
Beliavsky - Tal, Moscow 1 982.
9.h5 11Jc6!
In view of the situation on the kingside,
there is no point in castling too early.
a
b
c
d
e
f
g
h
12.�b5
The young Russian grandmaster wants to
keep the rook active, but it ends up resembling
the prey rather than the predator. 1 2.E1b2 was
safer, but Sokolov obviously wanted to play as
aggressively as possible in this game.
12oo.a6!
Karpov forces the rook to choose between
the b-file and the fifth rank.
lo.E1bl?!
The rook begins a journey which turns out
to be ill-fated. 1 0.d4 is more natural.
lOoo.'lMfc?
13.�c5
8
By defending the b7-pawn, Karpov prepares
to develop his light-squared bishop.
7
1 1 .ia3?!
5
Developing the bishop to this diagonal
does not fit well with the early advance of the
h-pawn.
4
6
3
2
ll oo.if5
Karpov develops the bishop with gain of
1
a
b
c
d
e
f
g
h
The Prime Years
36
13 ...'?Nd7!
This must have been an unpleasant move
to meet, especially taking into consideration
the match situation, whereby White was
effectively forced to play for a win. The queen
move sets up two important defensive ideas,
namely a potentially awkward check on e6 and
the blocking move . . . .id3.
1 5 .Wd1 Wxa2
Black is also fine after 1 5 . . . Wd7, as it is
worth sacrificing two tempos to stop White
from castling.
1 6.hxg6 hxg6!?
A simpler route to a draw is 16 ... fxg6 1 7.l"i:xc6
Wb I t ( l 7 . . . .id7?? 1 B.Wb4!) 1 B .'it>e2 Wd3t
with a perpetual.
1 7.l"i:xhBt .ixhB 1 B .l"i:xc6 Wb I t 1 9 . .ic 1
14.'?Nb3
8
Sokolov allows a queen exchange. With this
move he admits that his opening strategy has
not been a success, but he keeps his position
playable. What else could he do?
7
6
5
4
3
If 14 . .ie2 Black can obtain a good position
with 1 4 . . ..id3 or 1 4 . . . l"i:dB.
1 4.d4 b6! forces the white rook onto the
awkward c4-square. Alternatively Black can
play 1 4 . . . We6t 1 5 .We2 Wxe2t 16 . .ixe2 l"i:cB
when the queenless position is fine for him.
1 4.Wa4!?
2
a
b
c
d
e
f
g
h
1 9 . . . b 5 ! 20.Wh4 'it>d7!!
Black can play to trap the rook in a most
unusual way.
2 1 .l"i:b6
2 1 .l"i:c5 ? .ic2t is winning for Black.
2 1 . . .'it>c7 22.Wxe7t Wxb6 23 .Wd6t
The game ends with a perpetual check.
8
7
6
a
b
c
d
5
e
f
g
h
This prevents the exchange of the light­
squared bishops thanks to a nice tactical
shot.
1 4 . . . We6t!?
The trap is revealed after 14 ... .id3? 1 5 .h6!
.if6 16 . .ixd3 Wxd3 1 7.l"i:xc6! Wd7 1 B . .ixe7!!
.ixe7 1 9.1LJe5 and White wins.
14 . . . l"i:cB is possible though.
4
3
2
1
a
b
c
d
e
f
g
h
14... 0-O!
White has too many pieces on the queenside,
1 987
37
Andrei Sokolov - Anatoly Karpov
so Black can safely castle. Given the match
situation, many players would have opted for
safety with 1 4 . . ."\We6t 1 5 .i.e2 \Wxb3 1 6.axb3
when the position is about equal. But Karpov
is not desperate to exchange queens, especially
when it means improving his opponent's pawn
structure, and instead he plays the objectively
strongest continuation.
15. hxg6 hxg6
1 6.4:Jg5!? would have been the best attacking
attempt, but Black should be more than okay
here too, as long as he remains calm: 1 6 . . . 1%adS!
( l 6 . . . b6? 1 7.i.c4! e6 I S .1%xf5 is better for
White; 1 6 . . . b5 is possible, and after 1 7.1%d5
\Wc7 it is doubtful that White has anything
better than I S.1%c5 offering to repeat; Black
can also try 1 6 . . . i.f6!? 1 7.1%d5 \WcS I S . 4:J h7
when the position is unclear.) 1 7.i.c 1 ( l 7.d4?
i.xd4! wins.)
8
7
6
5
4
3
2
a
a
b
c
d
e
f
g
h
16.i.c4?!
Sokolov is looking for active play, but the
bishop takes away the escape route from the
rook. White had several other choices, although
none of them were particularly promising.
It is too late aim for a 'natural' set-up with
16.d4, due to 1 6 . . . b6! 1 7.\Wxb6 (or 1 7.1%d5
\We6t I s .'j;l d2 i.f6 1 9.i.c4 \We4 and White is
in trouble) 1 7 . . . \We6t I S .4:Je5 ( l S .i.e2 1%abS!
[18 ... ttJxd4 is cute, but much less effective]
1 9.\Wxc6 1%b l t-+) I S . . . 1%abS 1 9.\Wxc6 1%b l t
20.lt>e2 \Wxa2t 2 1 .Wf3 \Wxa3 and White has a
difficult position because of his shaky king.
White's objectively best continuation was
probably 1 6.i.e2 1%fdS 1 7. 0-0, but this is
inconsistent with his earlier play, and in the
resulting position Black stands better as he will
exert pressure against the hanging pawns.
b
c
d
e
f
g
h
1 7 . . . 4:Je5! (If I 7 . . . e5 I S .\Wc4! the queen transfer
to h4 is dangerous.) I S .\Wb4 4:J d3t 1 9 .i.xd3
\Wxd3 20.\Wh4 1%feS White is struggling, and it
is hard for him to develop his attack.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
16 .if6!
...
With this great move Karpov thwarts White's
attacking chances by taking away the g5- and
h4-squares from the knight and queen.
38
Th e Prime Years
1 6 . . . b5!?
Black can also play against the rook at once.
1 7.�d5 ttJ a5 I B .Wb4 ttJ b7 1 9. ttJ g5 ! ?
I n a n over the board situation, this appears
to lead to a terrifying attack, but objectively
it is a mistake.
Correct is 1 9.�xb7 Wxb7 20.0-0, but White
is only playing to equalize here.
1 9 . . . ttJxc5 20J:1hBt!?
The best chance.
If 20.Wh4 ElfdB 2 1 .�xf7t 'it>fB 22.0-0 ttJ e4!
Black keeps his huge material advantage.
20 . . . �xhB 2 1 .Wh4
In his Chess Informant analysis Podgaets
evaluates the position winning for White,
but it turns out that Black can escape.
23.�xa6 when White is worse but has drawing
chances.
18 ..id5 llJa5
a
b
c
d
e
f
g
h
19.�dl?
This is an unfortunate square for the queen.
a
b
c
d
e
f
g
h
2 1 . . . 'it>g7! 22.�xaB
After 22.Wh7t 'it>f6 23 .Wh4 Wxd5 White
has no strong discovered checks, so he loses.
22 . . . ttJ d3t 23.<;t>n ElxaB 24.f4 e5
And Black wins. Despite the viability of
this line, from a practical perspective Karpov's
move was clearly the best choice.
17.d4 b5!?
Karpov plays to win the present game, rather
than 'only' the match. His plan is to trap the
opponent's adventurous rook.
He could have simplified to an endgame
with 1 7 . . . b6 I B .Wxb6 ElfbB 1 9.Wxc6 Elb l t
20.<;t>e2 Wxc6 2 1 .Elxc6 Elxh l 22.Elxa6 Elxa6
1 9 .Wb4!
Bringing the queen to the fourth rank was
stronger, and in some lines she might even
find her way to h4.
1 9 . . . ttJ b7 20.ttJe5!
White's pieces are oddly arranged in the
centre, but they offer him just enough
dynamic chances to maintain the balance.
20 . . . �xe5?!
I have given this as the main line because it
is the most interesting move to analyse, but
objectively it is risky for Black.
Had this variation occurred at the board,
Karpov would almost certainly have opted
for the following simplifying line: 20 . . . ttJxc5!
2 1 .ttJxd7 ttJ d3t 22.'it>n ttJxb4 23.ttJxf6t
'it>g7 (23 . . . exf6 24.�xaB ElxaB 25 .�xb4 is
also drawish.) 24.�xb4 <;t>xf6 25 .�xaB The
position is absolutely equal.
2 1 .dxe5 'it>g7!
2 1 . . .ttJxc5 ?? 22.Wh4 wins.
2 1 . . .g5 ? takes away the h4-square from the
queen, but White has a great resource.
1 987
a
b
c
d
e
a
f
2 2.'lMfd4!! 4Jxc5 23.l'!hSt! and White wins
the queen.
22.\Wd4 i.e6 23.l'!c6!
After 23 .i.xe6 'lMfxd4 24.cxd4 4Jxc5 2 5 .i.xc5
Black is safe.
23 . . . l'!hS!
Black's best chance to maintain the balance.
23 . . . 'lMfxd5 24.i.xe7! g5 2 5 .'lMfxd5 i.xd5
26.l'!ch6 (If 26.l'!h5 i.xc6 27.i.f6t 'i!?g6
2S .l'!xg5t 'i!?h6 29.g4 White only achieves a
perpetual.) 26 . . . f6 (26 . . . l'!gS ? 27.i.f6t \tJfS
2S .l'!hS wins.)
6
d
e
f
g
h
White's chances are slightly higher, but it
will be hard for him to exploit his bishop pair.
8
7
6
5
4
3
a
19 ... tiJ b7!
5
2
c
1
7
3
b
2
8
4
39
Andrei Sokolov - Anatoly Karpov
vnnn'"h////' ' ' ' ' ",
b
c
c
d
e
g
This forces White to make an unpleasant
choice: he must either lose his 'active' rook, or
exchange his strong bishop, thus handing his
opponent full control over the light squares.
bn"j//nn,,///,///
a
b
d
e
f
g
h
27.f3!! Intending g4 to cage the black king.
(27.i.xfSt ? l'!xfS leaves White facing a
difficult ending) 2 7 . . . l'!gS (27 . . . 4Ja5? loses
to 2S.l'!h7t 'i!?g6 [2S . . . \tJgS 2 9 .exf6 wins]
29.g4 followed by mate in two.) 2 S . exf6t
\tJf7 2 9.l'!h7t 'i!?e6 30.f7 White has decent
winning chances.
24.l'!xhS l'!xhS 2 5 .i.f3 l'!h l t 26.'i!?d2 l'!al
27.l'!xa6 l'!xa2t 2S.'i!?e3 4J dS
20.tiJ e5?!
Objectively this is wrong, although Sokolov
was in a desperate match situation so it is
understandable that he chose to gamble with
the exchange sacrifice.
After 2 0.i.xb7 'lMfxb7 2 1 .'lMfe2 White's position
is no doubt depressing, but he still has chances
to survive. Black has a few tempting options
in 2 1 . . .i.e4 and even 2 1 . . .'i!?g7! ?, intending to
take over the h-file.
40
The Prime Years
20 ...he5 21.dxe5 tLlxc5 22.Y!¥d4
22.ixc5 l'l:fd8 wins.
22 ... tLld3t!
Karpov finds the clinical refutation. After the
imprecise 22 . . . l'l:ad8 ? 23 .Wlh4 ih3 24.l'l:xh3
Wlxh3 25.gxh3 l'l:xd5 26.ixc5 l'l:xc5 27.Wlxe7
White probably holds.
23.�f1
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
23 ... Y!¥a7!
This completely refutes White's attack.
24.he7
24.Wlh4 Wlxf2t leads to a winning ending for
Black.
24 ... Y!¥xd4 25.cxd4 tLl f4
0-1
White is a rook down so he resigned. Karpov's
opening and middlegame play were based on
well-known principles, but he also controlled
the tactics superbly, despite Sokolov's best
efforts to throw the game into turmoil.
The final score of7Y2-3Y2 was a great triumph
for Karpov, and it brought the previously high
flying Sokolov crashing back down to earth.
I know Andrei from j unior tournaments, and
he was the most confident player I ever met.
Maybe there were other factors, but it seems
to me that the outcome of this match broke
his confidence and he did not maintain his
world class status for much longer. Karpov's
dominant victory must also have come as a
bold message to Kasparov.
***
Not long after the Sokolov match, Karpov
participated in a top level tournament in
Brussels. He started with two hard fought
draws against Larsen and Tal, then defeated
Meulders in a nice endgame. In rounds 4 and
5 Karpov pressed for roughly sixty moves, but
had to settle for draws against Van der Wiel and
Winants. It looked like his game against Nigel
Short would end in a draw as well, but after
a period of accurate defence, the Englishman
blundered badly and lost the game in one
move. In the next two rounds Karpov drew
with Torre and Timman, then benefitted from
a huge slice of good fortune after Korchnoi
touched the wrong piece and left a knight en
prise in what should have been a dead drawn
position.
In the penultimate round Karpov and
Kasparov faced each other for the hundredth
time. Playing with the white pieces, it looked
like Karpov would obtain the kind of slight
endgame advantage that suited him so well,
but Kasparov reacted with a dynamic pawn
sacrifice in order to install a rook on the second
rank. The position should have been equal, but
Karpov misplayed it and in the end he had to
fight hard to draw. In the final round he had
a quick draw with Lj ubojevic, who had one of
the tournaments of his life, sharing first prize
with Kasparov on 8Y2/ 1 1 .
Karpov finished in third place with 7/ 1 1 .
It was a decent result, but not a spectacular
one, especially considering that a couple of his
wins could easily have ended in draws had his
opponents not blundered. My impression is
1 987
41
Anatoly Karpov - Viktor Korchnoi
that h e had played i n too many events i n the
previous six months. He prepared his openings
mainly for his matches against Kasparov and
Sokolov, which left him with insufficient time
to prepare for other top grandmasters.
Karpov's next tournament was the Euwe
memorial in Amsterdam, a four-player, double­
round-robin event. His opponent in the first
round was Korchnoi. Since their encounter in
London 1 984 (Game 71 of the first volume) ,
Karpov had defeated him twice (including the
recent touch-move accident) and they had
drawn three times.
5 ... cxb4 6.g3 b6
After this game Korchnoi switched to 6 . . . 0-0
and achieved decent results.
7.i.g2 .ib7 8.0-0 0-0 9.VNb3
Karpov provokes the following pawn move.
Later the queen might want to go to d3 , once
the reply . . . �e4 has been prevented.
9 ...a5
I Game 6 I
Anatoly Karpov - Viktor Korchnoi
Amsterdam 1987
l.d4 tLlf6 2.c4 e6 3 . .!iJf'3 �b4t 4.�dl
This was Karpov's usual choice, although he
sometimes played 4.ltJ bd2 as well. He scored
well with both of them.
4 ... c5!?
According to the database this move was first
played by P. Wolf (not to be confused with the
grandmaster Patrick Wolff) against Martz at
the 1 963 Illinois Open. However, it was the
Latvian genius of creative opening play, Alvis
Vitolinsh, who first started to play it regularly.
Remarkably, over the course of his long
career Karpov only ever faced this variation in
the present game. It probably did not come as
a surprise to him, as a year earlier Korchnoi has
outplayed Kasparov from the black side of it,
only to squander his advantage in time trouble
and concede a draw.
5.�xb4
Kasparov did not take on b4, but played
5.g3 .
a
b
c
d
e
f
g
h
10.a3!
Karpov wastes no time in undermining
Black's queenside and drawing his knight to a
passive position.
10 ... .!iJa6
Taking on a3 would give White easy play on
the b-file.
l 1 .tLlbdl d6 12J'Udl VNe7 13.tLlel!
Karpov anticipates that Korchnoi will want
to play . . . eS some point, and sees that with
no bishops on the board he will have better
chances to utilize the dS- and e4-squares.
13 ...hg2 14. .!iJxg2 �fd8?!
Korchnoi probably underestimated Karpov's
plan. The main line is 1 4 . . . Wfb7, a flexible
move which enables Black to liberate himself
with . . . bS in some positions. I S .ltJ e3 E1fc8 (The
42
The Prime Years
immediate 1 5 . . . b5?! is premature due to 1 6.a4!
bxc4 1 7.Wlxc4 CiJ c7 1 8 J=l:dc 1 CiJ cd5 1 9.CiJxd5
exd5 2o.Wld3 with some advantage for White,
Nikolic - Renet, Debrecen 1 992.) 1 6. CiJ c2
bxa3 1 7.bxa3 d5 1 8.!'lab l !'lab8 1 9 . CiJ e3 White
kept some pressure on the queenside but Black
remained solid enough, Zueger - Gostisa,
Graz 1 993.
tempo, Karpov has renewed the threat of CiJ c2.
By centralizing his queen, he also anticipates
Black's potential activity in that area of the
board.
1 6. CiJ c2?! would have been premature due to
1 6 . . . a4 1 7.Wlf3 (or 1 7.Wld3 b3 1 8 . CiJ e3 Wld7
1 9.!'lac 1 CiJ c7 and Black is doing fine) 1 7 . . . b3
1 8 . CiJ e3 b5 when Black has a pleasant game.
Interestingly Black's a-pawn is less vulnerable
on a4 than on a5, as White's knights are
currently a long way from being able to attack
the former.
8
7
�
��if�.m
��� "'' ' Y-� %%�3r��
��,
��,If.''',,%
�
'
� ��l!��
�%
.
tIIf.�%��
/"""
� �
�
a
b
c
d
e
f
g
h
15.CiJe3!
Karpov not only centralizes his knight, but
also creates the subtle positional threat of CiJ c2,
which would force Black to capture on a3 and
open the b-file.
4
lii
O lii Iii.
�
� '.V�Iii�
" '.
%
2 " ' ' ' �r%
3
�
& ,�
iA% •
/.:�
6 . %'.
""%
%
5
�
��:'
'%'''''%m'-:;;
�/�"''''
8 fmf!}
f�
,�ff!}
""%.�;
.
""%=""
a
b
,
c
d
e
f
g
h
16 ... e5?
1 5 ...Wle8
Korchnoi gets ready for CiJ c2, which can now
be met by . . . a4.
1 5 . . . d5 is possible, although in this case Black
is agreeing to take on a somewhat worse
position on account of the doubled b-pawns.
1 6.cxd5 CiJxd5 ( 1 6 . . . exd5 1 7. CiJ f3) 1 7. CiJ dc4
Wlg5 1 8 .CiJxd5 Wlxd5 1 9.!'lac 1 !'lac8 In this
position White keeps some advantage with
either 20.a4, or 20.e4!? Wlb7 (20 . . . Wlxe4?
2 1 . CiJ d6!) 2 1 .Wle3 .
16.Wld3!
Another fine multipurpose move. Since
the move . . . a4 no longer comes with gain of
Presumably Korchnoi was concerned about
the impending CiJ c2, and decided to take action
in the centre. However, it soon transpires that
the cure is worse than the disease, as Black's
pawn structure is weakened considerably,
along with some key central squares. Instead
Black should have played some kind of waiting
move.
1 6 . . . !'lac8 ? allows 1 7.c5! when Black is in
trouble.
1 6 . . . !'la7 is mentioned by Karpov and Zaitsev,
who evaluate the position as slightly better for
White without giving any further analysis.
White should carry out his plan with 1 7. CiJ c2
43
Anatoly Karpov - Viktor Korchnoi
1 987
bxa3 1 8 .bxa3 followed by attacking on the
b-file.
Karpov also mention 1 6 .. .'IWc6, giving it the
same evaluation, although it seems to me
that this is probably Black's best try. A likely
continuation is 1 7. ttJ c2 bxa3 1 8 .bxa3 Ei:ac8
1 9 .ttJe3 ttJb8 20.Ei:ab l ttJ bd7 when Black's
position with the weak b6-pawn is worse, but
not unbearable.
8
7
1 8.'i!Yxe4 exd4?!
Korchnoi simplifies, perhaps hoping to steer
the game towards a rook ending in which
drawing chances may occur. But he never gets
close to achieving it, and the backward d-pawn
becomes a serious weakness. Black would have
had a hard time after other moves as well.
1 8 . . . Ei:ac8 1 9.Ei:d2
This simple move works well.
A good alternative is 1 9.\Wb7!? exd4 20J'hd4
ttJ c5 2 1 .\Wf3 ttJ b3 22.Ei:e4 \Wd7 23.Ei:d l when
Black is in trouble.
1 9 . . . exd4 20.\Wxd4 ttJ c5
8
6
7
5
6
4
5
3
4
3
2
a
b
c
d
e
f
g
h
17.ttJe4!
With this move Karpov opens the d-file
and exchanges Black's better knight, thus
facilitating the occupation of the d5-square.
The disappearance of a pair of knights will
also accentuate the misplacement of Black's
remaining knight on a6.
Korchnoi played the opening and early
middlegame well below his normal level, as
at move 1 7 he is already close to being lost.
When one starts playing a new opening system
it gives motivation and some inspiration, but it
takes time to develop the deep understanding
which is necessary to succeed at a high level.
Credit to Karpov though - he prepared well
and did everything correctly.
17 ... ttJxe4
Black has to take the knight, otherwise he loses
a pawn: 1 7 . . . \We7 1 8 .ttJxf6t \Wxf6 1 9 .ttJd5
2
Y_="'_",,,,,/W,,'_,,
a
b
c
d
e
f
g
h
2 1 .ttJ f5 !
From here the knight attacks three key
squares.
2 1 . . .\Wf8 22.\We3!
22.\Wh4 Ei:e8! (but not 22 . . . ttJ b3 ? 23.Ei:d5!
ttJ xa l ? 24. ttJ e7t with a cute mating attack)
23.Ei:ad l bxa3 24.ttJxd6 a2 is not fully
convincing.
22 . . . Ei:e8
White was threatening a fork on e7.
23 .\wf3
Black will suffer due to his isolated pawn.
The best chance may have been:
1 8 . . . \We6!? 1 9 . ttJ d5
1 9 .\Wb7?! is less promising: 1 9 . . . exd4
20.Ei:xd4 ttJ c5 2 1 .\Wg2 (2 1 .\Wf3? does not
work, as 2 1 . . .ttJ b3 22.Ei:e4 ttJ d2! wins for
44
The Prime Years
Black.) 2 1 . . . tLl b3 22.E!:e4 Wlf6 23.E!:b l White
is a bit tied up.
1 9 .Wlxd4! ? E!:ab8 20.tLlf5 Wle5 2 1 .tLlxd6 Wlxe2
22.E!:e l Wlf3 23 .E!:e7 tLl c5 24.E!:ae 1 Black faces
serious problems here as well.
19 .. Jhe8 2oJ'hd4 �ad8
8
7
6
a
b
c
d
5
e
f
g
4
h
1 9 . . . E!:ab8!
Karpov and Zaitsev only mention the line
1 9 . . . exd4 20.Wlxd4 E!:ab8 2 1 .axb4 tLlxb4
22.tLlxb4 axb4 23.e3, after which they opine
that White is already winning.
20.axb4 tLlxb4 2 1 .tLlxb4 axb4 22.dxe5 dxe5
23 .E!:d5
Black's position is passive and unpleasant,
but he still has chances to survive.
8
3
2
1
a
b
c
d
e
f
g
h
2 1 .�adl
Karpov's style in endgames involves
improving his pieces first, followed by his king.
He follows the same formula in this game as
well.
2 1 ...�e6 22.<j;>f1 b3
7
Korchnoi hopes that by fixing the b2-pawn
he can get play against it, but his pieces are
too far away from that pawn. His situation
would have been gloomy after other moves as
well.
6
5
4
3
22 . . . g6 23.tLld5 bxa3 24.bxa3 E!:b8 25 .E!:b l
White wins a pawn.
2
1
a
b
c
d
e
f
g
h
19.�xe8t!
Karpov correctly judges that he can bring
down his opponent in the endgame, as Black
has too many weak pawns.
Nevertheless his chances would also have
remained excellent with queens on the board:
22 . . . bxa3 23.bxa3 Black would like to eliminate
all the queens ide pawns as he would have some
chances to draw with three versus four on the
kingside, but he cannot even accomplish this,
for instance: 23 . . . tLl c7 24.E!:b l E!:b8 2 5 . tLl f5
tLl e8 26.E!:d3 E!:e5 27.tLlxd6 tLl xd6 28.E!:xd6
White is winning.
23.ttJ d5
1 987
45
Anatoly Karpov - Viktor Korchnoi
This wins a pawn without allowing any
compensation.
A desperate exchange sacrifice. 30 ... ttJ xb2
was no good due to 3 1 .c5 when the knight is
trapped.
23 ... ttJc5
There is no point in trying to hang onto the
pawn with 23 . . J''\ b 8, as 24. ttJ c3 1"1d8 2 5 . ttJ a4
wins.
3 1 .�xd6 �xd6 32.�xd6 tLlxb2
Korchnoi gets the b-pawn in the end, but
he has sacrificed too much material in the
process.
24.ttJxb6 h6
33.�b6 tLlxc4 34.�b8t c.f,lg7
8
7
6
5
4
3
a
b
c
d
e
f
g
2
h
1
a
25.0!
b
c
d
e
f
Karpov takes away the e4-square to stop any
potential counterplay. His main idea is to win
the d6-pawn, but it will not run away.
35.a4!
25 ... g5 26 ..!::U d2!
35 ... c.f,lf6 36.e4 c.f,le5 37.c.f,le2
1-0
Karpov combines attack and defence, using
the rook to guard both the b2- and e2-pawns,
while also menacing d6. The idea is not hard to
find, but it is instructive nonetheless.
26 ... c.f,l g7
Bringing the king to the centre would not
have helped: 26 . . . �f8 27.ttJd5 1"1g6 28.ttJe3
!"lf6 29.1"1d5 Black's position falls apart.
27.ttJd5 tLla4 28.tLlc7 1"1f6 29.tLlb5
Karpov wins a second pawn, and with it the
game.
29 ... c.f,lfS 30.tLlxd6 �Uxd6!?
g
h
Karpov separates Black's queens ide pawns,
resulting in a trivial win.
The b-pawn will disappear, along with all
Black's hopes. Karpov completely outplayed
Korchnoi in the opening, and converted his
advantage ruthlessly. It is hard to say for sure,
but this game may be an example of how the
Kasparov matches stimulated Karpov to raise
the level of his opening preparation.
In the next round Karpov drew with
Timman, then defeated Van der Wiel to cap
an excellent first half of the tournament. In the
next three rounds he slowed down and drew
all three games. He was pressing against both
Korchnoi and Timman, then after a sharp
opening against Van der Wiel he settled for
46
The Prime Years
an equal endgame. This slower second half
enabled Timman to draw level and share first
place with 4/6.
to obtain an excellent posltion against Tal
although I later lost my way and was punished
by the maestro.
Karpov's next tournament was in Bilbao, where
world class grandmasters mixed with locals
and two top female players. In the first round
Karpov had the white pieces against Izeta
Txabarri. The endgame was close to drawn,
but Karpov kept pressing and eventually the
local player succumbed. He followed this with
a solid draw against Andersson, then faced
Zsuzsa Polgar, the eldest of the famous sisters
and a future women's world champion.
It is not easy to determine precisely how
many full length games Polgar played against
the world champions, as there are numerous
rapid, blind, and other exhibition games in
the mix. As far as I can see, she has played a
total of twenty two 'normal' games against the
world champions, from Smyslov all the way
to Topalov, with the exceptions of Fischer,
Kasparov and Kramnik. Against these giants
she scored one win, eleven draws and ten
losses. She lost her only subsequent encounter
with Karpov.
1 ...e5 V!ljc3 tLl c6 3.g3 g6 4..ig2 .ig7 5.e3
I Game 1 I
Anatoly Karpov - uuzsa Polgar
Bilbao 1987
l .c4
With his fourth match against Kasparov
approaching, Karpov does not want to reveal
too much information in his main openings.
He may also have been influenced by the fact
that Zsuzsa often played a sharp version of the
Queen's Gambit Accepted. Around that time,
your author worked with her and together we
developed some new ideas in this opening.
With one of them she beat Van der Sterren
and drew against Timman, and I managed
It looks like Karpov wanted to avoid
any opening lines involving an early
confrontation.
5 ... d6 6J�b l
In closed positions, gaining space with pawns
often constitutes an effective strategy.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
6 ... tLl ge7
In such positions the inclusion of the moves
6 . . . a5 and 7.a3 brings certain advantages for
both sides. Black will exchange a pair of pawns
on the queens ide, meaning one less target for
White, and he will temporarily obtain the a-file
for his rook. On the other hand, since White
naturally has the initiative on the queens ide,
the open file is likely to fall into his hands
during the middlegame. Overall the choice is
a matter of taste.
7.b4 a6 S.d3 .ie6 9.tLld5
Elite chess players cannot expect to succeed
at the highest level if they conduct the opening
in this manner in every game with the white
pieces, but every now and again it can be done.
With his last move Karpov prevents . . . d5.
1 987
47
Anatoly Karpov - Zsuzsa Polgar
9 ... 0-0
The position is a Closed Sicilian with reversed
colours. In that line White sometimes pushes
the h-pawn without castling, but it would be
rather risky to try it a tempo down.
10.lLle2 �d7 I l .lLlec3 lLld8 12.a4 lLlc8
Polgar carries out a well-known plan,
preparing . . . c6 to oust the strong knight,
without allowing an exchange on e7.
8
7
6
Black improves the knight and prepares . . . f5 .
Exchanging the bishop with 1 5 . . . ii.h3 was a
reasonable alternative. Later Black can develop
her knight to e6, where it stands well.
16.�c2 f5 17.lLld2 gS?!
Zsuzsa hopes to build an attack, but Karpov's
queenside play turns out to be more effective.
1 7 . . . d5 gains space but creates a hole on c5.
1 8 .cxd5 cxd5 1 9. ttJ b3! b6 20.a5 e4 2 1 .axb6
ttJxb6 22.ttJ c5 We7 Now after 23 . ttJ e2 or
23 .ii.b2 White remains somewhat better, as
Black's position is a bit loose.
My preference is 1 7 . . . a5 ! to disrupt White's
queenside play: 1 8 .b5 ttJ e7 1 9.ii.a3 ( 1 9.bxc6
bxc6 20.ttJe2 �ab8) 1 9 . . .f4 (Black can also
consider both 1 9 . . . g5 ! ? and 1 9 . . . c5 ! ? 20.liJd5
Wd8 .) 20.ttJde4 ttJ f5 Black's position is fully
playable.
5
4
3
2
a
b
c
d
e
f
g
h
8
13.lLle4
7
Of course Karpov was not entertaining any
thoughts of winning the black queen with a
fork on f6. Instead he vacates the c3-square
for the other knight, and meanwhile creates
the positional threat of lLl g5 which would
eliminate an important bishop.
6
13 ...£6
Zsuzsa decides to keep the bishop. The pawn
move appears ugly, but it will advance to f5
after suitable preparation.
It was also possible to 'sacrifice' the bishop with
13 . . . f5!? in order to obtain more fluent piece
play: 1 4.lLlg5 c6 1 5 .ttJ c3 ii.f6 ( 1 5 . . . ttJ e7!?)
1 6.liJxe6 ttJxe6 Black's pieces stand well, and it
is not easy for White to open the position for
his bishops.
14.0-0 c6 Is.lLldc3 ltJf7
5
4
3
2
1
a
b
c
d
e
f
g
h
18.aS!
Karpov does not panic in the face of the
impending pawn storm. White has time to
gain space and fix the weakness on b6.
1 8 ... hS
Continuing the pawn storm in a different way
with 1 8 . . . g4 1 9. ttJ a4 h5 20.f4 h4 transposes to
the following note. Instead 20 . . . gxf3?! 2 1 .ttJxf3
48
The Prime Years
is weaker as Black is slightly vulnerable on the
kingside.
Karpov embarks on a deep and powerful
plan, which will be revealed over the next few
moves.
19.f4 tlJe?
Pushing the kingside pawns first with
1 9 . . . g4! ? looks preferable, for instance 20.tlJa4
h4 2 1 ..ib2 h3 22 . .ih l Wc7. In the game
Black's pawns do not get this far. Black is still a
bit worse, but compared with the game White
has to pay more attention to the safety of his
king. And in the long term, the advanced
pawns could easily make a difference in some
endgames.
20.tlJa4! �ae8
The main alternative was 20 . . . E!ad8 2 1 .fxeS
(Also after 2 1 .fxgS tlJxgS 22. tLl b6 White is a
bit better.) 2 1 . . .tLlxeS 22.d4 tLlg4 23.tLlf3 .ih6
24.dS cxdS (24 . . . .if7 2S .h3 tLl f6 26.tLl d4)
2 S . tLl d4 .if7 26.tLl b6 We8 27.cxdS and White
keeps some advantage, as Black's pieces are not
well coordinated.
21.tlJb6!
It takes some time to appreciate the strength
of the knight on this outpost. It may not
attack anything (apart from a one move threat
to the queen, which is easily avoided) but it
takes away a number of squares from the black
pieces.
I could not help noticing that Gratvol, Karpov's
junior trainer, had a strong knight on b6, in
a game that we covered briefly on page 1 9
of the first volume. It is probably a complete
coincidence, but still it is possible that they
considered this type of idea when studying
together. Such ideas can remain in a player's
mind for a long time; for instance, Leko had
positions against world champions Kramnik
and Khalifman, which I taught him when he
was j ust over ten years old.
21. We? 22.tlJb3!?
••
A decent alternative was 22.fxgS tLlxgS 23 . .ib2
tLl c8 24. tLl a4 tLl e7 2S .E!be l and White is a bit
better.
22 tlJg6
•.•
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
23.�dl!?
Karpov continues with his plan, but it
turns out that there was an even stronger
continuation available.
23.fxgS! tLlxgS
If 23 . . . h4 24.gxh4 tLlxh4 2S .Wf2 tLlxg2
26.Wxg2 Black has problems, as White will
push his h-pawn shortly.
24.Wd l !
Black's kingside pieces turn out to be
surprisingly vulnerable.
24 . . . h4 2S .WhS! h3 26 . .ih l We7!
This clever defence keeps Black in the game,
but she is not out of the woods yet.
27.cS !
27.Wxg6 .if7 28.WxfS .ie6 29 .Wg6 :1'i:xfI t
30. mxfI .if7 is a repetition.
27.bS!? axbS 28.cxbS e4 29.tLld4 White's
chances are somewhat higher in this
complicated position.
1 9B7
27 .. .!:\dB
27 . . . d5? loses to 2B.Wfxg6, as the queen now
has the d6-square available.
a
b
c
d
e
49
Anatoly Karpov - Zsuzsa Polgar
f
g
White will neutralize Black's play on the e-file,
then look to exploit her weaknesses.
h
28.b5!!
This beautiful breakthrough decides the
game.
28 .�a3 can be met by 2B . . . d5!.
28 .. .f4 29.i.a3! Wff6 30.bxc6 f3 3 1 .l:%f2
White is winning.
23 g4 24.d4!
..•
Karpov executes his plan to gain space in the
centre.
24... exf4
Polgar opens the e-file, hoping to get some
activity there. Black cannot keep the position
closed, as 24 . . . e4 is well met by 25.d5! cxd5
26.cxd5 �d7 and now after 27.�b2 or 27.tLJ d4
White will dominate on the c-file, while Black's
pieces are achieving nothing on the kingside.
a
b
c
d
e
f
g
h
26.c:xd5 i.xd5
26 . . . cxd5?! 27.tLJ c5 is hopeless for Black.
27.tLJxd5 c:xd5 28.tlJc5!?
Karpov decides to postpone taking the d5pawn. In doing so he controls the e2-square
for longer, thus reducing the chance of a black
rook occupying the second rank.
Nevertheless 2B .i.xd5 was perfectly valid, and
after 2B . . . l:%d8 29.�e6 l:%feB 30.l:%e l Black has
no compensation.
28 .\Wc6 29 .ib2
•.
•
Karpov defends the d4-pawn securely.
25.exf4 d5?
Zsuzsa tries to fight for space in the centre,
and is even ready to sacrifice a pawn for it.
However, the lesser evil would have been to
avoid any further weakening of that area.
The best chance looks to be: 25 . . . h4 26.d5
(There is also 26.i.b2 d5 27. tLJ c5 h3 2B.l:%e l
ic8 29.�f1 when White's knights are superior
to Black's.) 26 . . . cxd5 (26 . . . i.d7 27.Wfd3 is
strong) 27.�xd5 �xd5 2B .Wfxd5 l:%e2 29 .�d2
a
b
c
d
e
f
g
h
50
The Prime Years
29 ... h4?
Zsuzsa probably saw no good ideas, and
just played this move in the hope that an
opportunity would arise on the kingside.
Karpov had finely judged that Black had no
time to improve her knight:
29 . . . lL'l d6 30.'lWb3
White can even go for a reversal of the
material balance with 30.:1'k l ! ? lL'l b 5 3 1 .'lWb3
lL'l e7 32.E!:fe l lL'lxd4 33 .iixd4 iixd4t 34.\t>h l
with tremendous compensation for the
pawn.
30 . . . lL'l c4
If 30 . . . lL'l e4?! 3 1 .E!:bc 1 or 30 . . . lL'l e7?! 3 1 .E!:fe l
cj;Jh7 32.E!:e6 Black i s hopelessly tied up.
3 1 .'lWxc4
There is also 3 1 .E!:fe 1 when Black is unable
to keep her position together.
3 1 . . . dxc4 32.iixc6 bxc6 33.E!:bc 1
White wins a pawn, and should have little
trouble in the endgame.
However, Black could have offered sterner
resistance with:
29 . . . E!:e3!
30.E!:e 1
After 30.'lWa4 'lWxa4 3 1 .lL'lxa4 h4 Black
continues to resist.
30 . . . E!:xe 1 t
Black should exchange in order to draw the
white queen away from the b3-square.
If 30 . . . E!:fe8? 3 1 .E!:xe3 E!:xe3 32. \t>f2 E!:e8
33 .'lWb3 E!:d8 34.E!:e l White is winning.
3 1 .'lWxe l E!:e8 32.'lWd l h4! 33 .'lWb3 lL'l e7
34.E!:c 1
Or 34.E!:e l h3 3 5 .E!:e6 lL'l d6 and Black is still
alive.
34 . . . h3 3 5 .iifl lL'l d6 36.lL'lxa6 iixd4t 37.\t>h l
lL'l c4 38.iixd4 bxa6
White is still clearly better, but the game is
not over yet.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
30.V:Vb3
The time has come for White to eliminate
the weak d5-pawn.
30 .. J�e2
a
b
c
d
e
f
g
h
Preventing the white queen from coming to
b3. They say it is easy to miss a threat along a
rank, but maybe it is true regarding defence
as well. Kasparov also missed a strong
defensive move on the third rank in Game
22 of his second match with Karpov.
A desperate attempt. I cannot be certain, but
I suspect that Zsuzsa, with her fine calculating
ability, had already seen the way she would
lose, but hoped that her opponent might have
missed something.
It is hard to suggest anything else, for instance
if 30 . . . lL'l e7 3 1 .E!:fe l h3 32.iih l \t>h8 33.E!:bc 1
Black is in a hopeless situation.
1 987
Anatoly Karpov - Zsuzsa Polgar
31 .ixd5 Wid6 32JUei
32.1Mrc4 was winning as well.
32 .. Jhb2 33Jhb2 i.xd4t 34.'it>f1 ixb2
If 34 . . . ttJxf4 3 5 .gxf4 1Mrxf4t 36.'it>g2 Black
soon runs out of checks.
8
7
6
4
3
2
a
b
c
d
e
f
g
Illescas and Ochoa de Echaguen. Interestingly,
he outplayed the last of the three on the
black side of a Berlin Defence. Considering
Kramnik's success with this opening against
Kasparov in 2000, we can only imagine what
might have happened had Karpov decided
to prepare it for use at world championship
level. He might even have forced Kasparov to
abandon 1 .e4 altogether, returning the 'favour'
given by Kasparov's Scheveningen.
In the last rwo rounds Karpov drew quickly, to
finish with 7/9, half a point clear of Andersson
and a point ahead of Chiburdanidze and
Lj ubojevic. It was an impressive performance,
especially considering that he was holding
back his main openings in order not to reveal
too much to Kasparov.
5
1
51
h
35J�e6!
This intermediate move pours water on the
flames and extinguishes Black's last hopes.
35 1Mrc7 36J�xg6t 'it>h7 37.�b6
•••
Karpov can choose to win in any way he
pleases. Another nice line was 37.1Mrxb2!?
'it>xg6 38.ttJe6 1Mrd6 39 .Wfg7t Wh5 40 .Wfh7t
tiJ h6 4 1 .ttJg7 mate!
37 i.g7 38.�xb7
.•.
Maybe Karpov was short of time, otherwise it
is hard to explain why Zsuzsa did not resign.
38 ...Wid6 39.lLle6!
Karpov could also take the free piece, but his
chosen move is even more accurate.
39 ... �c8 40.�xf7 'it>h8 41 .�xf5
1-0
In the fourth round Karpov made a quick draw
with the black pieces against Ljubojevic, then
he defeated the local players Fernandez Garcia,
Fourth World Championship match
versus Kasparov
In 1 987 the Spanish town of Seville won the
right to host the latest clash berween the rwo
titans. The match started in October and
finished in December. Kasparov had achieved
some sparkling results since their last match
and widened the rating gap to forry points.
However, the closeness of the last match
indicated that Karpov still stood a chance.
Kasparov's delegation consisted of Nikitin,
Dorfman, Azmaiparashvili and Dolmatov,
while Karpov's included Zaitsev, Ubilava,
Lerner and Podgaets. It is hard to spot when
the weakening Soviet state stopped paying
for the trainers of their chess superstars, but I
would estimate that from around the time of
the present match players probably had to dip
into their own pockets.
The first game saw both sides playing for
safery in a g3 Griinfeld, and a draw was agreed
in thirry moves. The next game was more
combative, and resulted in the challenger
drawing first blood.
52
The Prime Years
I Game 8 I
Garry Kasparov - Anatoly Karpov
Seville (2) 1987
l .c4
Kasparov unveils a surprise on the very first
move. Before this game he had only used the
English Opening sporadically. It looks like
he wanted to avoid Karpov's reliable Nimzo/
Queen's Indian repertoire.
1...e5
Maybe Karpov wanted to surprise Kasparov.
He abandoned 1 .e4 and thus the Scheveningen,
but is ready to play the position a tempo down.
During an interview with Kouatly, Karpov
once said he believed he could always make a
draw with the white pieces if he wanted to, but
it was not the case with Black.
Karpov chooses the same variation he used to
draw with Korchnoi in their 1 974 title match.
Interestingly, in the last game of the match
in which Karpov employed 1 . . .eS, he deviated
with 6 .. .l''1: e S and scored another valuable win:
7.d3 ixc3 S.bxc3 e4 9.tLJd4 h6 1 O.dxe4 tLJxe4
1 1 .Wc2 dS 1 2.cxdS WxdS 1 3 .e3 tLJaS 1 4.f3
tLJ d6 1 S .e4 WcS 1 6.ie3 tLJ dc4 1 7.if2 We7
1 S .Eiad 1 id7 1 9.f4 EiadS 20.eS ig4 2 1 .tLJ fs
We6 22.EixdS EixdS 23.tLJd4 WcS 24.fS ?! cS
2S .We4 cxd4 26.Wxg4 tLJxeS Kasparov was not
able to generate enough play for the sacrificed
pawn and went on to lose, Kasparov - Karpov,
Seville ( 1 6) 1 9S7. Amazingly, despite achieving
a plus score with the . . . tLJ f6/ . . . tLJ c6/ . . . ib4
system against the strongest player in the world,
Karpov never played it again after this match.
7.tlJg5
In the eighth game Karpov opted for
another set-up with 2 . . . d6 followed by . . . cs.
Kasparov showed his class and won a strategic
masterpiece, which is given in the notes to
Game 1 9 in the first volume.
Korchnoi opted for the other main line:
7.tLJ e l ixc3 S.dxc3 h6 9 . tLJ c2 b6 1 O. tLJ e3
ib7 1 1 .tLJ ds tLJeS 1 2.b3 EieS 1 3 .a4 d6 1 4.Eia2
tLJ ed7 1 S .h3 as 1 6.ie3 tLJxdS 1 7.cxdS Wf6
1 S .c4 Wg6 1 9.Wb 1 ics 2o.id4 tLJ cs 2 1 .Wh2
id7 22.Eig1 hS 23.Eib2 Eie7 White was unable
to make anything of his bishop pair and the
game was eventually drawn, Korchnoi Karpov, Moscow (9) 1 974.
3.tlJf3 tlJc6 4.g3 �b4 5.�g2 0-0 6.0-0
7 ... ,bc3 8.bxc3 ge8 9.f3
2.tlJc3 tlJf6
8
�
7
6
5
4
3
2
��.t)�
�:I.�
''',,%�
4i�
�,�
,, �".%�
,
�•.
� �
�
%
iil iiIiiI iii. iii
Wl% iiIiiI
" 'iii
iill!'! '
."�
iiIiiI
� m �lZJ�[J
.�,,3�.�..�
8��'
��
8��
��iL�
%=
..�%�""':.%���
....r�%•....
, ..
.
�
V
�
�
�
iii:.
. . . . .z . '
�
a
6... e4
/;
,
b
c
d
e
f
�
%
g
%
h
8
7
6
5
4
3
2
1
a
9 ... e3!?
b
c
d
e
f
g
h
1 987
Karpov surprises his opponent with a pawn
sacrifice to create an unusual pawn formation
in White's camp. In his next black game Karpov
chose not to repeat the idea and deviated with
9 ... exf3, but failed to equalize and eventually
lost. The move was a novelty at the time,
although Smyslov had utilized the same idea
in a similar position:
a
b
c
d
e
f
g
b
c
d
e
f
g
Although Black's idea was unsuccessful in
this game, it is possible that Zaitsev, one of
the most creative opening players of that era,
would have seen the game and conceived the
idea of using the . . . e3 advance in a different
scenario.
h
1 0 . . . e3 1 1 .d3 �b7 1 2.'lWe 1 tLl bd7 1 3.g4 h6
14.h4 tLlfS 1 5 .'lWg3 tLl g6 1 6. tLl h3 tLl h7 1 7.h5
lLlh4 l S .�h 1 f5 1 9.�b2 'lWf6 20.f4 �xh 1
2 1 .g5 hxg5 22.fxg5 'lWe5 23 .'lWxh4 �c6 24.l'l:f4
g6 25.hxg6 tLl fS 26.'lWh6 'lWg7 27.l'l:xf5 tLlxg6
28.l'l:afl l'l:fS 29.l'l:f6 'lWxh6 30. gxh 6 White has
won a pawn, and soon converts his advantage
using a beautiful mating net: 30 . . . l'l:xf6 3 1 .l'l:xf6
Wh7 32.�c l l'l:gS 33.tLlg5t �xh6 34.�xe3
Wh5 3 5 .l'l:f7 l'l:eS 36.l'l:h7t �g4 37.�f2 tLl e7
38.lLle6 tLl f5 39.tLld4 tLlxd4 40.cxd4 l'l:cS 4 1 .d5
ia4 42.�d4 a6 43.e4 c5 44.�f6
a
53
Garry Kasparov - Anatoly Karpov
h
1-0 Botvinnik - Smyslov, Moscow 1 964.
a
b
c
d
e
f
g
h
IO.d3
Kasparov chooses to live with the unpleasant
e3-pawn, hoping to surround it and win it later.
If White can carry out his ambitious strategy
successfully then the game will virtually be
over, on the other hand if he does not manage
to remove the pawn, then it might serve to cut
his position in two.
1 0.dxe3
Taking the pawn is also playable, and has
since acquired quite a large body of theory.
Kasparov later tried this move in one simul
and one regular game.
1 O . . . 'lWe7
1 0 . . . d6 1 1 .e4 h6 1 2. tLl h3 �e6 1 3 .g4 'lWd7
1 4.c5 dxc5 1 5 .'lWe 1 tLle5 1 6.'lWg3 �c4 1 7.l'l:e 1
tLl g6 l S . tLl f2 'lWe6 1 9 .h4 White went on to
win this irregular position, Kasparov - Rao,
Harvard (simul) 1 9S9.
1O ... b6 is the main line.
1 1 .tLl h3 'lWc5 1 2. tLl f4 'lWxc4 1 3 .e4 d6 1 4 .'lWd3
tLle5 1 5 .'lWxc4 tLlxc4 1 6.g4 l'l:bS 1 7.l'l:d 1 b6
l S .g5 tLl d7 1 9 .tLld5 l'l:b7 20.f4 tLl fS 2 1 .l'l:d4
54
The Prime Years
tLl a5 22.f5 i.d7 23J':1b 1 c5 24.Eld3 i.a4
25 .i.f4!? i.c2 26.Elb2 i.xd3 27.exd3
Karpov diverts the queen from the nice b3square.
12.�a3
1 2.1Wa4 is also possible, but it has been less
popular than Kasparov's move.
78
6
a
b
c
d
5
e
f
g
h
This was Kasparov - Sadvakasov, Astana 200 1 .
White has lovely positional compensation for
the exchange, nevertheless Black remains quite
solid and Sadvakasov eventually managed to
hold a draw.
It is hard to draw definite conclusions from
the above two games, especially since neither
of Kasparov's opponents employed the main
line of 1 O . . . b6. It should be noted that overall
White has scored less than fifty percent after
1 0.dxe3 .
10 ... d5
Karpov steers the play along a fundamentally
different course from the Botvinnik - Smyslov
game, and opens the position. Despite
White's two bishops, Karpov's choice is the
most logical and principled continuation,
as White is behind in development and his
uncoordinated pieces are not ideally placed for
a direct confrontation.
1 1 .�b3
Kasparov defends the c4-pawn and exerts
pressure on the d5-pawn. His move has grown
into the main line of the variation.
After 1 1 .cxd5 tLlxd5 one thing is clear: hardly
any game ends in a draw from this position.
n ... llJ a5!
4
3
2
1
���}�!
�
i;
if��
i
I% i if� �
%%
,,,,,/,� ,,,,,/,� _"".
/,�
. �� T�f N m
"�n "' .%%�'
.!.�B'
N.
% ��,% %%�,�
if� 8 if� 8 if�
'�/,� ,,,,,/,� '�/,� Jf�
��
&:,i{""
� ��
M /,�
�� ��7;;
" "
a
b
c
d
e
f
g
h
12 ... c6!
Karpov defends his knight while preparing to
recapture on d5 with the c-pawn, which could
offer him play on the c-file while also taking
the e4-square away from White's knight.
13.cxd5
If 1 3 .c5 h6 1 4. tLl h3 b6 Black has a healthy
position.
13 ... cxd5 14.f4
Kasparov wants to avoid retreating to the
h3-square with his knight. In the event of
other moves, Karpov would probably have
developed his light-squared bishop and played
. . . Elc8 soon after.
With the white pawn on f4, the position
is oddly reminiscent of a reversed Leningrad
Dutch, as in that opening White sometimes
ends up with a pawn on e6 following a minor
piece exchange on that square.
14 ... ttJc6
Nowadays 1 4 . . . i.g4 is the main move.
1 987
55
Garry Kasparov - Anatoly Karpov
15J�bl
1 5 .lLlf3 has also been tried, most notably by
Mamedyarov, who used it to defeat Papa in
Lausanne, 2004.
Zaitsev recommend 1 7.Elfe 1 which they
evaluate as slightly better for White. Such
conclusions should always be taken with
a pinch of salt, as it is unlikely that a world
class player will reveal all his ideas in a sharp
variation so soon after the event. In any case,
1 7 . . . Elad8 seems like the right response, and
Black has done well from this position in
practice.
1 7. lLl f3
a
b
c
d
e
f
g
h
15 ...�c7
This has been by far the most popular
choice in the position. Another plan would
be to prepare the move . . . ih3 to exchange the
light-squared bishops. If Black can carry out
this plan successfully, it will reduce White's
dynamic potential considerably.
16.ib2
Hunting the e3-pawn with 1 6J''1 f3 ?! is too
artificial: 1 6 . . . h6 1 7. lLl h3 (White could try
the piece sacrifice 1 7.Elxe3 Elxe3 1 8 .ixe3 hxg5
1 9.fXg5, but he can hardly hope for more
than equality here.) 1 7 . . . if5 1 8 .Elxe3 Elxe3
1 9.ixe3 Ele8 20.�e l We7 Black has a healthy
position.
1 6 ...ig4
Karpov develops quickly.
17.c4?!
Kasparov is somewhat impatient, and this
premature attempt to open the game brings
him no advantage.
In their Chess Informant notes, Karpov and
a
b
c
d
e
f
g
h
This looks like the critical move. The knight
will need to come back into play �t some
point, so White may as well combine it with
the defence of the e2-pawn.
1 7 . . . Wd7!?
My preference is for Black to meet the
impending c4 with . . . dxc4, rather than
blocking the centre with . . . d4. With this in
mind, the queen vacates the c-file.
The alternative is 1 7 . . . Elad8 1 8 .Elfe l h6
1 9 .c4 (Or 1 9 .ia1 Eld7 20.c4 d4 2 1 .Wc5
with a small initiative on the queenside.)
19 ... d4 20.ia 1 and White's queenside
initiative seems more relevant that Black's
kingside play.
1 8 .ia 1
1 8 .c4 dxc4 looks reasonable for Black.
1 8 .Elfe l ixf3 1 9.ixf3 Wh3 20.ig2 Wh5
2 1 .c4 d4! On this occasion Black blocks the
centre, and his kingside attacking chances
are quite real.
56
The Prime Years
This looks critical. It would be interesting to
know how Karpov would have reacted had it
occurred over the board.
22.Ei:c2 ttJ d5
The position is roughly balanced.
8
7
6
5
fm/.,J///·///· ,·,=,
4
3
2
a
b
c
d
e
f
g
h
1 B . . JhcB ! ?
This would b e m y preference, although
several other moves can be considered.
1 B . . . �h3 1 9 .�xh3 Wxh3 20.Ei:xb7 ttJ g4
2 1 .Wc5 ! is clearly better for White, as
pointed out by Karpov and Zaitsev.
1 B . . . b6 1 9.Ei:fc l �h3 20.c4! (20.�xh3
Wxh3 2 1 .c4 ttJ g4 22.cxd5 [22.Wd6 ttJ e7!J
22 . . . ttJ e7 23 .Ei:c7 ttJ f5 24.Wc3 f6 Black has
good chances on the kingside.) 20 . . .�xg2
2 1 .�xg2 d4 22.c5 White's queens ide
initiative is stronger than Black's kingside
play.
1 B . . . Ei:e7 1 9.Ei:fc l �h3 is given as unclear
by Karpov and Zaitsev, who mention the
continuation 20.�h 1 ttJh5 when ideas of
. . . ttJxf4 are in the air. However, White can
improve with 20.c4! when his chances are
higher as Black's kingside play is less strong.
1 9 .Ei:fc l b6 20.c4 dxc4 2 1 . Ei:xc4 �e6
a
b
c
d
f
e
g
h
17 dxc4!
•.•
Karpov corrected judges that he can allow
his kingside pawns to become doubled.
18.hf6 gxf6 19.ttJe4 @g7
1 9 . . . Ei:xe4? is not only risky, but also
ultimately incorrect: 20.�xe4 f5 ! (20 . . . �xe2?
2 1 .Ei:fe 1 ) 2 1 .�f3 ttJ d4 (Even worse is
2 1 . . . �xf3 ? 22.Ei:xf3 ttJ d4 23.Ei:xe3 c3 [23 . . . ttJ c2?
24.We7J 24.Ei:e5 c2 25 .Ei:c l b6 26.�f2 when
Black is in trouble as his knight will be sent
back.)
a
b
c
d
e
f
g
h
22.�d5! (Karpov and Zaitsev only mention
22.dxc4 �xf3 23.exf3 e2 24.Ei:fe l Wxc4 when
Black is fine.) 22 . . . ttJxe2t 23.�g2 Black has
problems as his pieces lack coordination.
a
b
c
d
e
f
g
h
a
a
b
c
d
e
57
Garry Kasparov - Anatoly Karpov
1 987
f
g
h
20.dxc4?!
This inaccurate move leads to trouble for
White. It was also necessary to avoid 20.h3?
ixe2 2 1 .'Dxf6 .txfl 22.Wl'c3 as Black takes
over with the lovely move 22 . . . E!:e5 ! . Three
other moves deserved attention though.
2o.1Wc3 Wl'd8! (20 . . . Wl'e7? runs into 2 1 .E!:xb7)
was pointed out by Andersson in the press
centre. Play continues 2 1 .E!:xb7 'D d4 22.Wl'xc4
I2lxe2t 23.'it>h l .te6 24.Wl'c5 Wl'd4 25 .Wl'c2 .tg4
and the position is unclear.
2o.1Wb2!? According to Kasparov this is one of
the most interesting options in the position.
20 . . . 1We7
2 1 .dxc4
(The
overaggressive
21 .l"1fcl ?! cxd3 22.E!:c5 ? backfires after 22 . . . h6!
23.exd3 E!:ad8 24.h3 .te6 when Black is much
better.) 2 1 . . . .tf5 22.'D c3 .txb l 23.'Dd5 Wl'd6
24.c5 1We6 The position is equal according to
Kasparov.
20.tiJd6
This leads to heavy complications, but the
position remains approximately balanced.
20 ... l"1ed8!
This is the best square for the rook, as it is
important to attack the d3-pawn.
Karpov and Zaitsev focus on another move,
which could lead to perpetual check. Their
analysis is nice, but it contains a hole.
20 . . . E!:e6 2 1 .'Dxc4 E!:d8
b
c
d
e
f
g
h
22.f5 (Instead the calmer 22.E!:fe l ! looks
good for White.) 22 . . . E!:ee8 23 .E!:b2 'D d4
24.E!:xb7 'Dxe2t 25.�h l 'Dxg3t! 26.�g l !
(26.hxg3 ? Wl'xg3 Black is winning) 26 . . . 'D e2t
(26 . . . 'Dxfl ? 27.E!:xc7 e2 28.Wl'xa7 e 1 =Wl'
29.E!:xf7t 'it>h6 30.E!:xf6t! 'it>g5 3 1 .Wl'g7t
'it>f4 32.Wl'h6 is mate.) 27.�h l 'D g3t White
cannot escape the checks, and Black also
has the option of continuing the fight with
27 . . . Wl'f5 ! ? It all looks good, except for the
improvement noted at move 22.
2 1 .'Dxb7
2 1 .'Dxc4 .txe2 does not help White.
2 1 . . . .txe2
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
22.E!:fe l
After 22.'Dxd8 .txfl 23 . 'D e6t fxe6 24.'it>xfl
Wl'd7 the position is unclear.
22 . . . E!:xd3 23 .Wl'c5 E!:c8!
Black can get away with this temporary piece
sacrifice thanks to the unfavourable position
of the knight on b7.
The Prime Years
58
24 . .ixc6
24.l"i:xe2? is refuted by 24 . . . tLl d4!.
24 ... Wxc6 25 .Wxc6 l"i:xc6 26.l"i:xe2 l"i:b6
27.l"i:xb6 axb6 28.Wg2 l"i:d7
Black traps the knight and will emerge with
an extra pawn in the rook ending, although
White should be able to hold it.
a
b
c
d
e
f
g
h
23.�hl?
Hopeless is 23 .Wxa7?! .ixe2 24.l"i:b 1 (24.l"i:e 1
Wb4 25 .Wf2 .ic4 and Black wins.) 24 . . . .id3
2 5 . tLl d2 Wc3 26.tLle4 Wc2 and Black wins.
a
b
c
d
e
f
g
h
20 J�ad8!
The best chance was 23.tLlf2! l"i:xe3 24.Wxe3
tLlxe2t 25.cJlh 1 .ie6 when Black has a solid
extra pawn, but the win is a long way off.
.•
Karpov bases his strategy on his fluent piece
play. With the last move he develops his last
dormant piece and prevents 2 1 .tLld6.
21 .13b3?
2 1 . tLl c3 should have been preferred, although
after 2 1 . . . l"i:d2 Black's active pieces give him the
advantage.
2 1 ... liJd4 22.13xe3
23 ... liJrs
This move does not spoil the win, but it was
not the most efficient route to victory. Stronger
was 23 . . . tLl c2! 24.Wb2 tLlxe3 2 5 .Wxf6t WgS
26.Wg5 t cJlfS 27.Wh6t We7 2S .Wg5 t Wd7
when Black wins as his king can escape to the
queenside.
24.13d3 Le2 25.13xd8 13xd8 26.13el
Finally White removes the troublesome
pawn, but it has cost him too much time.
22 ...Wxc4!
22 . . . tLl c2? would have let White off the
hook: 23 .Wc3 tLlxe3 24.Wxf6t WfS 25 .Wh6t
cJle7 26.Wg5 t! White must control the c5square. (Interestingly Karpov and Zaitsev
missed this move in their annotations, giving
only 26.Wf6t? cJld7 when the king escapes
and Black wins.) 26 . . . cJld7 27.tLl c5t Wd6
2S.tLlxblt White escapes with a draw.
a
b
c
d
e
f
g
h
1 987
Garry Kasparov - Anatoly Karpov
After making this move Kasparov famously
forgot to press his clock, a mistake which
cost him two precious minutes. By the time
he noticed his mistake and finally pressed the
button, he had less than one minute to reach
the time control.
26 ... !!e8!
With this excellent move Karpov sets up
various mating motifs based on White's
vulnerable first rank.
27.W1a5
One of the nice points behind Black's last
move is revealed after 27.lZl d6 lZlxd6 28 .W1xd6
i.f3! when White is defenceless.
27.l"lc l is not much better, and after 27 . . . Wi'd4
28 .'iWc5 l"lxe4 Black wins easily.
27 b5
..•
A more straightforward route to victory
was 27 . . . l"lxe4 28.Wi'xf5 , when 28 . . . iJ3 and
28 . . . 'iWc3 both win.
a b
29 ...1f3!
c
d
e
f
g
h
Once the defensive bishop has been
eliminated, White's kingside will collapse.
30.hf3 Wi'xf3t 31.'it>gl !!xel t 32.Wfxel lLle3
0-1
59
White can only avoid checkmate by giving
away his queen.
The next game was another g3 Griinfeld, which
resulted in a peaceful draw. In Game 4 Karpov
avoided repeating his . . . e3 pawn sacrifice in
the English, but was unable to equalize and
Kasparov converted his advantage to level the
match. But Karpov immediately struck back.
First he introduced a new idea in one of the
main lines of the Griinfeld, and accepted
a gambit pawn in a position where Black's
compensation has previously been assumed to
be adequate. It became known as the Seville
Variation, as it featured prominently in the
match. Kasparov reacted well, but fell into
time trouble and blundered at the end.
The next two games were drawn, without
either player being in serious danger, Then in
Game 8 Kasparov drew level after producing
one of his greatest ever positional masterpieces,
on the white side of a semi-blocked English
position. It can be found in the notes to Game
1 9 in the first volume.
The Griinfeld debate continued in the next
game. Karpov played another Zaitsev idea and
got some advantage, but Kasparov managed
to hold it. In Game 1 0 Kasparov switched to
l .e4 for the first time in the match, but got
nowhere against the Caro- Kann. In Game 1 1
Karpov once again got some advantage in a
Seville Griinfeld, but he returned Kasparov's
favour of Game 5 and also blundered and lost.
Game 12 was a Queen's Gambit Declined,
via an English move order. The position was
poised for a tense struggle, but it seemed
neither player wanted to risk anything and a
draw was agreed on move 2 1 . Thus Kasparov
led by three wins to two at the halfway point.
In Game 1 3 Karpov switched to the lZl f3
Exchange Variation against the Griinfeld,
but got no advantage and in the end it was
Kasparov who had chances to be better before
60
The Prime Years
the draw became inevitable. In the next game Kasparov once again got nowhere against the Caro,
and a quick draw ensued. Game 1 5 was more interesting, and Karpov has some chances to be
better in a Griinfeld, but eventually it ended in another draw.
In Game 1 6 Kasparov returned to the English, as shown in the note to Black's sixth move in the
main game above. He sacrificed a pawn but did not get enough play for it, and resigned the game
at the adjournment stage. The scores were equal. The next six games were all drawn. Both players
had to defend moderately unpleasant rook endings with the black pieces, but on each occasion
the defender was able to hold firm.
All this meant that the match was tied at 1 1 - 1 1 with two games to play. In his final white game,
Karpov played a rare line against the Griinfeld and kept a lot of pieces on the board. Black was
somewhat worse, but he defended well and the position remained complicated until Kasparov
blundered on move 50 when he overlooked a claver tactical finesse.
Kasparov now had to win the final game of the match to retain his title. When Karpov was
faced with the same situation two years before, he opted for a policy of outright attack. Kasparov
on the other hand went for a quiet opening. Karpov equalized, but perhaps the tension got to him
as he fell behind on time and drifted into an uncomfortable, though still defensible position. On
move 33 Kasparov blundered by missing a clearly winning continuation, but Karpov returned the
favour and missed an equalizing line. Once the dust had settled and the time control had been
reached, Kasparov had queen, bishop and four pawns on the kingside, versus Karpov's queen,
knight and three pawns. Crucially, Karpov's pawns were split into two islands, and he was unable
to hold it. Thus the match ended in a draw, and Kasparov retained his title.
Compared with the London/Leningrad match of the previous year, both players seemed to attach
more importance to safety. They both enjoyed a degree of success with their opening preparation.
Curiously, there were many more blunders then before, with both players winning two games
as a result of a blunder by the opponent. The only true masterpiece came in Game 8, which was
won by Kasparov. The final result demonstrates that the margin between success and failure can
be razor-thin at this level, and it was the closest Karpov ever came to dethroning his successor.
1987 Summary
Candidates final match versus A. Sokolov, Linares: Won 7Y2-3Yz (+4 =7 -0)
Brussels (3rd place) : 7/ 1 1 (+3 =8 -0)
Amsterdam ( l st-2nd place) : 4/6 (+2 =4 -0)
Bilbao ( l st place) : 7/9 (+5 =4 -0)
World Championship match versus Kasparov, Seville: Drew 1 2- 1 2 (+4 = 1 6 -4)
Total 6 1 . 5 % (+ 1 8 =39 -4)
iI:I Wins . Draws . Losses
1988
Rating 27 1 5 (2 in the world)
For the first time since 1 983, Karpov went a whole year without playing a world title match
against Kasparov. He made the most of this respite and played more tournaments then ever
before, making 1 98 8 the busiest year of his career in terms of the number of regular games
played.
First was Wijk aan Zee, where Karpov started by beating Curt Hansen (an extract from the
game is shown in the notes to Game 2 1 in the first volume) before suffering a temporary setback
against Nikolic. His remaining eleven games followed a nice pattern whereby he drew - sometimes
very quickly - with the black pieces, and won with white. We will look at his win from Round 5
against the former World Junior Champion, Kiril Georgiev.
The Bulgarian grandmaster faced all the world champions from Smyslov to Anand, with the
exceptions of Fischer and Petrosian. Out of thirty five games, he scored three wins, twenty two
draws and ten losses. Before the present encounter he had drawn one game with Karpov. Their
lifetime score stands at four wins to Karpov with four draws and one win to Georgiev. The last of
these results is rather tragic. Karpov had a queen and two pawns versus a queen, and was easily
winning, but was under the mistaken impression that more time would be added to his clock,
and lost on time on move 96.
I Game 9 I
Anatoly Karpov - KirU Georgiev
Wijk aan Zee 1988
I.c4
At the 1 986 Dubai Olympiad Karpov tested the Bulgarian grandmaster in the Dragon but
could only draw, so this time he switches to a calmer opening.
1...g6 v!lJf3 i.g7 3.d4 ltJf6 4.g3 0-0 5.i.g2 c6
According to the database Georgiev mainly played the Nimzo-Indian and had never tested the
present system prior to this game.
6.tLlc3 d5 7.\Wb3
Karpov chooses a line which he never used before and never repeated in the future. He obviously
wanted to avoid Georgiev's preparation based on the numerous encounters with Kasparov. The
fact that the Bulgarian was straying from his normal repertoire may also have contributed to
62
The Prime Years
Karpov's decision to depart from the main
theoretical paths.
7... e6
Georgiev keeps the position closed. The other
main possibilities are 7 . . .dxc4 and 7 . . . lMfb6.
8
7
6
5
4
8.0-0 lLl bd7
Black can also play 8 . . . b6 9 .j,f4 j,a6.
3
2
9.i.f4
1
a
b
c
d
e
f
g
h
12.b3!?
Karpov decides to give up his nice bishop. In
return he gains space and obtains a clamp in
the centre. His pawns restrict the g7-bishop,
and later he can look to develop play along the
g-file.
a
b
c
d
e
f
g
h
9 ... lLl b6?!
Georgiev wants to clarify the situation on
the queens ide, but he loses time and concedes
space.
Better options include 9 . . . b6 and 9 . . . dxc4
1 O.lMfxc4 tLl b6 I l .lMfb3 tLl fd5.
In the only other game that reached this
position White kept his bishop with 1 2.j,g5 ,
which led to an interesting fight: 1 2 . . . f6
1 3 .j,c 1 e5 1 4.b3 tLl a5 1 5 .dxe5 fxe5 1 6.j,g5
lMfc7 1 7.e4 d4 1 8 . tLl a4 b5 1 9 . tLl b2 j,e6 20.j,d2
h6?! 2 1 . tLl d3 White was a bit better in Lobron
- Keitlinghaus, Bad Woerishofen 1 989.
12 ... lLlxf4 13.gxf4 lLl a3 14.Wfd2 b5
Georgiev's strategy is to close the position as
much as possible. He manages to do it to a
large extent, but Karpov will still have ways to
improve his position.
10.c5 tLl c4 1 l .Wfc2
Karpov clears the way for his b-pawn,
anticipating the undermining move . . . b6.
1 1 . .. tlJh5
After 1 1 . . .b6 1 2.b3 tLl a5 1 3 .b4 ( 1 3 .:§:fe l !?)
13 ... tLl c4 1 4.tLle5 tLlxe5 1 5 .j,xe5 bxc5 1 6.bxc5
White obtained a small advantage and went on
to win in Korchnoi - Zhukhovitsky, Vilnius
1 953.
Black could also have considered preserving
some tension on the queens ide with:
1 4 . . . b6!? 1 5 .lMfb2
After 1 5 .tLl a4 Black can obtain the game
position a tempo up with 1 5 . . . b 5 , while
1 5 . . . j,a6 is also alright.
1 5 . . . tLlb5
Karpov also mentions the piece sacrifice
1 5 . . . bxc5 ? in his Chess Informant analysis.
1 6.lMfxa3 cxd4 1 7. tLl a4 j,a6 ( 1 7 . . . d3 1 8 .exd3!
1 98 8
63
Anatoly Karpov - Kiril Georgiev
Surprisingly Karpov misses this exchange
sacrifice in his analysis. 1 8 . . . i.xa l 1 9.1''lxa l
White's pieces are dominating and Black is
in big trouble.) 1 8 .�fe l d3 1 9.�ac l dxe2
20. ttJ c3 'lWb6 2 1 .ttJxe2 Black's two pawns are
no match for White's extra piece.
1 6.ttJxb5 cxb5 1 7.�fc l
Black's position i s solid with n o serious
weaknesses, but his pieces are passive.
17.e3
Karpov opens the path to transfer the bishop
to d3, which will also help to clear the g-file
for his rooks.
17... S?
Georgiev continues to block the position.
Indeed he prevents some of White's active
ideas, but in doing so he renders his pawn
structure too rigid. From now on Karpov has
to pay virtually no attention to Black's play,
and can focus all his energy on cracking Black's
defence.
If 1 7 . . . b4 1 8 . ttJ a4 ttJ b 5 1 9.i.f1 White is
somewhat better according to Karpov, who
suggests the subtle plan of ttJ b6 and later a3 .
Karpov also mentions 1 7 . . . 'lWc7!? when Black
still has the possibility of playing actively in the
centre with . . . f6 and . . . e5.
a
b
c
d
e
f
g
h
15J�fel
Karpov probably did not expect Georgiev to
handle the position in the way he did, and was
instead anticipating . . . f6 and . . . e5 to open the
centre.
15 .. J�b8?!
This rook move proves to be of little use in
the game. Maybe Georgiev played it because
of the following variation: 1 5 . . . 'lWc7 1 6.ttJe5
(White can also play 1 6.e3) 16 ... f6 1 7. ttJ d3 e5?
(l7 . . . �b8!) 1 8 .ttJxd5! White wins material.
Perhaps the best move was 1 5 . . .f6!? when it is
not easy for White to play for an advantage, as
if he plays e4 then the f4-pawn will become a
target.
16.�hl as?!
Like Black's previous move, this proves to be
largely superfluous.
a
b
c
d
e
f
g
h
1 8.iofl !
Karpov transfers the bishop and clears the
g-file for his rook.
18 ...iod7?!
Georgiev is overconfident regarding the
strength of his fortress, and thinks he can hold
it with natural moves.
1 8 . . . b4!?
64
The Prime Years
Black's best chance is to exchange his passive
bishop using the a6-square. However, he
must be careful to do it in a way that does not
allow White to crack open the queenside.
1 9. ttJ a4
a
b
c
d
e
f
g
23.V!fcl !
a
b
c
d
e
f
g
1 9 . . . 1"1aB!
Karpov mentions the line 1 9 . . . ttJ b 5 20. ttJ b6
ttJ c7 (20 . . . �a6 2 1 .a3!) 2 1 .a3! when Black's
queens ide pawns are vulnerable.
20.ttJb6 1"1a7
Black's disadvantage is smaller than in the
game, as his bishop gets to a6. Nevertheless
after securing the g-file White can exchange
the a3-knight from c2, and later open the
queenside with a3 .
19.i.e2 i.f6 20J�gl
Karpov starts his attacking operation using
the only file open to his rooks.
Karpov finds a clever way to improve his
knight.
23 h4 24.lLla4!
•••
The knight is heading for b6. On this square
it does not hurt Black directly, but it will
constantly threaten to infiltrate to the truly
fearsome d6-square via cB. For the moment
Black has several pieces guarding against this
invasion, but later those pieces will be needed
to defend Black's kingside, and could easily
become overloaded.
24 J�g7 25.V!ff1 !
•.
Karpov switches between flanks again.
25 1"1bg8 26.V!fh3 i.e8
•••
20 i>h8
•.•
In the long run the king will not be
particularly safe here. It was worth considering
a king transfer to f7 or e7, followed by
bolstering the pawn on g6.
2 U :�g3 V!fe7 22J�agl
Karpov not only builds his attack, but also
ensures that Black will not be able to free his
position with . . . g5 .
22 1"1g8
•••
Georgiev decides to defends the cB-square
with the queen. It is rather a pitiful role for the
strongest piece, but the position is closed, so
Black has chances to survive with his passive
pieces.
The other option was to leave the bishop on
d7 and use it to guard the cB-square, but
this was also not easy: 26 . . . WfB 27.Wh6 Wf7
2 B . ttJ b6 ttJ c2 29 .�d3 ttJ a3 30.h4 (30.1"1 l g2
We7) 30 . . . �eB (if 30 . . . We7 3 1 .ttJxd7 Wxd7
32.h5 We7 33.1"1xg6 White breaks in) 3 1 . ttJ cB
WfB 32.h5 (32. ttJ g5 ! ?) 32 . . . 1"1c7 33.WxfB 1"1xfB
1 98 8
34.hxg6 .txg6 3 5 .;gh3 i>g8 36.4:Jd6 Black's
position is extremely passive.
27.tt:l b6 YlVd8
a
b
c
d
e
65
Anatoly Karpov - Kiril Georgiev
f
g
h
28.Wlh6!
Karpov tightens the screw. The way he
has combined his play on both flanks is
exemplary.
28 .. J:i:c7 29.lLle5 �cg7
Karpov soon creates nasty threats with
his knights, so perhaps Black could have
considered 29 . . . .txe5!? 30.fxe5 ;gcg7 although
his position would remain desperately passive.
Karpov might have transferred a rook to d l
and prepared e4, while pushing the h-pawn is
another option.
30 ... lLl b5?!
Georgiev probably wanted to exchange to
b6-knight from c8, but he allows Karpov to
obtain a powerful passed pawn.
30 . . . .te7?
Blocking the seventh rank allows White to
tighten the noose.
3 1 .4:J bd7!
This witty move was mentioned by Karpov
in his annotations, although he did not give
any additional analysis. The point is that if
White can exchange Black's dark-squared
bishop without giving up the kni ght outpost
on e5, he will be able to break through on
the kingside.
Karpov also gives 3 1 . .ta6 as an alternative
winning plan, the idea being to transfer the
bishop to c8. This idea looks frightening,
but it turns out that after 3 1 . . ..tf6! 32 . .tc8
Vfie7 Black keeps his position together.
Although the main line is strong, the simplest
solution is probably 3 1 .;gh3! when White will
win the g6-pawn: 3 1 . . . .tf6 (3 1 . . . .tf8 32.;gxg6)
32.4:Jxg6t .txg6 33.;gxg6 White is a pawn up
and he has the better pieces as well.
30 .td3
•
a
b
c
d
e
f
g
h
3 1 . . . .th4
Alternatives are no better.
3 1 . . . .txd7 32.;gxg6 wins, as 32 . . . .tf8
33.4:J f7t! ;gxf7 34.;gxg8 is mate.
3 1 . . . .tf6 32.4:Jxf6 Vfixf6 33.;gg5 Vfie7 34.h4
Vfif6 3 5 .h5 White breaks in.
a
b
c
d
e
f
g
h
66
The Prime Years
32.1"1f3 j,xd7
32 . . . Wfe7 allows 33.ltJ b8! j,f6 34.lO bxc6
Wfc7 3 5 .1"1fg3 Wfc8 36.1"1 1 g2 Wfc7 37.h4 and
White wins.
33 .1"1h3 j,e8 34.1"1xh4 1"1b7 3 5 .1"1h3 Wff6
36.1"1hg3 lO b 5 37.j,xb5 1"1xb5 38 .h4
White will break through on the kingside.
3 1 .,hb5!
Karpov does not pass up the opportunity to
create a passed pawn. It may not hurt Black
immediately, but it could play a key role in the
subsequent course of the game.
31 ... cxb5 32.f3!?
Karpov sets a trap with this waiting move.
Black could have offered sterner resistance with
30 . . . Wfc7! , guarding the c8-square and using
the queen to defend along the seventh rank.
A good alternative was 32.h4 j,xe5 (32 . . . j,xh4?
33 .1"1h3 j,f6 [33 . . . j,xf2 34.1"1fl ] 34.lOxg6t
wins.) 33.fXe5 j,c6 34 .<;t>h2 Wfe7 3 5 .1"1g5
and with h5 coming soon, Black's position is
depressing.
32 ...�h4
The trap occurs after 32 . . . j,e7?:
a
b
c
d
e
f
g
h
3 1 .h4 (3 1 .1"1h3 j,d8 defends) 3 1 . . .Wfd8 32.1"1h3
(or 32.lOf3) 32 . . . Wfc7 3 3 . lO f3 White can
squeeze for a long time to come, although it is
not clear ifhe can win by force. Even supposing
that Black can draw with perfect defence, it
would be next to impossible to survive such a
position over the board against Karpov.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
33.lO bd7! when 33 . . . j,xd7? is refuted by
34.1"1xg6 j,f6 3 5 . 1"1xg7 1"1xg7 36.1"1xg7 j,xg7
37.1O f7t <;t>g8 38.Wfxg7t wins. Black can
survive for slightly longer with 33 . . . j,h4, but
after 33 .1"1h3! White wins as Black cannot
defend the g6-pawn.
Black's best chance to survive would have
been: 32 . . . Wfc7! Karpov mentions this move
in his annotations, adding the reply 33.h4
without further comment. Play might
continue 33 . . . Wfd8 34.1"1h3 (it is too early
for 34.h5?! gxh5) 34 . . . Wfc7 and here White
should probably make some preparations such
as transferring his king to the centre before
1 988
67
Anatoly Karpov - Kiril Georgiev
playing h5. It is hard to say if he is winning
outright, but over the board his chances would
certainly have been high.
Black could have kept himself in the game
with 34 . . . We7! 3 5 . tt:l cB! when there are two
possibilities.
a) 35 . . . Wc7 is refuted beautifully: 36.2"i:g5 !
WxcB (36 . . . .txe3 37.2"i:hg3 is winning, as
37 . . . .txf4 3B .Wxh7t! leads to mate.)
a
b
c
d
e
f
g
h
33J'�h3
This move looks obvious and it works well
in the game, nevertheless 33.1'Bg2! would have
been objectively stronger, for instance 33 . . . .tf6
34.h4 with excellent chances for White.
33 .tfl 34Jhg6
...
a
b
c
d
e
f
g
h
37.Wf6! Wc7 3B.2"i:g6!! Black is defenceless
against the sacrifice on h7 followed by mate.
b) However Black has a stunning tactical
defence: 3 5 . . . Wf6!!
It looks as though Black is on the verge
of collapse, but his resources are not yet
exhausted.
a
a
b
c
34 he3??
...
This loses a piece.
d
e
f
g
h
b
c
d
e
f
g
h
Exploiting the possibility of mate on g 1 .
36.2"i:xg7 Wxg7 37.Wxg7t 2"i:xg7 3B .2"i:g3! In
a way White is lucky to have this resource.
3B . . ..txg3 39.hxg3 ct?gB! (39 . . . 2"i:xg3 40.tt:ld6
2"i:gB 4 1 .c6 wins) 40.g4 �fB Despite being the
exchange down White is in full control of the
position, nevertheless it is not clear if he can
win.
68
The Prime Years
I GamelO I
35J3xg7!
Karpov now simplifies and win a piece.
Anatoly Karpov - John Van der Wiel
35 J3xg7
Wijk aan Zee 1 988
•.
l .c4 e6 2.c!lJc3 i.b4
The Dutch grandmaster played this line six
years earlier against Tal and drew. He employed
it later as well. Despite its resemblance to the
Nimzo-Indian, Karpov never got the appetite
to play it with Black.
3.g3
a
b
c
d
e
f
g
h
This is one of the most popular moves,
although there is no clear main line for
White.
3 llJe7
•..
36J3g3
The kingside threats, combined with the
passed c-pawn, are too much for Black to
handle.
36 '?;Ve7 37J3xg7 '?;Vxg7 38.'?;Vxg7t i>xg7
39.c6 hc6 40.c!lJxc6 .bf4 41.c!lJd7
1-0
.•.
After a quick draw with against Tal, Karpov
continued his winning run with the white
pieces against John van der Wiel, the only
Dutch player to have won the European Junior
Championship. Van der Wiel became a decent
grandmaster and on one rating list he was even
the Dutch number one, although I imagine
that the Dutch chess fans were hoping for even
more from him.
Van der Wiel faced the world champions
forty seven times. He won four of those games,
drew eighteen and lost twenty five. He found
Karpov a difficult opponent, as he scored j ust
seven draws against him, with eleven defeats
and no wins.
Van der Wiel develops the knight in a slightly
unconventional way.
4.i.g2 0-0
So far nobody seems to have tried the
provocative 4 . . . d5!?, which appears risky but
playable, for instance 5 . cxd5 (or 5 .Wfb3 ct:J bc6)
5 . . . exd5 6.Wfb3 ct:J bc6 7.ct:Jxd5 .te6 8.e4 Wfd6
and Black is in the game.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
5.'?;Vb3
Karpov decides not to play with doubled
pawns on the c-file.
1 98 8
69
Anatoly Karpov - John Van der Wiel
5 c5 6.a3 .ta5 7.e3
...
Karpov is not tempted to snatch a pawn
with 7.�xb7 �xb7 8.'lWxb7 lLl bc6 9.'lWb3
1"1b8 1 0.'lWd 1 lLld4 when Black has decent
compensation thanks to his play on the
b-file.
7 lLlbc6 8.lLlge2
...
Karpov avoids lLl f3 in order to inhibit . . . d5.
8
a
b
c
d
e
f
g
h
�b6?!
...
This is too slow.
8.. .'tJe5 also falls short of equality: 9 . 0-0
(9.'tJe4? d5!) 9 . . . 'lWb6 (After 9 . . . lLl d3 1 O.'lWc2
'tJxc1 l 1 .Ei:axc 1 d6 1 2.d4 Black has the bishop
pair, but White has more space in the centre
and is better developed.) 1 0.'lWc2 lLlxc4 1 1 .lLl a4
1Mfb5 1 2.d4 Black will not be able to keep his
extra pawn, and his pieces lack harmony.
The most principled continuation is:
8 ... d5!
This pawn sacrifice offers promising
compensation.
9.cxd5
If9.0-0 lLl e5 1 O.d3 dxc4 1 1 .dxc4 'lWd3 Black
will take the c4-pawn and equalize.
9 . . . 'tJe5!
Black's play is based on the weakness of the
d3-square.
a
b
c
d
e
f
g
h
1 0.dxe6
After 1 0. lLl f4 c4 1 1 .'lWc2 exd5 1 2. lLl cxd5
lLlxd5 1 3 .lLlxd5 �h3! White has problems.
There is also 1 0. 0-0 c4 1 1 .'lWc2 exd5 1 2. lLl f4
�f5 1 3 .'lWd1 �xc3 1 4 . dxc3 �e4 when Black
has a free position and White's dark-squared
bishop is unimpressive.
1 O . . . lLl d3t
1O ... c4 is also promising.
1 1 . \t>fl �xe6
Also after 1 1 . . .fxe6!? 1 2. lLl f4 lLl g6 1 3 .�e4
lLl ge5 Black has compensation.
1 2.'lWxb7 Ei:b8 1 3 .'lWxa7 �c4
White has three extra pawns, but he is
dangerously tied up.
9.0-0 tlJa5 IO.'lWal
White chooses this square instead of a4 as
the queen will be less exposed, and she also
helps to keep the d5-square under control.
IO ... d5!?
The Dutch grandmaster plays actively. His
cards are similar compared with 8 . . . d5! as
analysed above, but since then White has been
dealt an ace.
1 O . . . lLl ac6 stops d4 but allows l 1 .b4! (There
is also l 1 .Ei:dl ! ? when Black should play
1 1 . . .e5.) 1 1 . . .cxb4 1 2.axb4 lLlxb4 1 3.'lWb3
lLl bc6 and now after 1 4.d4 or 1 4.�a3 White
has beautiful compensation in a position
somewhat reminiscent of the Benko Gambit.
?O
The Prime Years
12.dxc4 .id7
Black does not mind White's queens ide
expansion. With 1 2 . . . ttJ ac6 he could try to stop
it. 1 3 .Ei:b l (Another interesting idea is 1 3 .Ei:dl
We? 1 4.b4!? cxb4 1 5 . ttJ b 5 Wb8 1 6.axb4
ttJxb4 1 ? Wb3 ttJ a6 1 8 .i.a3 with promising
compensation for the pawn.) 1 3 . . . a5 1 4.Ei:dl
We? 1 5 . ttJ b 5 Wb8 1 6.b3 White's bishops will
be well placed on the long diagonals, and he
remains in control of the game.
13.b4! ttJac6 14J:�b l Wfc7
1 l .d3!
Black's idea was to meet I I .cxd5 with
1 1 . . . c4! 1 2.dxe6 i.xe6 when his grip on the
light squares offers him decent compensation.
Instead Karpov opts for a simple and strong
continuation, which highlights the fact that
although Black was able to gain some space, he
lacks the ability to support his pawn centre.
8
7
6
5
4
3
1 l ... dxc4
2
Having covered the c4-square, White was
threatening to take on d5. Van der Wiel opts
for a symmetrical pawn structure, but in the
resulting position White's pieces are more
favourably placed, especially his "Catalan"
bishop on g2.
1
Black can no longer sacrifice under favourable
conditions: 1 1 . . .ttJ ac6?! 1 2.cxd5 (White could
even try the gambit 1 2.b4!? cxb4 1 3 .axb4 ttJxb4
1 4.'1Wb l ttJ bc6 1 5 .i.a3 with lovely play for the
pawn.) 1 2 . . . exd5 1 3 .ttJxd5 ttJxd5 ( 1 3 . . . i.e6?!
1 4 . ttJ ef4) 1 4.Wxd5 Wxd5 1 5 .i.xd5 Ei:d8
1 6. ttJ c3 Black's compensation is insufficient.
1 1 . . . d4 is possible, but it allows White to obtain
a turbocharged reversed Benoni position:
1 2.exd4 cxd4 1 3 . ttJ a4 ( 1 3 . ttJ e4 ttJ ac6 1 4.b4
is also good) 1 3 . . . i.c? 1 4 . b4 ttJ ac6 1 5 .ttJc5
White has mobilized his queens ide pawns and
keeps some advantage.
a
b
c
d
e
f
g
h
15.Wfb3?!
The idea behind this somewhat mysterious
move is to develop the bishop on b2, but
if White wanted to play this way then he
should have done so on the previous move,
instead of wasting time with Ei:b 1 . Objectively
he should have preferred one of the
alternatives:
1 5 .b5 could lead to something resembling the
game continuation after 1 5 . . . ttJ e5, although
Black also has the option of 1 5 . . . ttJa5!?
1 5 .ttJ e4 ttJ d8 ( 1 5 . . . f5 ? 1 6.ttJxc5) 1 6.bxc5 i.xc5
1 ? ttJxc5 Wxc5 1 8 .i.d2 We? 1 9. ttJ d4 Black
has a sounder pawn structure, but White has
the better placed pieces and keeps the overall
advantage.
1 98 8
71
Anatoly Karpov - John Van der Wiel
1 5 .ttJb5 Wb8 1 6.�d l �d8 1 7.Wc2 and White
keeps the upper hand.
IS ... lLl eS!
Van der Wiel attacks White's only weakness,
forcing a closing of the queenside.
16.bS
Apart from safeguarding the c4-pawn, it
is also useful to take away the c6-square from
Black's pieces. His light-squared bishop will have
a particularly hard time getting into the game.
16 .. J�ad8?
Van der Wiel occupies the only open file.
He was probably hoping to exchange at least
one pair of rooks, in order to create some
extra breathing room for his minor pieces.
Nevertheless the text move is not purposeful
enough. Instead Black should have taken
immediate steps to improve the prospects of
his minor pieces.
1 6 .. .f5 ! ? looks provocative, but is not easy to
refute: 1 7.f4!? ( 1 7 . .ib2 .ie8 Black improves his
bishop.) 1 7 . . . lLl d3 (another idea is 1 7 . . . lLl g4
1 8 .e4 �adS) 1 8 . .id2 .ieS Black's position
looks playable.
Curiously, this move reminded me of a game
from the 1 975 World Junior Championship
in Graz, when I lost to none other than Van
der Wiel himself! In that game, a French
Advance, the Dutchman played a fairly
early . . .f6 and later activated his bishop via
eS and g6, to good effect. I think he should
have employed a similar idea in the present
game.
1 7 . .ib2
If 1 7.f4 lLlf7 the knight stands much better
than the way it ends up in the game. After
the natural 1 8 .e4 Black can try 1 s . . . lLl d6
1 9. 1Ll a4 .ia5 , as well as 1 S . . . �adS followed
by . . . .ia5 , . . . b6, . . . .ics and . . . .ib7.
1 7 . . . .ie8 1 S . lLl f4
1 S .f4 lLl g4 100ks alright for Black.
1 S . . . .if7
Black should be able to live with his spatial
disadvantage.
8
7
6
5
4
3
1 6 . . .f6!
2
1
a
b
c
d
e
f
g
h
17.f4!
Over the next three moves Karpov gains
considerable space in the centre.
a
b
c
d
17 ... lLl 5g6?!
e
f
g
h
This would be my preference. Black gives
his pieces some additional options, without
weakening his dark squares.
When deciding where to place a knight, one
should not only think of the first move, but also
take into account the second and sometimes
further steps which the knight might take.
From its present post on g6, the knight only
The Prime Years
72
has the miserable hB-square available. With
this in mind, it becomes clear that 1 7 . . . tLl g4!
was correct, and after I B .h3 tLl h6 1 9.e4 f6
Black has better defensive chances than in the
game.
18.e4! f5 19.e5
8
7
6
5
4
3
20.�e3!
White should play against the c5-pawn.
Instead after 20.�b2? �cB 2 1 .E1fd l b6 Black
is close to equalizing.
20 . . . b6!
The immediate 20 . . . �cB allows White to
increase his initiative in a surprising way:
2 1 .b6!! axb6 (Other captures are worse as
they do not enable Black to keep the extra
pawn: 2 1 . . .�xb6 22.tLla4!±; 2 1 . . .'lWxb6
22.'lWa4 'lWa6 23 .�xc5±) 22.tLlb5 'lWbB
23.tLld6 tLlhB 24. tLl c3 tLlf7 2 5 . tLl cb5 White
maintains strong pressure on the queenside.
2 1 .E1bd l
Black must tread carefully, as shown by some
of the following variations.
2
1
a
b
c
d
f
e
g
h
19 ....ie8?
In a way Van der Wid is playing in a similar
manner to Karpov, in the sense that, having
chosen a plan, he follows it through. However,
he would have done better to readj ust and
abandon his plan of playing on the d-file.
1 9 . . . �a5 !
This was the best chance. Black's idea is to
bolster the c5-pawn with . . . b6 and then
exchange his light-squared bishop on b7.
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
2 1 . . .E1feB!?
This waiting move may well be the best
chance.
2 1 . . . tLl hB 22.tLla2 ! tLlf7 (22 . . . �cB 23 .'lWa4!?
[23 . tLl b4 is also good] 23 ... �b7? [23 ... a6
24.tLl ac3] 24.�xb7 'lWxb7 25 .�xc5 White
wins a pawn.) 23.tLlb4 �cB 24. tLl c6 tLlxc6
2 5 . bxc6 The strong c6-pawn may hurt
Black.
2 1 . . . �cB 22.tLld5! exd5 (22 . . . 'lWbB 23.tLlxe7t
tLl xe7 24.'lWa4 'lWc7 2 5 .tLl c l ! The bishop on
a5 is vulnerable.) 23.cxd5 tLlxd5 Black has to
give back the piece. 24.�xd5t i>hB 25 .'lWc4
White is clearly better.
22.tLla2 �cB 23 .'lWc2
After 23.tLlb4 �b7 Black seems to be okay.
1 98 8
73
Anatoly Karpov - John Van der Wid
23 . . . ib7 24. 4:l ac l a6 25 .ixb7 'lWxb7 26.4:l b3
White keeps the advantage, but Black has
chances to resist.
2 1 ...i.h5 22.�d l;d7
If 22 . . . 2"1c8 23.4:ld3 ie2 24.4:l dxc5 wins.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
20.ie3!
Karpov spots that the c5-pawn is weak,
despite the fact that Black is defending it
twice. In the next few moves he goes after it
purposefully until it falls.
a
b
d
c
f
e
g
h
23.lLl d3!
Having won the strategic battle, Karpov
shows that he is equally adept at controlling
the complications.
23 .'lWc2 was also good enough: 23 . . . 'lWc8
24. 4:l d3
20 ... 4:lh8?!
Van der Wid hopes to counter White's
assault on the c5-pawn by becoming active
on the d-file, so he makes way for the light­
squared bishop to come to h5. Nevertheless
the knight is pitiful in the corner, and Van der
Wiel never gets around to moving it again .
The attempt t o disrupt the opponent with
20 . . 2"1d3 was not much of an improvement,
and after 2 1 .if2 ia5 22.'lWc2 the rook has to
retreat.
a
.
b
c
d
e
f
g
h
24 . . . 2"1fd8 (24 . . . 2"1c7 25 .'lWf2) 2 5 . 4:l dxc5 ixc5
26.4:lxc5 2"1c7 27.4:lxb7 2"1xb7 28 .ixb7 'lWxb7
29.c5 And White wins.
Black's best chance was to defend passively:
2o . if7 2 1 . 4:l a4 2"1fe8 22.4:l c l 4:l f8 23.4:ld3
ttJ d7 Black has avoided a direct collapse,
although the chances of surviving such a dismal
position against Karpov would be slim.
23 ...i.e2
21.tlJa4!
24.�dxc5
. .
Karpov continues his plan.
This is the critical move, but Karpov has
everything under control.
Finally the c5-pawn succumbs.
74
The Prime Years
24 .. J�d3!?
Black tries to stir things up, but it is not
enough. If 24 . . . i.xfl 2 5 .<j;lxfl White wins as
the e6-pawn falls.
a
b
c
d
e
f
g
h
25.liJxd3! Le3t 26.i>hl �d8 27.tLlb4 i.xfI
28.�xfl
White is not only a pawn up, he is also
playing with virtually an extra piece, as the h8knight exists but does nothing.
28 ...�c5 29.liJd3
29.ltJ c6 is also convincing.
29 ... b6 30.tLlaxc5 bxc5 3 1 .a4 �d4 32.a5
Wlxa5 33.tLlxc5 i>f7 34.liJb7 Wld2 35.b6
1-0
Black resigned, as the e6-pawn will soon fall
as well.
For the rest of the tournament Karpov
continued his pattern of drawing, often
quickly, with Black and winning with White.
Over the next four rounds his draws came
against Sosonko and Hubner, and he defeated
Lj ubojevic and Farago. His final draw came
against Andersson, who was trailing Karpov
by half a point, but made no serious attempt
to win. Thus Karpov finished in outright first
place with a score of 9/ 1 3 .
Karpov remained i n the Netherlands for his
next event, the Euwe Memorial in Amsterdam,
a four player double-round-robin tournament.
Karpov drew with Timman and beat Short
but then lost to Lj ubojevic. He then won
an interesting endgame with rook and three
connected pawns versus rook and bishop
against Timman. In the next round Karpov
narrowly failed to beat Ljubojevic, then in
the final game he took a quick draw against
Short, and finished in second place half a point
behind the Englishman.
Karpov's next tournament took place in
Brussels, and was the first in a series of events
known collectively as the World Cup, which
took place over the course of 1 988 and 1 989.
Karpov began by drawing with Salov, then he
beat Nogueiras. Two more draws with Tal and
Portisch followed. Karpov then lost to Beliavsky
and dropped back to fifty percent, but not for
long as he beat Timman in an imaginative
game. He followed it with a quick draw with
Nunn, then powered ahead with three fine
victories over Andrei Sokolov, Lj ubojevic and
Speelman. In the last six games Karpov scored
two more wins over Korchnoi and Winants,
and four draws. His total of 1 1 / 1 6, with seven
wins, eight draws and one loss, was enough to
secure first place a point ahead of Salov.
Karpov's next event, in Amsterdam, was an
extended version of the traditional round­
robin format, as each contestant played four
games against each of his three opponents. The
event was especially notable as it 'reunited'
Karpov with his nemesis Kasparov. The other
two invitees were Timman and Van der Wiel.
Karpov crushed the latter by the score of 3 Yz­
Y2, but he only managed four draws against
Timman. Karpov's result against Kasparov
was more disappointing: he lost both of his
black games, and could only draw the white
ones. One of the losses could easily have gone
1 98 8
75
Anatoly Karpov - Garry Kasparov
the other way, as Kasparov embarked on an
unsound sacrificial attack and Karpov missed
several wins; nevertheless in the end he was
unable to cope with the problems and lost on
time in what was by then a losing position.
Karpov finished in second place with 6Y2/ 1 2 ,
a massive 2Yz points behind Kasparov. There
were times in the tournament when Karpov
played extremely good chess, nevertheless
the result made it clear that Kasparov was on
another level.
For his next tournament Karpov travelled to
Belfort, in eastern France, for the next stage
of the World Cup series. He started badly
with a loss to Andrei Sokolov, but bounced
back immediately with a win over Hjartarson.
Over the next eleven games Karpov performed
solidly, winning four games against Nogueiras,
Beliavsky, Ehlvest and Timman, and drawing
the other nine. Then in the penultimate round
he faced Kasparov, who had already won the
tournament.
I Game 11 1
Anatoly Karpov - Garry Kasparov
Belfort 1 988
1 .d4 ttJf6 2.c4 g6 3.ttJc3
Karpov was never able to put much pressure
on Kasparov with the 3.g3 Grunfeld, so he
plays something more combative.
3 ... d5 4.cxd5 ttJxd5 5.e4 ttJxc3 6.bxc3 i.g7
7.ic4 c5 8.ttJe2 ttJc6 9.i.e3 0-0 10.0-0 i.g4
1 1 .8 ttJa5 12.i.xf7t!?
The players continue the opening debate
which began in Seville the previous year. Igor
Zaitsev's idea has grown into a major line, and
nowadays approximately 500 games can be
found on the database.
12 .. .l:hf7 13.fxg4 �xf1t 14.�xf1 �d6
1 5.e5
The move safeguards the e-pawn and freezes
the g7 -bishop, but gives up some squares in
the centre. In the eleventh game of the Seville
match Karpov played 1 5 .�gl and obtained
some advantage before he blundered and lost.
1 5 ...�d5 16.i.f2
Four years later Kasparov faced 1 6.g5 and
demonstrated a reliable antidote: 1 6 . . . '\W e4
1 7.i.f2 �fS I S . ttJ g l ttJ c4 1 9 . ttJ f3 ttJ e3t
20.i.xe3 '\Wxe3 2 1 .'\Wb3t �hS 22.�e l Black
soon gave a perpetual, Beliavsky - Kasparov,
Linares 1 992.
16 ... �d8
In the very first game featuring the Seville
Variation, Kasparov chose 1 6 .. J'l:fS and was
doing all right until he blundered shortly before
the time control. Afterwards he switched to the
text move and seemed to be content with the
positions he got.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
17.�a4!?
Karpov unveils an important novelty. White's
idea is to provoke . . . b6 in order to weaken the
c6-square.
The two players had already debated this
position twice before. The first game continued:
1 7.'\We l '\We4 I S .g5 '\Wf5 1 9.h4 ttJ c4 20.�gl
76
The Prime Years
'lMfg4 2 1 .a4 h6 22.�a2 hxg5 23.'lMfb 1 gxh4
24.'lMfb3 'lMfe6 2 5 . lZl f4 'lMffl 26.lZlxg6 'lMfxg6
27.'lMfxc4t cj;Jh8, Black was okay and the game
was later drawn, Karpov - Kasparov, Seville (7)
1 987.
At the more recent Amsterdam tournament
Karpov deviated with: 1 7.'lMfc2 'lMfc4 1 8 .'lMfb2
i.h6 1 9.h4 'lMffl (Interestingly, despite his
success in the main game below, Karpov came
back to the present line the following year
against Timman. The game continued: 1 9 . . . �f8
20.g5 'lMfd3 2 1 .'lMfb 1 'lMfe3 22.'lMfe 1 i.g7 23.cj;Jg1
'lMfe4 24.lZlg3 'lMfxh4 2 5 . lZl e4 �xf2 26.lZlxf2
cxd4 27.�d 1 White had a clear advantage and
went on to win, Karpov - Timman, Rotterdam
1 989.) 20.i>gl �f8 2 1 .lZlg3 lZl c4 22.'lMfe2
'lMfxf2t 23.'lMfxf2 i.e3 24.'lMfxe3 lZlxe3 25 .dxc5
�c8 26.�b 1 �xc5 Black was fine and in the
end it was White who had to work for the
draw, Karpov - Kasparov, Amsterdam 1 98 8 .
17... b6
19.i>gl %Vc4
a
b
c
d
e
f
g
h
20.%Vd2!
Karpov is flexible in his thinking. 20.'lMfe4
is no longer as attractive in view of 2o . . . lZl c6!,
taking advantage of the vulnerable knight
on e2, and meanwhile Black threatens
. . . lZlxe5.
20 %Ve6?!
.••
On 1 7 . . . �f8 White must avoid 1 8 .'lMfxa5 ??
'lMffl, but should instead step aside with
1 8 .i>g l ! .
Kasparov tries to play safe. White's extra
pawn may be doubled, yet its influence can
still be felt.
18.%Vc2! �f8?!
20 . . . 'lMffl was also not ideal, as after 2 1 . lZl g3!
intending lZl e4 and 'lMfe2, White improves his
pieces considerably.
Kasparov is widely considered the greatest
opening player of all time, but even he can
have trouble when confronted with a novelty.
It is understandable that he chose not to repeat
his play from the recent game when he had the
same position with the pawn on b7 instead of
b6, as Karpov would obviously have analysed
that continuation in depth. In the event of
1 8 . . . 'lMfc4 Karpov would probably have replied
1 9 .'lMfe4! when the black knight does not have
the use of the c6-square.
Subsequent analysis revealed that 1 8 . . . �c8! is
the right move, and Black has scored reasonably
well with it in practice.
Exchanging or improving the bishop looks like
the best idea:
20 . . . i.h6! 2 1 .'lMfxh6
After 2 1 .i.e3 i.xe3t 22.'lMfxe3 cxd4 23.cxd4
'lMfc2 Black's kingside is a bit open, but he
should be able to live with it.
However 2 l .'lMfe 1 ! ? is a promising alternative,
and after 2 1 . . . lZl c6 22.h4 e6 23.g5 i.g7
24.'lMfd2 White has somewhat better
chances.
2 1 . . .'lMfxe2 22.'lMfe3
Karpov ends his analysis here, evaluating the
position as slightly better for White.
1 98 8
77
Anatoly Karpov - Garry Kasparov
Karpov prepares to bring his rook into
play and also ensures that Black's bishop
remains passive. In addition Black has to start
calculating whether 23.lLlf4 or 23 . .ih4 are
threatened.
22 h6!?
...
a
b
c
d
e
f
g
h
22 1Wb2!
This must be the critical continuation.
23J�d l
23.Ei:fl is also possible.
23 . cxd4 24.cxd4 'lWxa2 2S .dS lLl c4
Here White can choose between 26.'lWe l
and 26.'lWd4, to which Black would probably
reply 2G . . . lLlxeS. In both cases Black's position
is dangerous as his king is exposed and White
has a potential passed pawn in the centre, but
objectively his position may be playable.
. . .
. .
Kasparov voluntarily weakens his kingside
and restricts his bishop in order to disrupt
White's smooth development. Alternatives
were uninspiring, for instance 22 . . . 'lWf7 23.�fl
(23 . .ih4 comes into consideration as well)
23 . . . e6 24. lLl g3 'lWd7 2 S . lLl e4 and Black has no
compensation for the missing pawn.
23JWcl
8
7
6
5
4
21 .h3
Karpov consolidates his extra pawn and also
gives his king an escape square, which can
make a big difference in some lines.
2 l . .. tD c4
3
2
1
a
b
c
d
e
f
g
h
23 'i'f7!
...
The attack on the bishop prevents White
from getting organized.
If23 . . . 'lWdS 24.'lWc2! i>h7 2 S . lLl g3 lLl a3 26.'lWb2
lLl c4 27.'lWe2 White is firmly in control. Soon
he will develop his rook and then put his
knight on the dominating e4-square.
a
22.'lWg5!
b
c
d
e
f
g
h
Karpov mentions 23 . . . hS to open the kings ide,
but exchanging the opponent's doubled pawns
is a big concession for Black, and after 24.gxhS
gxhS 2S .'lWgS �fS 26.'lWg3 White is a pawn up
for nothing.
78
The Prime Years
24 .ig3
.
The bishop has to block the route of the
knight from g3 to e4, on the other hand the
possibility of ltJ f4 now becomes attractive.
a
b
c
d
f
e
g
h
24 g5?
...
Kasparov stops ltJ f4, but he weakens his
kingside too much.
24 . . . 1.Wd5 !
Activating the queen was Black's best
chance.
2 5 . ltJ f4
And not 25 .1.Wc2?? ltJ e3 .
25 . . . 1.We4 26.ltJe6
26.1.Wb l ! ? �xf4 27.1.Wb3 is tricky, but does
not lead to a clear-cut win: 27 . . . 1.We3t!
(27 ... 1.Wd5 ? 2S . .ixf4 cxd4 29.�d l d3
30.1.Wb l +-) 2S.�h2 �h7 29.1.Wxc4 cxd4
30.�e 1 1.Wd2 3 1 ..ixf4 1.Wxf4t 32.c;f;>gl 1.Wd2
33 .1.We2 1.Wxc3 White certainly has winning
chances, but it is hard to say if it will be
enough for a full point.
a
b
c
d
e
f
g
h
26 . . . �cS!
The exchange sacrifice is not sufficient:
26 . . . cxd4? 27.ltJ xf8! (After 27.cxd4 ltJ e3!
2S .1.Wd2 [2S .1.Wb2 �cS] 2S . . . ltJ c4 29 .1.We l
ltJ e3 30.1.We2 �cS Black is very much alive.)
27 . . . ltJ e3 28 .1.Wd2 dxc3 29.1.Wf2! (This move
is even stronger than Karpov's 29 .1.We2,
which is good but not immediately decisive.)
29 . . . .ixfB 30.�e 1 c2 3 1 ..if4 White wins.
26 . . . ltJ e3 ? looks active but Black just steps
into a lethal pin. 27.1.Wd2 ltJ c4 (27 . . . cxd4
28.�e l !+-) 2S .1.We l ltJ e3 29 .1.We2 cxd4
30.ltJxfS (30.cxd4? �cS!) 30 . . . .ixfS (30 . . . d3
3 1 .1.Wf2) 3 1 .cxd4 1.Wxd4 32.�e l White is
winning, as all the discovered checks are
harmless.
27.1.Wb l !
This looks best, although 27.1.Wf4 is also good:
27 . . . 1.Wxf4 (It is risky for Black to keep the
queens on the board: 27 . . . 1.Wd5 ?! 2S.ltJxg7
�xg7 29.e6! �fs [29 . . . 1.Wxe6 30.�e l ] 30.1.Wc7
Black has problems.) 2S.ltJxf4 g5 29.ltJd3
White is better, but his advantage is smaller
than in the main line below.
a
b
c
d
e
f
g
h
27 . . . 1.Wxb l t
27 . . . 1.We3t ? 2S . .if2 1.Wxc3 29.1.Wxg6 leads to
a checkmate.
2S.�xb l c;f;>f7 29.ltJc7!
This cute finesse is a useful resource.
29 . . . e6 30.�f1 t c;f;>e7 3 1 . ltJ b 5
White has decent winning chances, but the
game is far from over.
79
Anatoly Karpov - Garry Kasparov
1 98 8
25.�c2!
Karpov prepares to activate his rook,
meanwhile the queen stands excellently on the
b l -h7 diagonal.
tLlxe5 (27. . . ixe5 28 .Wg6t ig7 29.tLl d4 :9:f6
30.tLlf5 +-) 28.:9:dl Wc4 29.tLlg3 e6 30.tLlh5
Black is lost.
27 :9:f7
.••
25 ...�d5
In the event of a waiting move, Karpov could
have improved his position with :9:e l and e6.
a
b
c
d
e
f
g
Kasparov must make a defensive move,
as 27 . . . b4 is refuted by 2 8 . tLl f5 :9:0 29.e6
Wxe6 30.:9:e 1 Wd7 3 1 .cxb4 as pointed out by
Zaitsev.
h
26.ifl!
Karpov defends the d4-pawn and vacates
the g3-square for the knight. The game has
reached a critical phase. Both players must
have been conscious of the fact that if White
were able to bring his rook into play and
activate his knight, then the game would be
almost over. But Kasparov is one of the most
tenacious players of all time, and possesses an
unparalleled ability to stir up complications
from difficult positions.
a
b
c
d
e
f
g
h
28.:9:el
Karpov brings his last piece in play, although
the more aggressive 28.tLlf5 ! ? cxd4 29.:9:d l was
also good enough.
Another route to victory was 28 .Wg6!?
cxd4 29 .ixd4 tLlxe5 30.ixe5 Wxe5 3 1 . tLl f5
:9:xf5 32.gxf5 Wxc3 33.:9:dl with a winning
position.
28 b4 29.Wg6
•.•
26...b5!
Attempting to undermine White's pawn
structure is the best chance, although ultimately
it is insufficient.
27.lLlg3
True to his style, Karpov gives as little
counterplay as necessary.
White's advantage is so great that he could
also afford to open the position: 27.dxc5 !?
a
b
c
d
e
f
g
h
The Prime Years
80
29 ... �f8
A desperate attempt to confuse White.
29 . . . bxc3 was no better: 30.tt:lh5 (30 . tt:l f5 mf8
3 1 .e6 2::1 xf5 32.gx:f5 tt:l d6 33.dxc5 also wins)
30 . . . cxd4 (30 . . . e6 3 1 .tt:l f6t) 3 1 .e6 White
wins.
8
7
6
30.lLle4
Karpov keeps improving his pieces. This
does not spoil anything, but it was possible
to win by more direct means: 30.tt:lh5! e6
(30 . . . bxc3 3 1 .e6) 3 1 .tt:l f6 j,xf6 32.exf6 Black
is hopelessly lost.
5
4
3
2
1
30 .. Jhf2!?
Objectively this desperate exchange sacrifice
hastens the end, but practically it was probably
the best chance.
The calmer 30 . . . cxd4 3 1 .cxd4 gives Black
little chance of saving the game, for instance
3 1 . . .a5 (Or 3 1 . . .Wg8 32.2::1 e2 a5 33.tt:lg3 and
Black will not be able to live with the knight
on h5.) 32.tt:lc5 mg8 33.tt:le6 a4 34.2::1 fl b3
3 5 . axb3 axb3 36.j,g3 with a mating attack.
3 1 . �xf'l bxd
8
7
6
5
4
3
2
1
a
32 ... �g8 33.Wfc8t! �h7 34.Wfxc5 Wff7t
35.�gl c2 36.lLlg3 i.f8 37.llJrs �g8
b
c
d
e
f
g
h
32.Wfrst!
With this and the following check White
wins the c5-pawn and the game.
a
b
c
d
e
f
g
h
38 ..Ekl
1-0
The c-pawn was Black's last hope, and it will
perish on the next move. Amazingly this was
Karpov's only tournament win over Kasparov;
the other nineteen wins all occurred in their
world championship matches.
Karpov also won his final round after Spassky
blundered a piece. Karpov finished in second
place with l OYz1 1 5 , a point behind Kasparov
and a point and a half ahead of Ehlvest who
was third.
Soon after came the USSR Championship,
in which the two Ks would cross swords
once again. Karpov started well with a win
over Kharitonov, a draw with Beliavsky, and
a win against Smirin. He followed with five
solid draws against Yudasin, Salov, Khalifman,
Kasparov, and Eingorn. Interestingly Kasparov
avoided the Grunfeld this time, in favour of a
Nimzo-Indian.
Karpov then resumed the pattern of winning
with white and drawing with black, scoring
wins over Yusupov, Mikhail Gurevich and
Gavrikov, and drawing with Ivanchuk, Andrei
Sokolov and Smyslov. He drew his next white
game with Vaganian, then met Vladimir
1 98 8
81
Anatoly Karpov - Vladimir Malaniuk
Malaniuk. The Ukrainian grandmaster played
fourteen games against the world champions
from Smyslov all the way to Topalov, with the
exception of Fischer. He did not win any of
these games, but he performed solidly, with
twelve draws and only two losses. He drew
his one previous game with Karpov, and
the two players never met again in regular
tournaments.
I Game II I
Anatoly Karpov Vladimir Malaniuk
-
Moscow 1 988
l.d4 f5 2.g3 ttJf6 3 .ig2 g6
•
Malaniuk is one of the world's greatest
experts on the Leningrad Dutch. For much of
his career he played virtually nothing else. He
scored well with it, and even used it to draw
with Kasparov.
a
b
c
d
e
f
g
h
1 6. lLl b 5 ! !
This combination requires a n exceptional
level of calculation.
1 6 . . . cxb5
1 6 . . . Wib6 1 7.lLlxe5 wins material.
1 7.cxb5 ttJc5 1 8 . .ixe5 Wib6 1 9 . .ixf6 .ixf6
2o.Wid5t ttJ e6 2 1 .Wixd7 Ei:ad8
It looks as though White's queen is trapped.
4.c4 ig7 5.ttJa d6 6.0-0 0-0 7.llJc3
Karpov almost always opted for the main
line against the Leningrad Dutch, and he won
most of his games with it.
7 . Y;Ye8
The following year Yusupov tried a different
system, but Karpov exhibited an impressive
combination of strategic and tactical ideas:
7 . . . c6 8.b3 Wic7 9 . .ia3
Preventing . . . e5.
9 . . . a5 l OJ"kl lLl a6 I 1 .Wid2 .id7 1 2.Ei:fe l
White prepares e4, gaining space and
opening the e-file.
1 2 . . . ttJb4 1 3 . .ib2 e5?!
Finally Black plays his desired pawn break,
but it contains a tactical flaw.
14.a3 ttJa6
14 . . . exd4!? would have left Black positionally
worse, but he would have avoided the
unpleasant game continuation.
1 5 .dxe5! dxe5
.
.
a
b
c
d
e
f
g
h
22.Ei:c6!!
Karpov had to foresee this move all the way
back when he exchanged on e5 on move 1 5 .
In fact he would have needed to calculate
even further and judge the endgame in
certain variations.
22 . . . Ei:xd7
22 . . . bxc6 23 .Wixe6t �g7 24.lLle5 (24.Wixc6
Wixc6 2 5 . bxc6 Ei:d6 26.lLlh4 is also good)
24 . . . Wixb5 2 5 . ttJ d7 Ei:de8 26.Wixc6 Wixc6 (If
26 . . .Wixb3 27.ttJxf8 Ei:xf8 28.a4 White's extra
pawn gives him decent winning chances.)
27 . .ixc6 .ic3
82
The Prime Years
a) The first continued: 1 1 .�d4 e5 1 2.dxe6
ttJe5 1 3 .l"i:ad 1 'lMrxe6 1 4. ttJ a4 'lMrxc4 1 5 .'lMrxc4t
ttJxc4 1 6.�xg7 �xg7 1 7.b3 ttJ b6 1 8 .l"i:xd6
ttJxa4 1 9.bxa4 ttJc5 Black equalized and after a
long fight the game ended in a draw, Karpov ­
M. Gurevich, Reggio Emilia 1 989.
a
b
c
d
e
f
g
h
28.l"k 1 ! (28 .ttJ xf8 �xe 1 29.�xe8 'iii xf8
30.�b5 Black has decent drawing chances,
despite the two pawn deficit.) 28 . . . �b2
29.l"k2 �xa3 30.ttJxf8 l"i:xf8 The game
is not over yet, but White's extra pawn
should provide excellent winning chances,
especially since Black's kingside has been
compromised.
23.l"i:xb6 ttJc5 24.b4 axb4 2 5 . axb4 ttJ e4 26.e3
White converted his advantage in Karpov Yusupov, Linares 1 989.
b) Two years later Karpov deviated: 1 1 .�f4
ttJc5 1 2.'lMrc2 h6 1 3 .h3 e5 1 4.dxe6 ttJe5
1 5 .l"i:ad 1 ttJxe6 1 6.l"i:xd6 ttJxc4 1 7.l"i:d3 ttJxf4
1 8 .gxf4 �e6 1 9 .1"i:fd 1 'lMre7 Black had equalized
again, but this time Karpov was able to grind
out a win, Karpov - M. Gurevich, Amsterdam
1 99 1 .
After the turn of the millennium Karpov
preferred 8.d5 and won a complex struggle
in Karpov - Zhang Zhong, Cap d'Agde (4)
2000.
8 ... lLl a6
Malaniuk had played this move previously,
and later he also tried 8 . . . e5.
9 ..ia3
Preventing . . . e5. Malaniuk had already
faced this move twice, and was successful both
times.
a
b
c
d
e
f
g
h
8.b3
This was the first time Karpov encountered
the present variation, and at the time of the
game the text move was considered critical.
Karpov went on to try 8 .'lMrb3 c6 9.d5 ttJ a6
1 O.�e3 ttJ g4 in two games against Mikhail
Gurevich:
a
b
c
d
e
f
g
h
9 .. c6 10.�d3
.
Karpov follows Dautov's play. Two rounds
1 98 8
83
Anatoly Karpov - Vladimir Malaniuk
earlier Malaniuk defeated Gavrikov, who
preferred 1 O.1"kl .
Malaniuk can do little but wait. If Black's
play grinds to a halt like this in the Leningrad
Dutch, it is almost always bad news for him.
l o ...id7?!
For some reason Malaniuk deviates from his
successful Dautov game, in which he preferred
the main line of 1 O . . . 1"i:b8 preparing . . . b 5 .
White's main replies are I l .tLl d2 and 1 1 .e4, but
Black scores all right against both of them.
Nevertheless 12 . . . tLl c7 was a more useful way
of waiting; the knight has to come to c7 at
some point, and it would have been better
to do it immediately. On the other hand the
king move to the corner may not be essential.
Play might continue: 1 3 .e4 ( 1 3 .h3!?) 1 3 . . . fxe4
1 4.tLlxe4 if5 1 5 .tLlxf6t 1"i:xf6 1 6.iMfe3 ( 1 6.iMfd2
ig4) 1 6 . . . �e6 1 7.iMfd2 1"i:xe l t 1 8 .1"i:xe l iMfd7
Black is still worse but the exchange of a pair
of rooks has eased his position somewhat.
13.e4!
Having performed all the necessary
preparations, Karpov opens the centre.
13 ... fxe4 14. .!Llxe4 if5 1 5 . .!Llxf6
a
b
c
d
e
f
g
h
l U�fel!
Karpov prepares the thematic advance e4.
Karpov exchanges and gets another step
closer to the e7-pawn. 1 5 .tLl h4! ? tLlxe4 1 6.ixe4
ixe4 1 7.iMfxe4 would also have brought White
some advantage.
1 5 ...ixf6
1 1 . .. �d8?!
It is hard to work out the motivation for this
move. Perhaps Malaniuk was anticipating the
witty idea of 1"i:ad l followed by tLle5, although
if that happened Black could always just retreat
his bishop. As it turns out, the rook stays
on this square for the rest of the game while
contributing very little to Black's position.
The plan involving 1 1 . . . b5?! is ineffective here
as the knight on a6 is unprotected, and after
12.cxb5 cxb5 1 3 .e4 White is better.
Perhaps Black should have regrouped with
1 l ...lLlc7 1 2.1"i:ad l b5!?, although White keeps
a modest edge after 1 3 .e4 or 1 3 .tLle5!?
12.�adl c;t>h8?!
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
16JWe3!
This move is more subtle than it looks.
Normally it is said that one should attack
along a file with the rook(s) positioned in front
84
The Prime Years
of the queen, yet Karpov is happy to put his
queen in front. The clever idea behind the text
move is to make it harder for Black to find a
good square for his queen. In some variations
the attack on the a7-pawn might also prove
significant.
16 Wfl
..•
Had the white queen retreated to d2 instead
of e3, the reply 1 6 . . . Wd7 would have been
possible. In the present position this move
would be met by 1 7.h3! .ixh3 l S .ttJe5! when
the advantage of having the queen on the e3square becomes apparent, and after 1 8 . . . Wf5
1 9 .ttJxg6t Wxg6 20 . .ixh3 White dominates
the light squares.
Another possibility was 1 6 . . . ttJ c7 1 7.h3! cj;JgS
l S .g4 ( l S . .ib2 and l S .:ge2 are also possible)
l S . . . .ie6 and now White keeps a stable
advantage after 1 9.:gd2 or 1 9 . .ib2.
17.h3!
Karpov takes away the g4-square from the
bishop and sets up the possibility of g3-g4.
As usual he makes all the small improvements
before taking direct action.
1 9 .94 .icS 20.g5 .ig7 2 1 .Wxe7 bxc4 22.Wxf7
:gxf7 23.bxc4 .ixh3 (23 . . .r;i{gS!? 24.:ge3 .ie6
Black has chances to hold the position a
pawn down.) 24 . .ixh3 :gxf3 25 .:ge3 :gf7
White is only a little better.
1 9 . . . WgS
After 1 9 . . . WeS 20.d5! c5 2 1 ..ib2 Black has a
hard time keeping his position together.
20.cxb5 ttJxb5
After 20 . . . cxb5 2 1 .d5! White takes over on
the queens ide.
21 . .ib2 .id7
White is still better, but the position is more
complex and less one-sided than the game
continuation.
8
7
6
5
4
3
2
1
a
17 c!lJc7 18J�e2!
..•
Defending the f2-pawn in advance.
18 ...i.c8?!
It is not Malaniuk's day; this passive move
does nothing to alleviate his problems.
l S . . . b5!
With this move Black could at least have
created a distraction on the queenside.
1 9 .ttJg5!
1 9.:gc 1 bxc4 20.bxc4 (If 20.:gxc4 ttJd5 the
knight stands well in the centre.) 20 . . . .id7
Black's problems are smaller than in the
game as White must keep an eye on the c4pawn.
b
c
d
e
f
g
h
19.c!lJg5!
Karpov pushes his opponent back even
more.
19 Wg8
•.•
If 1 9 . . . Wg7 20.ttJe4 is strong.
20.Wdl
The queen has done her duty, and now she
vacates the e-file to make way for the rooks.
20 c!lJe6?
•.•
Malaniuk does not wish to suffer slowly, and
tries to alter the course of the game. In a way
he succeeds, as the text move loses by force.
1 98 8
20 ... l'!de8 is uninspiring, but Black is not yet
losing outright: 2 1 . ltJ e4 (2 1 .Wla5 ltJ a6 defends)
2 l . . .ig7 22.l'!de l White is in full control
and has several ways to improve his position,
including softening the black kingside with
h4-h5 , and/or playing in the centre with ib2
followed by ltJ g5 and d5.
21.ltJxe6 Le6 22J:�del .ad7
22 . . . ic8 avoids the game continuation,
but Black is still in big trouble: 23.d5 Wlg7
(23 ... l'!f7 24. dxc6 bxc6 25 .ixc6 ixh3 26.id5
wins) 24.ib2 White wins a pawn, while
his pieces also stand better than their black
counterparts.
Black tries to survive by exchanging a pair
of rooks, but it costs him another pawn and
barely slows down the attack.
29.�xe5 dxe5 30.'lWxe5 �f7 3 1 .d6 .ifS
3 1 . . .l':le8 32.Wlf6t �g8 33 .Wlh8t leads to a
checkmate.
32.c5 h5 33.g4 hxg4 34.hxg4 .id3
If 34 . . .ixg4 3 5 .Wlf6t c;t>e8 36.Wlxg6t wins
the bishop.
a
a
b
c
d
e
85
Anatoly Karpov - Vladimir Malaniuk
f
g
b
c
d
e
f
g
h
35 ..id5t!
1-0
h
23.l'!xe7!
Karpov's strategic plan prevails. The exchange
sacrifice looks nice, although it is not difficult
to appreciate its power.
23 ...ixe7 24J3xe7
White has a pawn for the exchange, and
more importantly, his dark-squared bishop is
poised to commit bloody murder on the long
diagonal.
24 .. J3f6
24 . . d5 25 .id6 wins.
.
25.d5 Wlf8 26J3e3 �g8 27 ..ib2 �fS 28.'lWd4
l'!e5
Black resigned, as it is mate in four more
moves. It was not one ofMalaniuk's better days,
nevertheless Karpov's play was exemplary.
In the last round Karpov and Kasparov both
drew, with Ehlvest and Eingorn respectively.
Thus the two Ks shared first place, with
1 1 Yz/ 1 7.
Karpov's hectic schedule continued with an
eight player, double-round-robin tournament
in Tilburg. He drew his first game against
Hjartarson, defeated Van der Wiel then won a
nice game with black against Timman. Karpov
drew his next four games, then won three in
a row against Hjartarson, Van der Wiel and
Timman. After a quick draw with Hubner
he faced Portisch. Since their 1 982 London
86
The Prime Years
encounter (Game 6S in the first volume)
Karpov had beaten the Hungarian number
one twice, and drawn with him seven times.
grandmaster had considerable experience
with the Benoni structure. A year earlier,
when Karpov reached the same position from
the opposite side of the board, he preferred
9 .. .lZlxd2 and drew with Torre.
Anatoly Karpov - Lajos Portisch
Not long before the present game, Karpov
faced 9 . . . dS. The game continued: 1 O.cxdS
exdS 1 1 ..ie3 liJ a6 1 2.Wa4 cS 1 3 .E!:fd 1 We8
1 4.Wa3 liJxc3 1 S .Wxc3 We7 1 6.Wd2 E!:fe8
1 7.b3 E!:ad8 1 8 .E!:e 1 cxd4 1 9 . .ixd4 liJcS 20.Wb2
.ixd4 2 1 .liJxd4 Wf6 22.E!:ed 1 liJ e6 23.E!:c2 E!:c8
24.E!:cd2 White obtained a stable advantage and
went on to win after his opponent blundered,
Karpov - Van der Wiel, Amsterdam 1 98 8 .
I Game 13 I
Tilburg 1 988
l .d4 lLlf6 2.c4 e6 3.lLlf3 b6 4.g3 .ib7 5 ..ig2
.ie7 6.lLlc3 lLle4 7..idl
Portisch had faced this variation only once
before, drawing with Petros ian in 1 974. Before
that he had used it from the other side, drawing
with Smyslov in 1 97 1 . He never encountered
it again after the present game.
A year later Karpov entered the complicated
8 J'!c 1 .ixd4!? variation and eventually lost a
tough battle against Salov.
A few years later Karpov faced another main
line in 9 . . . d6, and was able to exact some
revenge over Salov: 1 O.dS liJxd2 1 1 .Wxd2 We7
1 2.e4 liJ d7 1 3 . liJ d4 .ixd4 1 4.Wxd4 eS 1 S .Wd2
as 1 6.f4 liJcS 1 7.fS f6 1 8 . .if3 gS 1 9 .b3 .ic8
20 . .id 1 .id7 2 1 .a3 E!:a7 22.b4 liJ b7 23.E!:f2
liJ d8 24.We3 axb4 2S.axb4 liJ f7 26.h4 �h8
27.<j;Jf1 White is somewhat better, and went
on to win a tense and difficult game in Karpov
- Salov, Reykjavik 1 99 1 .
8 ... 0-0 9.l'3c1
10.d5 exd5 1 1 .cxd5 lLlxdl 12.<�xdl d6
7....if6
In the aforementioned Petrosian game the
Hungarian grandmaster preferred 7 . . . cS.
8.0-0
a
b
c
d
e
f
g
h
��
Portisch goes for the main line; the Hungarian
a
b
c
d
e
f
g
h
1��
At the time o f the game, this whole variation
1 98 8
87
Anatoly Karpov - Laj os Portisch
was not new, but nor had it been analysed
extensively. The text move is a decent one
- a white knight rarely stands badly on c4
in Benoni positions - but the main focus
gradually shifted in the direction of 1 3 . ltJ de4!,
a move which was first played by Kasparov
back in 1 980. Five years after the present game,
Karpov used this move to good effect to score a
fine win over Salov: 1 3 . . . j,e7 1 4.f4 ltJ d7 1 5 .g4
a6 1 6.a4 E1e8 1 7.g5 j,fB I B .'j;Jh l b5 1 9 .axb5
axb5 20.ltJxb5 �b6 2 1 .ltJ bc3 �b4 22.�d3
IiJ b6 23 .�g3 <j;JhB 24.E1cd l ltJ c4 25 .b3 ltJ b6
14...hc4 15.VNxc4
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 5 ...a6
This move has been criticized, but it is hard
to see how Black can do without it in the long
run. For example 1 5 . . . �c8 1 6. ltJ e4 j,e7 1 7.f4
and White is better, as after 1 7 . . . ltJ a6 I B . f5 ! ?
looks dangerous.
a
b
c
d
e
f
g
h
26.g6!! fxg6 27.f5! gxf5 2B.E1xf5 ltJd7 29.E1dfl
liJe5 30.E15f4 �b6 3 1 .ltJg5 ltJg6 32.ltJf7t <j;JgB
16.a4 liJd7 17.e3 ltJe5?!
Portisch was a world class player, who
remained the Hungarian number one for
thirty years. However, when I investigated his
games I noticed a subtle weakness: he always
wanted to move forwards. (See also Game 65
in the first volume.) In the present position,
the knight move is too committal and it
turns out that the rest of the black position
is not strong enough to support the advanced
steed.
13 ia6 14.VNb3
A subsequent game demonstrated the
correct way for Black to handle the position:
17 . . . �e7 1 8 .�e2 E1fbB 1 9 .E1al (Or 1 9.b3 b5
20.axb5 j,xc3 2 1 .E1xc3 axb5 and Black has
no problems.) 1 9 . . . E1cB 20.E1fc 1 c4 2 1 . ltJ a2
(If 2 1 .j,h3 E1a7 22.f3 �e5 Black has a good
grip on the position.) 2 1 . . . b5 Black was fine in
Vyzhmanavin - Khalifman, Lvov 1 990.
After 1 4.b3?! b5 Black has counterplay on
the queenside.
1 8.VNe2 c4?!
a
b
c
d
e
f
g
h
33.�xg6! 1-0 Karpov - Salov, Linares 1 993.
This fantastic game serves as a sharp reminder
of Karpov's vicious attacking capabilities.
...
The Prime Years
88
Continuing the faulty strategy. The problem
with this move is that it gives White an easy
lever with which to open the queenside.
The lesser evil was: 1 8 . . . g6! ? 1 9 . 4J e4 (Another
line is: 1 9 .�e4 Wie7 20.�c2 �g7 2 l .f4 4J d7
22.�d3 l"i:fe8 23.l"i:fe 1 Wid8 24.Wg2 Wic8 Black
is passive, but it is not easy to tighten the screw
on his position.) 1 9 . . . �e7 ( 1 9 . . . �g7?! 20.f4
4J g4 2 1 .Wixg4 f5 22.Wie2 fxe4 23.l"i:c4! l"i:e8
[or 23 . . . Wif6 24.l"i:xe4] 24.l"i:xe4 l"i:xe4 25 .�xe4
b5 26.b3 Black has little for the pawn.) 20.b3
White is definitely more comfortable, but a
long fight lies ahead.
8
7
6
5
but his coordination is less than ideal and it
is not easy to improve his position.) 20 . . . Wid7
2 1 .b3 (2 1 .f4 4J g4 22.h3 �xc3 23.bxc3 4J f6
24.Wh2 b5 is unclear) 2 1 . . .cxb3 22.�xb3
l"i:fc8 23.4Je4 �e7 Black may be slightly worse,
but his position looks preferable to the game
continuation.
20 ..ic2!
Karpov prepares to open the queens ide with
b3.
20.f4? would be asking for trouble: 20 ... 4Jd7!
(20 . . . 4J g4? 2 1 .Wixg4 �xc3 22.�xh7t Wxh7
23.l"i:xc3 b5 White has an extra pawn, although
even here it will not be easy to convert his
advantage.) 2 1 .Wixc4 In this position 2 1 . . . 4J c5!
is better for Black, while 2 1 . . .l"i:c8!? is also
interesting, and after 22.Wid3 4J c5 23 .�xh7t
Wh8 24.Wic2 g6 25 .�xg6 fxg6 26.Wixg6 l"i:xe3
the position is unclear (analysis by Karpov) .
4
3
8
2
7
1
6
a
b
c
d
e
f
g
h
5
19 ..ie4
4
This move prevents 4J d3, but it is also part
of a more subtle plan.
3
Another interesting
intending b3.
idea
was
1 9 .l"i:b 1 ! ?
2
1
a
b
c
d
e
f
g
h
19 . . 3 eS
20 .. J3cS?!
Portisch was probably expecting f4, but after
the game continuation the rook move turns
out to be of limited usefulness.
Portisch tries to prepare for the impending
b3, but his rook does not achieve much on
the open file. Black would have been better off
looking for a way to sacrifice a pawn.
J
1 9 . . . g6! ? deserved attention, III order to
secure the future of the bishop on the long
diagonal, for instance: 20.�c2 (After 20.f4
4Jd7 2 1 .Wixc4 l"i:e8 White has an extra pawn,
20 . . . b5!? 2 1 .axb5 axb5 22.4Jxb5 (22. f4 4J g6
23.4Jxb5 Wib6) 22 . . . Wib6 23 . 4J d4 Karpov
evaluates this position as clearly better for
1 98 8
89
Anatoly Karpov - Lajos Portisch
White, but matters are not so dear: 23 . . Jh2
(23. .. g6!?) 24. tLl c6 (24.b3 tLl d3; 24.13:b 1 g6)
24...lLlxc6 2 5 . dxc6 d5 (25 . . . \Wxc6 26.b3 13:b2
27.�xc4) 26.\Wh5 \Wxc6 27.\Wxh7t 'it>f8 Black
has good drawing chances.
20 . . . g6!? 2 1 .b3 (2 1 . tLl e4 is possible, although
after 21 . . . ,ig7 22.b3 f5 Black is active enough.)
2 l . . .lLld7! 22.tLle4 13:xe4! 23.,ixe4 cxb3 Black
has fine play for the exchange and should not
be worse.
White could have improved with 24.f4!
tLl d7 2 5 . bxc4 13:xc4 26.tLld2 13:c5 27.tLl b3
13:c4 28.tLld4 with some advantage.
24 . . . bxc4
If 24 . . . tLl xc4 25 .,id3 \Wa5 26.13:fd l White
has a small edge.
25 .,ia4 13:f8 26.,ib5 f5 27. tLl c3 ,if6
Black obtained an active position and went
on to win in Kuzmin - Timoshenko, Moscow
1 989.
8
21.lLle4 !e7?!
Portisch saves the bishop, but thanks to
Karpov's skilful play this piece will play no
significant role in the game, apart from being
the subject of the eventual coup de grace.
7
A year later two strong players repeated the
game all the way up to the present point, when
Black deviated with:
2 l . . .b5 22.axb5
White can also play without inserting taking
o n b5: 22.b3!? cxb3 (After 22 . . . bxa4 23.bxc4
a3 24.,ia4 13:f8 2 5 . f4 White is better.)
23.,ixb3 13:b8 24.13:b 1 Black is not much
worse, although the b5-pawn is a target.
22 ... axb5 23.b3 ,ie7
Black could consider 23 . . . cxb3 24.,ixb3
�b6! ? 25.tLl xf6t gxf6, although White's
chances are higher here too.
4
6
5
3
2
1
a
b
c
d
e
f
g
h
22.b3!
7
Karpov's previous three moves have all served
to prepare this pawn break. The exchange of
White's b-pawn for Black's c-pawn helps the
first player in three main ways: it liberates
his bishop, opens the lines of fire towards
Black's two remaining queens ide pawns, and
also removes the b2-pawn, which could have
become a long term weakness had it become
fixed on a dark square.
5
22 ... cxb3
8
6
4
3
2
a
24.bxc4?!
b
c
d
e
f
g
h
It was worth considering the pawn sacrifice:
22 . . . g6! ? 23.bxc4 f5 (23 . . . tLlxc4? 24.,id3! is
not what Black wants) 24.tLld2 ,if6 Black
does not have full compensation, but his
position will be tough to break down. The
following continuation may not represent
best play, but it is interesting to analyse all the
same:
90
The Prime Years
view of White's shaky king, I would agree with
that assessment.
24...�xc8 25.Elhl !
Karpov gets closer to Portisch's queens ide
pawns.
a
b
c
d
e
f
g
h
25.e4!? ig7! 26.exf5 ! ? (26.Elfe l is safer,
although after 26 . . .tZJ d7 White's advantage is
not huge.) 26 . . . ih6! 27.f4 tLlxc4 2B .�f2 tLl e3
29 .ib l tLlxfl 30.Elxfl It looks as though White
has fabulous compensation for the exchange,
but Black can keep the position unclear with
30 . . . g5! when any result is possible.
a
23.i.xh3 �d7 24.Elxc8
It was too early for pawn-grabbing: 24.�xa6?!
tLl f3t 25.'it>g2 with two options for Black:
a
b
c
d
e
f
g
h
a) The fancy 25 . . . �g4? does not quite work:
26.ElxcB tLl h4t 27.<;t>gl tLl f3t 2B.<;t>h l �h3
(2B . . . ElxcB 29.id l ! Elc2 30.tLlg5!+- This is
not the only win, but it is certainly the most
attractive one!} 29.ElxeBt ifB 30.ElxfBt <;t>xfB
3 1 .�aBt <;t>e7 32.�b7t <;t>fB 33 .�bBt White
is winning.
b) 25 . . . Elxc l ! 26.Elxc l �g4 Karpov stops here,
concluding that Black has compensation. In
b
c
d
e
f
g
h
25 ...�f5?!
I have already mentioned that Portisch has
a tendency to move forwards, and here we
see another example where a more restrained
approach would have served him better.
A better idea was:
25 . . . g6
This is also a 'forward' move, but it also has
the advantage of giving the black king a
flight square.
26.tLld2
26.<;t>g2!? Karpov likes to make small
king moves like this. 26 . . . �b7 (Also after
26 . . . 'it>g7 27.h3 f5 2B.tLld2 �c3 29 .ic4
White is somewhat better.) 27.ic4 Ela8
White can try to crack Black's position in
several ways, but the defender remains quite
solid.
26 . . . �c3!
The queen is annoying here.
27.tLlc4 tLl d7
After 27 . . . tLlxc4 2B.�xc4 �xc4 29 .ixc4 b5
it is possible that Black can hold the opposite-
coloured bishop position a pawn down, but
he will have to suffer for a long time.
28.ia2 l"i:b8 29.mg2 if8
Black is not much worse.
a
b
c
d
e
91
Anatoly Karpov - Lajos Portisch
1 98 8
f
g
If followed up correctly, this enables Black to
make a real fight of the game.
After 27 . . . tLl f3t?! 2B.mg2 IMrxd5 29.mxf3 f5
30.mg2 fxe4 3 1 .ixa6 Black is in trouble.
Passive defence with 27 . . . l"i:bB?! 2B .l"i:b4 a5
29.l"i:b l g6 leads to a depressing defence for
Black.
27 . . . l"i:c7!? is playable though: 28.l"i:xb6
l"i:c 1 t 29.mg2 tLl f3 30.lMrxf3 l"i:g l t 3 1 .mxgl
IMrxf3 32.l"i:xa6 IMrd l t 33 .ifl IMrxd5 White
is certainly not risking anything with this
queen sacrifice, but it will be hard for him
to win.
2B .lMrxd3
h
26.�d2!
Karpov refuses to give Black any counterplay.
This is not only an objectively strong move, it
also avoids the need to calculate any tricky
and time-consuming variations. Apart from
removing itself from danger and guarding
the f3-square, the knight also moves one step
closer to its ideal home on d4.
26.ic2
Objectively this move might still be good for
White, but it is unnecessarily complicated.
26.J'kB! 27.id3
27.l"i:xb6?! leaves White vulnerable on the first
rank: 27 .. .'IWf3! 2B .lMrxa6 l"i:xc2 29.l"i:bBt ifB
30.l"i:xfBt mxfB 3 1 .IMrxd6t cj;>eB 32.lMrxe5t
cj;Jd7 White has no more than a perpetual.
27.f4 tLlg4 2B .id3 IMrh5 (There is also
28. . .�xd5!? 29.lMrxg4 f5 30.tLlf6t ixf6
3 1 .�xf5 [3 1 .ixf5 l"i:eB] 3 1 . . .lMrxf5 32.ixf5
l'l:bB when Black has decent drawing chances
thanks to the opposite-coloured bishops.)
29.ixa6 l"i:c7 30.tLlg5 l"i:c3 3 1 .h3 1Mrg6 1t will
not be easy for White to play for a win, as his
kingside is dangerously loose.
27 . . . tLlxd3!
a
b
c
d
e
f
g
h
2B . . . g5 !! 29.lMrxa6
The point behind the previous move is
revealed after 29.l"i:xb6?? l"i:c 1 t 30.cj;>g2 g4!
when Black wins.
29 . . . l"i:c5 !
It is not easy for White to maintain control
over the position.
26 J�b8
..
Portisch quite rightly prefers to maintain
the flexibility of his queenside pawns. Instead
26 . . . a5? 27.id l id8 would give White a free
hand to exploit his 'half a pawn' advantage.
27 .idl
.
Now Karpov threatens to take the a6-pawn.
92
The Prime Years
Exchanging queens would not have solved
Black's problems: 27 . . . Wd3?! 2B .Wxd3 (but not
2B.f4? lLlf3t!) 2B . . . lLlxd3 29.lLlc4 b5 30.lLla5
Black is struggling.
White to force it with the slightly weakening
move f2-f4.
a
a
b
c
d
e
f
g
h
28 . .!tJ b3!
The knight is heading for d4.
28 i.f6
•••
Objectively Black's position may not be
drastically worse, nevertheless it is not much
fun when one has to keep choosing which
inferior position to defend.
b
c
d
e
f
g
h
33 .!tJ c6!
•
The knight completes its mission and arrives
at a square from which it can choke Black's
entire position. It is so well placed here, it does
not even feel the need to move again for the
rest of the game.
33 J3a8 34.Wlb4 Wlc7
.•
One other option was 2B . . . Wb7 29.lLld4 l'!cB
(29 . . .Wxd5 30.Wxa6) 30.Wfl g6 3 1 .i.e2 and
White continues to press.
29 .!tJ d4 Wlb7 3o.Wlc2 g6
•
30 . . . Wxd5? loses material after 3 1 .l'!xb6!,
as 3 1 . . .WaB runs into the beautiful 32.We4!,
exploiting the weak back rank.
3 1 .i.e2 c;!;> g7
If 3 1 . . . Wxd5 32.i.xa6 lLl f3t (or 32 . . . Wc5
33 .Wd l with a lasting advantage) 33.lLlxf3
Wxf3 34.Wc7 Black loses a pawn.
32.Wlb3 ttJd7
Rather than glVlng up the c6-square
voluntarily, it may have been better to wait for
a
b
c
d
e
f
g
h
3S. c;!;> g2!
Such small improvements of the king's
position are so typical of Karpov. He achieved
what he wanted on the queens ide and now
prepares to advance on the other flank.
1 988
93
Anatoly Karpov - Lajos Portisch
There was nothing to be gained by forcing
matters on the queenside: 35 . .tg4 lD c5
36.�xb6? Wxb6 37.1'hb6 lDxa4 White has
squandered his advantage.
Karpov continues with his plan.
40 ...�xd7
35 ...h5 36.h3!
Karpov continues his play on the kingside.
36 ...lfig8 37.i.dl ge8 38.g4 hxg4
a
b
c
d
e
f
g
h
41 .�f4!
White must avoid 4 1 .Wxb6?? l"lhS 42.l"lh l
Wf5 when he even loses.
a
b
c
d
e
f
g
h
39.ixg4!?
Karpov probably timed this changing of the
position to arrive just before the time control
in order to present his opponent with fresh
problems. His plan is to exchange the enemy
knight, after which the b6-pawn will be a
sitting duck.
The alternative was also good: 39.hxg4 �g7
40.�f4 l"lhS 4 1 . .te2 White keeps some
pressure.
39 ... lfig7!
Portisch reacts well and immediately looks
for counterplay on the h-file.
39 . . lDc5? is feeble, and after 40.Wxb6 Wxb6
4 Uhb6 lDxa4 42.l"lxa6 White has excellent
winning chances thanks to his extra pawn and
the weakness of the d6-pawn.
.
4o.hd7
It would be premature to exchange queens:
4 1 .Wg4?! Wxg4t 42.hxg4 l"le4! 43.�f3 (If
43 .l"lb4 l"lxb4 44.lDxb4 b5 Black is not worse
in the endgame.) 43 . . .l"lxa4 44.l"lxb6 l"la2 and
Black gets counterplay.
41 ...gh8 42.�g4 �e8
Portisch elects to exchange a pair of pawns.
He could also have considered 42 . . .Wc7,
defending the b6-pawn. 43.e4 (Karpov's
suggestion of 43.f4 is by no means clear after
43 . . . b5! 44.axb5 Wb6 when Black suddenly
becomes very active.) 43 . . . l"lh4 44.We2 Wd7
45 .We3 .tg5 46.Wg3 �h6 Both sides must
be very careful, although Black probably has a
narrower margin for error.
43.gxb6 gh4 44.�f3 gxa4 45.gb8 �d7
In his analysis Karpov prefers exchanging
queens.
45 . . . We4
This reduces White's activity against the
black king; on the other hand the d6-pawn
becomes more vulnerable.
94
The Prime Years
46J:�a8 Ah4
46.Wxe4 E!:xe4
Portisch
wisely
refrains
from
the
overoptimistic idea of trying to improve his
queen.
a
b
c
d
e
f
g
h
47.E!:b7!?
It seems to me that this is White's best
chance.
Karpov gives 47.E!:aS E!:a4 followed by
the awkward 4S.lLlbS?! (4S .E!:a7 is better,
although after 4S . . . E!:a2 49.mg3 E!:a4 White
is unlikely to be able to squeeze a full point.)
but fails to consider the active 4S . . . E!:a2!
49.E!:xa6 E!:b2 5 0 . lLl c6 .ih4 when White has
no winning chances at all.
47 ... E!:a4 4S.E!:d7 E!:a2
There is no saving the d6-pawn, so Black
should concentrate on making his a-pawn
into a nuisance.
49. mg3 a5 50.E!:xd6 a4 5 1 .E!:d7 a3 52.E!:a7 E!:d2
53.E!:xa3 E!:xd5
It is hard to tell whether White can win
this endgame. I would estimate that Black's
drawing chances are higher than White's
winning chances.
46 . . . E!:a2 47.e4
47.E!:a7 WeS is okay for Black.
47 . . . Wb7?
Instead Black should revert to the waiting
plan seen in the game: 47 . . . E!:a4 4S .Wd3
Wb7 49.E!:bS (49.E!:a7 Wb5) 49 . . . Wc7 Black
continues to resist.
4S.E!:a7 Wb2
If Black is looking to activate his queen, then
this seems like the most principled way of
doing it.
4S . . . Wb6 loses material after 49.E!:d7 (but
not 49.e5? dxe5when White has nothing) .
a
b
c
d
e
f
g
h
49.e5! dxe5 50.lLldS!! .ixdS 5 1 .Wxf7t mh6
52.WfSt!
This precise check leads to mate in a
maximum of eight moves. The main line
runs as follows:
52 . . . mg5 53.h4t mxh4 54.Wh6t mg4
5 5 .Wxg6t .ig5 56.We4t mh5 57.E!:h7t .ih6
5 S .Wf5 t mh4 59.E!:xh6#
47.e4 .if6
Black could have considered sacrificing the
a6-pawn: 47 . . . E!:a2!? 4S .Wc3t .if6 49.Wb3
E!:b2 50.Wa3 (50.Wc4 .ih4) 50 . . . E!:e2 5 1 .Wd3
E!:el 52.E!:xa6 (52.f4 E!:e l ) 52 . . . Wb7 It will be
a
b
c
d
e
f
g
h
1 98 8
hard for White to make his extra pawn count,
as Black's pieces are active and the c6-knight is
out of play.
48.Wd3
a
50.We2!
�b7 49J�� b 8 �d7
b
c
d
e
95
Anatoly Karpov - Lajos Portisch
f
g
After 5 1 . . . l"i:xb3 52.1Mfxb3 Karpov evaluates the
position as clearly better for White, due to the
impending 1Mf a3 when the a-pawn comes under
fire. However, Black can create problems with
52 . . . 1Mfe8! 53 .1Mfc4 \j;>g8 54.f3 (or 54.<;f;>f3 1Mfd7)
54 . . . 1Mff8! 5 5 .1Mfxa6 1Mfh6 when he should be
active enough to hold the balance.
h
Karpov notices that the black rook is short of
squares, and sets out to exchange it.
50 ... 1:!a3
If 50 . . l"i:al 5 1 .l"i:b 1 White can still exchange
the roo ks .
.
a
b
c
d
e
f
g
h
52.1:!f3!
Suddenly it transpires that the rook on al
is trapped, and Black has no defence against
l"i:xf6.
52 ... �b7
Portisch tries a desperate counterattack.
53.�xf6 �b5 54.�c3!
Karpov has seen that he can easily evade the
checks.
54...�fl t 55.i>g3 �gIt 56.i>h4
1-0
a
b
c
d
e
f
5 1 .1:!b3! 1:!al??
Portisch commits a fatal error.
g
h
Black had no choice but to exchange rooks.
The way Karpov applied pressure in the
middlegame was highly instructive and
powerful, notwithstanding the fact that Black
could still have held the position after the
knight transfer to c6.
In the penultimate round Karpov defeated
Predrag Nikolic, then finished with a quick
draw against Short. His final tally of l OYz1 1 4,
with seven wins and seven draws, was enough
96
The Prime Years
to win the tournament by a remarkable two point margin ahead of Short.
Karpov's final event of the year was the Thessaloniki Olympiad, where he represented the
Soviet Union on the second board. He started with a somewhat surprising draw with the white
pieces against Mascarinas of the Philippines, who played well. But then he found his form and
dispatched Panno, Speelman and Andersson in succession. Karpov followed with a long draw
against Nikolic, a win over Gulko, draws with Ribli and Van der Wiel, and two final wins over
Lars Bo Hansen and Ftacnik.
Karpov's final tally was a superb 8/ 1 0, with six wins and four draws. He won the individual
gold medal for the best performance on board two, and also helped the USSR to claim the gold
medal in the main competition.
According to the database Karpov played two games in Riga in 1 988, but I was not able to locate
the full details of the tournament or the time of year when it took place. He won one of these
games and drew the other. Karpov also took part in the first World Rapid Chess Championship
in Mazatlan, Mexico. He tied with Gavrikov, then they drew the playoff but Karpov took the title
on a tie break. Sadly, no rapidplay world championship has taken place since then.
In 1 988 Karpov's results were highly impressive, and left no doubt that he was still head and
shoulders above every chess player on the planet not called Garry Kasparov. Nevertheless
Kasparov's results were even more superlative, which suggested that the gap between them was
still widening.
1988 Summary
Wijk aan Zee ( l st place) : 9/ 1 3 (+6 =6 - 1 )
Euwe Memorial, Amsterdam (2nd-3rd place) : 3 Yzl6 (+2 = 3 - 1 )
World Cup, Brussels ( l st place) : 1 1 1 1 6 (+7 =8 - 1 )
Amsterdam (2nd place) : 6Y2/ 1 2 (+3 = 7 -2)
World Cup, Belfort (2nd place) : 1 0Y21 1 5 (+7 =7 - 1 )
USSR Championship ( l st-2nd place) : 1 1 Yz/ 1 7 (+6 = 1 1 -0)
Tilburg ( l st place) : l OYz/ 1 4 (+7 =7 -0)
Thessaloniki Olympiad (Board two gold medal) : 8/ 1 0 (+6 =4 -0)
Total 68.6% (+45 =54 -6)
Ii Wins • Draws • Losses
1989
Rating 2750 (2 in the world)
The next world championship cycle was underway, and this time Karpov was seeded directly
into the quarter-finals. The match took place in Seattle, and Karpov's opponent was Johann
Hjartarson. The Icelandic grandmaster played all the world champions from Smyslov to Topalov,
with the exceptions of Petros ian and Anand. He won two of those encounters, drew nineteen and
lost eighteen. He found Karpov a difficult opponent, and scored a mere six draws and seven losses
against the former champion, without a victory.
Karpov drew the first game of the match with the black pieces, then won the second in style.
1 Game 14 1
Anatoly Karpov - Johann Hjartarson
Seattle (2) 1 989
I .c4
Hjartarson knew the black side of the Queen's Gambit as well as the Nimzo- and Queen's
Indian Defences, so Karpov probably wanted to reduce his options.
l. ..e5
Hjartarson immediately prevents a transposition to a l .d4 opening.
2.lLlc3 tLlf6 3.g3
Karpov played this move only three times in his career, and the first two occasions were way
back in 1 972.
3... d5
Hjartarson is a l .e4 player so he is familiar with Sicilian positions, yet he had never before
played this particular continuation, so overall his decision was rather risky.
4.cxd5 tLlxd5 5.i.g2 tLlb6 6.tLlf3 tLlc6 7.0-0 i.e7 8.a3 i.e6 9.b4 0-0 10J!b l f6 1 l .d3 �d7
This is something of a sideline, and that was especially true in 1 989 when the move had hardly
ever been seen. The main lines are 1 1 . . .tLl d4 and 1 1 . . .a5.
12.lLle4 tLld5
1 2 . a5? 1 3 . tLl c5 is strong.
.
.
The Prime Years
98
14 . . . CiJ a7 1 5 .d4 CiJxb5 1 6.dxe5 f5 1 7. CiJ fg5
c6 1 8 .CiJxe6 Wxe6 1 9.a4 leads to complex
play, although White's position remains
slightly preferable.
1 5 .CiJxd4 exd4 1 6.�b2 b6
1 6 . . . !'Iad8 1 7.�xd4 is good for White.
The game Gacso - Kerekes, corr. 1 990,
reached this position via a different move
order.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
13.'lWc2
Karpov indirectly prevents the exchange of
the g2-bishop.
Karpov's play must have made an impression on
Hjartarson, as he went on to play this line with
the white pieces, and two years later he reached
the same position. At this point he deviated
from Karpov's play: 1 3 .�b2 !'Iad8 1 4.Wc2 �h3
1 5 .tlk5 �xc5 1 6.Wxc5 �xg2 1 7.<;t>xg2 <;t>h8
1 8 .h3 We8 ?! 1 9.b5 CiJ ce7 20.Wxa7 Wxb5?
(20 ... b6) 2 1 .�a l Wd7 22.Wxb7 White was a
pawn up and went on to win in Hjartarson all, Budapest 1 99 1 .
13 ... b6?!
Weakening the c6-square is not a good idea.
13 . . . �h3? ? is refuted by 1 4.�xh3 Wxh3
1 5 .Wc4 We6 1 6. CiJ c3 !'Ifd8 1 7.e4 and White
wins a piece.
1 3 . . . a5! ?
This untested move leads t o a complex
fight.
The more conservative 1 3 . . . a6! ? has become
the main line, although it does not really
challenge the white position and the first
player should maintain a slight plus.
1 4.b5 CiJ d4
1 4 . . . CiJ d8 1 5 .d4 is strong.
a
b
c
d
e
f
g
h
1 7. CiJ d2!?
The game continuation of 1 7.a4 CiJ b4
1 8 .Wd l �d5 was okay for Black.
1 7.Wa4!? f5 ! 1 8 .Wxd4 !'IV 1 9 . CiJ c3 �f6
20.Wa4 �xc3 2 1 .�xd5 �xd5 22.�xc3 f4 is
unclear.
1 7 . . . Wxb5 1 8 .�xd4 Wd7 1 9 .�b2
The position is rather complicated, but
White should have slightly better chances
thanks to his central pawn majority.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 989
99
Anatoly Karpov - Johann Hjartarson
14.ib2 gac8?
Black cannot spare the time for this move,
and he should have acted at once.
1 4 ... aS !
This was the only way for Black to make
sense of his position.
1 5 .bxaS
If 1 5 J�fc 1 axb4! 1 6.Wl'xc6 bxa3 1 7.Wl'xd7
ixd7 1 8 .ic3 c6 Black has decent
compensation for the piece.
1 5 .bS lLl a7 1 6.d4 lLlxbS 1 7. dxeS lLlxa3
( l 7 . . . ixa3 1 8 .l"lfd l ) 1 8 .i.xa3 i.xa3 1 9 .1"lfd l
Most human players would prefer White's
position, but it is rather double-edged.
1 5 . . . :1haS 1 6J''1 fc 1 lLl a7
16 . . . lLld8 1 7.d4 favours White.
1 7.d4 tUbS 1 8 . dxeS lLl xa3 1 9 .ixa3 l"lxa3
20.l"ld l
opening well below his usual level; according
to the database he had never played it before
this game, and he may also have been nervous
at the start of his first candidates match.
l S . . . lLl d8 would have given a few more
chances, but after 1 6.d4! exd4 (or 1 6 . . . lLl f7
1 7.dxeS fxeS 1 8 .l"lfd l ) 1 7.tUxd4 i. f7 1 8 .l"lfd l
Black is in trouble.
16.i.xd4!
Karpov starts by taking with the bishop,
in order to make the second capture with his
knight.
Nevertheless 1 6.lLlxd4 was also strong:
1 6 . . . exd4 1 7.Wl'c6! l"lcd8 ( l 7 . . .l"lfd8 1 8 . lLl d2)
1 8 . tU d2! White goes after the d4-pawn.
1 8 . . . tU c3 1 9.i.xc3 Wl'xc6 2o.ixc6 dxc3 2 1 .l"lxc3
Black has nothing for the pawn.
White's superiority in the centre should
count for more than Black's queenside pawns.
16 ... exd4
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
17.�c6!
Karpov fixes the c7-pawn and exchanges the
queens, thus sealing the fate of the d4-pawn.
15 ... tlJd4?!
This virtually amounts to a pawn sacrifice,
After the hasty 1 7.lLlxd4? lLlxb4 1 8 .axb4
( l 8 .Wl'b2 lLldS) 1 8 . . .Wl'xd4 1 9 .bS White is only
slightly better according to Karpov.
but Black will not obtain any compensation
for it. Overall Hjartarson has played the
17 �xc6!
.••
100
The Prime Years
If 1 7 . . . ltJ c3 ? 1 8 .ltJxd4 wins, while after
1 7 .. .1"!:fd8 1 8.ltJxd4 White wins a pawn while
maintaining a better position.
is smaller than in the game. The knight has
better prospects on d2 than a6.
8
18Jhc6 i.d7
7
6
5
4
3
2
1
a
a
b
c
d
e
f
g
b
c
d
e
f
g
h
22 ... lLIf6 23.lLIxa7 i.d6
h
19.ttJxd4!
This exchange sacrifice seals the outcome of
the game. White will obtain two pawns and
the black pieces will remain painfully passive.
19 ...i.xc6 20.ttJxc6 13ce8
Black has to defend the bishop.
2 1 .13c1!
Karpov sees that the second pawn will not
run away, so he develops calmly.
21 ...f5
Black vacates a square for his knight. Instead
2 1 . . .i.d8 22. ltJ d2 ltJ e7 23.ltJxa7 is winning for
White.
Sacrificing a piece with 2 1 . . .a5!? does not
break White's grip: 22.ltJ d2 ltJxb4 23.axb4
i.xb4 24.i.d5t Wh8 2 5 . ltJ e4 f5 26.ltJg5 1"!:xe2
27.ltJxb4 axb4 28.1"!:xc7 White wins.
22.ttJd2!
The flashy 22.ltJc5?! is not so effective, and
after 22 . . . ltJ f6 23 . ltJ a6 i.d6 White's advantage
If 23 . . . c5 24. ltJ c6 g6 25 .e3 cj;>g7 26.b5 ltJ d7
27.ltJ c4 i.d8 28.a4 i.c7 29.1"!:al White is in
full control and will soon break through on
the queenside.
24.e3 cS 2S.lLIc4
This is not the only way to convert the
advantage. A good alternative was 2 5 . ltJ b5
1"!:d8 26.ltJ c4 i.b8 27.d4 cxd4 28.ltJxd4 1"!:c8
29.i.c6 and White wins another pawn.
2S ....tb8
8
7
6
5
4
3
2
1
a
26.lLIc6
b
c
d
e
f
g
h
1 989
101
Anatoly Karpov - Johann Hjartarson
Material i s roughly equal, but the position is
completely winning for White as his pieces are
so dominant. Black's rooks can only dream of
finding an open file.
The extravagant 26.bxc5 !? was also good
enough, as after 26 . . . i.xa7 27.cxb6 i.b8 28.a4
8
7
Karpov gives up his great knight in order to
invade with his other pieces. Other options
were equally effective, for instance 30.'.t>n!? f4
3 1 . l!?e2 or 30.i.f3!? f4 3 1 .gxf4 gxf4 32.e4.
30 .. J�xb8
30 . . . ttJxb8 3 1 .Ei:c5
33.i.b7 wins.
ttJ a6
32.Ei:xb5
Ei:b8
3 1 .Ei:c7
The rook takes up a dominating position on
the seventh rank.
6
5
4
3 1 ...tiJf6
3 1 . . . ttJ b6 32.ttJc6 Ei:be8 33.Ei:b7
34.Ei:xb5 picks up another pawn.
3
2
a
b
c
d
e
f
g
h
White's four pawns are stronger than Black's
rook. Of course there was no need for Karpov
to play like this in such an important game,
when simple play will suffice.
ttJ a4
8
7
6
5
4
26 b5 27.tiJ4a5 c:x:b4 28.axb4 tiJd7
...
3
2
1
a
b
c
d
e
f
g
h
32.tiJc6! l3b6 33.tiJe7t c;!;>h8 34.tl)xf5
White not only picks up a third pawn, but
also opens the floodgates for the advance of his
central pawns.
a
b
c
d
e
f
g
h
29 d4!
.
White was already dominating, and now
he takes away the e5-square from the black
pieces.
29 g5 30.tiJxb8
...
34... l3a6
Finally one of the black rooks gets to an open
file, but it is too little too late.
35.l3c1
Karpov gives no counterplay at all.
35 ... l3a2
If 35 . . . Ei:a4 the simplest route to victory is:
36.Ei:b l Ei:a2 37.e4 White simply pushes his
The Prime Years
102
pawns. 37 . . . tLl g4 (37 . . . tLl d5 38.1"i:b3) 38.f3
White wins.
36.h3 �b2 37.e4 �xb4
a
b
c
d
e
f
g
h
38.g4!
Karpov stabilizes the knight, and is now
ready to push the e-pawn.
38 ... h5 39.e5 hxg4
39 . . . tLlh7 40.e6 wins.
40.exf6 gxh3 41 .Lh3
If 4 1 ..ie4 1"i:xf6 42.1"i:c8t <;t>h7 43.tLle3t <;t>g7
44.1"i:c7t <;t>f8 45.tLld5 wins.
41 ...�xf6 42.�c8t <.fIh7 43.�c7t <.fIg6
43 . . . <;t>h8 was also hopeless: 44.1"i:b7 1"i:b3
45 ..ig4 1"i:b2 46.<;t>g2 b4 47.d5 1"i:d2 48.d6+-
44.�g7t <.fIh5 45.8!
1-0
Karpov finished the game in characteristic
style, weaving a mating net with a few pieces
in the endgame.
In Game 3 Karpov won again, after
Hjartarson erred in a Zaitsev Ruy Lopez.
After taking a two point lead at the halfWay
point of the six-game match, Karpov took no
further chances and played solidly in the next
two games, drawing both with ease and thus
securing victory with a game to spare.
After his convincing match win, Karpov took
part in the prestigious Linares tournament. He
suffered a setback in the first round against
Short, but after drawing with Gulko, he
bounced back with wins over Beliavsky, who
blundered in an equal endgame, and Yusupov,
in a fine game that can be found in the notes to
Game 1 2 (Karpov - Malaniuk) . Next Karpov
drew with Andrei Sokolov, and in Round 6 he
encountered Ljubomir Ljubojevic.
Since their meeting in 1 977 (Game 47 of the
previous volume) , Karpov and Ljubojevic had
met twenty eight times over the board. Karpov
had won ten of those encounters, drew fifteen
and lost three.
I Game l� I
Anatoly Karpov - Ljubomir Ljubojevic
Linares 1 989
l .d4 .!iJf6 2.c4 e6 3 . .!iJc3 d5
This was the first time Ljubojevic played
this opening against Karpov. Curiously, in the
games where Karpov had white, they hardly
ever repeated the same opening as one of them
would almost always deviate.
a
b
c
d
e
f
g
h
4.cxd5 exd5 5 ..ig5 c6 6.e3 .!iJ bd7
Vaganian survived Karpov's squeezing in the
6 . . ..if5 7.1Wf3 .ie6 8 . .ixf6 1Wxf6 9.1Wxf6 gxf6
1 989
Anatoly Karpov - Ljubomir Ljubojevic
variation at the 1 988 USSR championship,
but no-one else was brave enough to play this
position against Karpov.
103
The main line continues 1 4 . . . liJ g6 1 5 .b5
axb5 1 6.axb5 ig4 1 7.ixe4 dxe4 1 8. liJ d2
when White is somewhat better and risks very
little.
7.id3 ie7
Sometimes Ljubojevic preferred a set-up
with the bishop to d6, with mixed results. On
this occasion he opts for the main line. Overall
his opening selection does not seem ideal, as
Karpov has obtained a position that suits his
style, with a stable pawn structure and clear
strategic contours.
8.�c2 0-0 9.ttJO ge8 10.0-0 ttJ f8 1 1 .gab l
Karpov deviates from his usual I I .h3 line,
and instead starts the minority attack at once.
He probably anticipated that Ljubojevic had
prepared something. The long-time Yugoslav
number one was a world class player, but
openings were never his strong point, so it
is logical that Karpov would try to set him
unexpected problems.
1 1 ...ttJe4
Black elects to simplifY; it is one of the main
lines. Many players like to insert the moves
1 1 . . .a5 1 2.a3 before deciding what to do next.
Perhaps the most testing line is 1 1 . . . id6!?
1 2.b4 liJ g6 1 3.b5 h6.
1 2.ixe7 Wfxe7 13.h4
Karpov follows the well-known plan of
staging a minority attack, to create a weakness
in Black's queens ide structure.
1 3 ... a6
Exchanging the a-pawn has a good and a bad
side. Black exchanges off a potential weakness,
on the other hand White has another open file
with which to organize his queenside play. Less
obviously, Black can also encounter problems
on the seventh rank and even the back rank.
14.a4 if5
a
b
c
d
e
f
g
h
I S.ttJeS!?
Karpov introduces a novelty rather than play
the usuaI 1 5J'�fc l .
It is worth mentioning that the immediate
implementation of the minority attack with
1 5 .b5 offers White very little, as Black gets
counterplay on the a-file: 1 5 . . . axb5 1 6.axb5
liJ xc3 1 7.'lWxc3 �a3 1 8.�b3 �xb3 1 9.'lWxb3
ixd3 20.'lWxd3 c5 2 1 .dxc5 'lWxc5 22.�d l �d8
White's advantage is symbolic, as Black can
easily live with his solitary weakness on d5.
I S ... gad8
Ljubojevic faces an unpleasant choice
between several continuations which all lead
to slightly worse positions. In addition he
is more than aware of Karpov's exceptional
ability in squeezing small advantages. His last
move contains a couple of ideas. In certain
positions Black may be able to meet b5 with
. . . axb5 followed by . . . c5 when the rook will be
useful on the d-file. Another idea would be to
leave the pawn on c6 and swing the rook to
g6 or h6, where it partakes in both attack and
defence.
1 04
The Prime Years
1 5 . . .tZJg6 does not equalize: 1 6.ltJxg6 (There
is also 1 6 . .ixe4 .ixe4 1 7.ltJxe4 ltJxe5 I B. ltJ d2
ltJ g4 1 9.1tJf3 when White maintains some
pressure.) 1 6 . . .hxg6 1 7 ..ixe4 .ixe4 I B.ltJxe4
lMfxe4 1 9.1Mfxe4 E!:xe4 20.E!:fe l Intending b5,
with pressure in the endgame.
I B . . . exd4 1 9.ltJf5
Karpov evaluates this posltlon as clearly
better for White, but this seems much too
optimistic.
1 9 . . . lMfe4 20.lMfxe4 E!:xe4 2 1 .ltJxd4 ltJ e6
Black is safe.
1 5 . . . ltJ d7 1 6 . .ixe4 dxe4 1 7.ltJxd7 .ixd7 I B.b5
cxb5 1 9.axb5 a5 20.ltJd5 lMfg5 2 1 . ltJ b6 .ih3
22.f4 exf3 23.E!:xf3 .ig4 24.E!:g3 The pin is
unpleasant for Black.
1 5 . . . f6!?
This has been tried in a couple of subsequent
games, and looks like Black's best chance to
equalize.
1 6 . .ixe4
Simplifying with 1 6.ltJxe4 gives little for
White: 1 6 . . . fxe5 1 7. ltJ c5 .ixd3 I B.lMfxd3
exd4 1 9.1Mfxd4 ltJ e6 Black equalizes.
1 6. ltJ f3!? soon led to a draw in Rendboe
- Bank Friis, Bellinge 1 99 1 . It is doubtful
that it is worth investing two tempos just to
provoke the slight weakening of . . . £7-f6.
1 6 . . . .ixe4
1 6 . . . dxe4 1 7. ltJ c4 is good for White.
1 7. ltJ xe4 fxe5!
1 7 ... dxe4 I B. ltJ c4 maintains the pressure.
a
b
c
d
e
f
g
h
16JUc1!
Karpov plays a smart move and develops his
final piece before playing b5.
1 6.ltJxe4 .ixe4 17 . .ixe4 dxe4 I B.E!:fe l ! (After
the premature I B.b5?! axb5 1 9.axb5 c5! Black
solves his problems.) It looks as though White
has a similar advantage as in the game, but
there is a difference. I B . . . E!:d5! We can see
another point behind Ljubojevic's 1 5th move.
1 9.b5 axb5 20.axb5 c5 2 1 .lMfxe4 lMfe6 22.lMff4
cxd4 23.exd4 lMfe7 24. ltJ f3 ltJ e6 25.lMfe3 lMfd6
Black's pressure against the d4-pawn gives him
equal chances.
16 li:) g6�!
...
a
b
c
d
e
f
g
h
I B.ltJg3!?
After I B.ltJc5 exd4 1 9.exd4 ltJ e6 the target on
d4 gave Black enough play in De Lagontrie
- Ackerley, corr. 1 992.
Black exchanges all the minor pieces, but
underestimates the power of White's queenside
initiative in the major piece ending.
Chasing back the knight with 1 6 . . . f6 was
stronger: 1 7. ltJ f3 ( 1 7 . .ixe4 .ixe4 I B.ltJxe4 fxe5
1 9.1tJc5 exd4 20.exd4 ltJ e6 The weakness of the
d4-pawn gives Black enough play.) 1 7 . . . ltJxc3
1 989
105
Anatoly Karpov - Ljubomir Ljubojevic
1 8.�xc3 i.xd3 1 9.Wxd3 lLl e6 I t will not be
easy for White to make progress, as a future
b5 will be met by . . . axb5 and . . . c5, when Black
will have excellent chances to hold.
17.he4 he4 1 8.lbxe4 dxe4?!
Black continues the faulty plan of simplifying.
Karpov mentioned the line 1 8 . . . lLlxe5 1 9.1Lld2
liJg4 (or 1 9 . . . lLl g6 20.b5 with strong pressure.)
20.liJf3! White makes sure ... lLlxf2 ideas will
not work, and will soon carry out the b5
advance. Black is worse, yet still he has more
chances for counterplay than in the game.
19.�xg6
With the knights removed, Black's chances
of launching a successful kingside attack are
almost nonexistent.
a) 2 1 . . .cxb5?! This is the least attractive path.
Not only is b7 the hardest weakness to defend,
but Black must also reckon with the white rook
switching to e5 to win the e4-pawn. 22.l'!xb5
Wd7 (22 . . . l'!d7 23.l'!e5; 22 . . . l'!b8 23.l'!cb l )
23.Wb l l'!c8 24.l'!d l l'!c7 25 .d5 l'!e5 26.h3
Black is struggling.
b) Keeping the weakness on c6 with 2 1 . . .l'!d6
would have led to a passive defence as well:
22.Wa4 l'!c8 23.h3 Wd7 24.l'!c5 Black faces an
unpleasant defence.
Before we go any further, let me show you
another wonderful example in which Karpov
wears down his opponent using a minority
attack.
19 .. hxg6
.
8
7
6
5
4
a
3
2
1
a
b
c
d
e
f
g
h
20.b5!
Karpov carries out the thematic minority
attack, and ensures that at least one pawn
weakness will appear in Black's camp.
20 ... cxb5
Black has to take on b5 one way or another,
as Black would find it impossible to hold a
position with weaknesses on a6 and c6. After
the alternative 20 . . . axb5 2 1 .axb5 Black has
two options:
b
c
d
e
f
g
h
26.lLl b3!
Karpov improves the knight. Later it may
move to c5, but it has another purpose as
well.
26 . . . i.c8?!
Black follows the principle stating that
pawns on their starting squares are not easy
to attack.
He should have preferred 26 . . . b6! 27.lLld2
l'!eb8 28.l'!bc 1 lLl e7 with a playable
position.
27.a5!
Karpov fixes the b7-pawn.
27 . . . lLl e7 28.lLl g3!?
In the minority attack White must pay
attention to the c4-square. With his last move
The Prime Years
1 06
Karpov prevents . . . if5 , as the exchange of
light-squared bishops would have enabled a
black knight to occupy the outpost on c4.
28 . . . g6
Black decides to chase the knight. Like in so
many ofKarpov's games, when his opponents
try to create counterplay on the kingside,
Karpov gradually shifts his attention to that
side of the board and eventually exploits the
weaknesses created there.
Maybe Black should have preferred
28 . . . lilef5 , continuing to play for a bishop
exchange. After 29.lilxf5 ixf5 3o.ixf5
lilxf5 3 1 .lilc5 lild6 32.'lWb6 :B:ec8 Black has
reasonable chances to hold the position.
but still the minor piece exchanges allow
him to get closer to Black's queens ide
weaknesses.
34 . . . ie6
Kharitonov wants to manoeuvre his knight
to d6, which is usually a good square for
it in this pawn structure. The immediate
34 . . . lilc8? loses to 35 .'lWxd5, so he defends
the d-pawn first.
a
a
b
c
d
e
f
g
h
29.:B:cc l !
I t i s not easy to spot the point o f this rook
move. Karpov wants to put his knight on
c5, but to do it he needs a rook on al to
defend the a5-pawn. With the text move he
ensures that the rook will be defended, thus
preventing . . . b6.
29 . . . h5
Black continues his plan. 29 ... lilef5 was still
reasonable.
30.:B:al h4 3 1 .lilfI !
Karpov keeps the e2-square vacant.
3 1 . . .if5 32.ie2!
Karpov keeps the bishop in order to control
the c4-square.
32 . . . lile4 33.lilc5 lilxc5 34.'lWxc5
Karpov's strong knight has been exchanged,
b
c
d
e
f
g
h
35.lild2!
Karpov spots that the h4-pawn can be
attacked.
3 5 . . . lilf5 36.lilf3!
Karpov plays carefully.
36.id3 would allow Black to stir up
complications: 36 . . . lilxe3!? 37.fxe3 ixh3!?
38.lilfI ixg2 39.<j;Jxg2 :B:xe3 40.'lWc2! Black
does not have quite enough compensation,
but White still has to be careful.
36 . . . 'lWd8
36 . . . lild6 37.lilxh4 is a safe extra pawn for
White.
a
b
c
d
e
f
g
h
1 989
37.a6!
Once again Karpov's strategy reigns supreme,
and Black's pawn structure collapses.
37 . . . bxa6 38.l"lxa6 l"lxa6 39.�xa6 1M'aB
After 39 . . . �d7 40.tLl eS tiJ e7 White can
choose between 4 1 .l"lb 1 , invading to the
seventh rank, and 4 1 .�b7 winning the c6pawn. In both cases Black is helpless.
40.�xc6
A one-pawn advantage is usually enough to
decide the game by itself, and here Black also
suffers from weaknesses on dS and h4.
40 ... �xc6 4 1 .l"lxc6 l"laB 42.�d3
1 -0 Karpov - Kharitonov, Moscow 1 9BB.
Black resigned in view of 42 . . . l"la l t 43.Wh2
l"la2 44.l"lc2 when the h4-pawn falls as well. It
was another great strategic performance from
Karpov.
Back to the game.
a
b
c
d
e
f
g
This excellent move creates the threat of
l"lc7, while also setting up the idea of l"lcS after
which the rook can go to as or eS according to
circumstances.
23 ...�d7?
Ljubojevic commits a common mistake,
thinking that a queen exchange in the endgame
will help him to get closer to a draw. The
idea has a major drawback here, as it allows
the white king to play an active role in the
game.
23 . . . l"laB?!
This move was also unsatisfactory.
24.l"lcS! 1M'h4
Karpov stops here saying that Black has
counterplay, but further analysis reveals that
it is insufficient.
2S.l"lb7!?
White has a subtle way to go after Black's
king.
2S .l"leS l"lf6 26.1M'c2 probably wins as well.
2S ... l"lf6 26.1M'c2 as
h
2 1 .axb5l"1d6
Ljubojevic decides to have his soft point on
a6. Perhaps his reasoning was that it would be
harder for the white queen to attack the a6and e4-squares at the same time.
2 1 . . .aS? would have been a mistake, as after
22.b6! Black must worry about three vulnerable
pawns instead of two.
22.bxa6 bxa6 23.�a4!
1 07
Anatoly Karpov - Ljubomir Ljubojevic
a
b
c
d
e
f
g
h
27.l"leS!
The immediate 27.l"lcBt?! is premature, as
after 27 . . . l"lxcB 2B.1M'xcBt cj;lh7 29.g3 1M'hS!
White's king is exposed to checks.
27 . . . a4 2B.l"lc7 a3 29.l"lcBt! l"lxcB 30.1M'xcBt
Wh7 3 1 .g3! a2
3 1 . . . h6? 32.l"leB wins immediately.
32.l"laS
The Prime Years
1 08
White wins the a-pawn, and should not have
too much trouble converting his advantage.
23 . . . 'We6!
This was Black's chance. The queen move
prepares . . . l"k8, and enables Black to offer
serious resistance.
8
7
6
5
4
3
2
1
��������=-�
a
b
c
d
e
f
g
h
24.'Wxd7
Swapping queens will enable Karpov to
activate his king.
a
b
c
d
e
f
g
h
24.l"k5
If 24.l"k7 :gc8 25 .:ga7 :gcc6 Black remains
fairly solid.
24.h3 :gcB 25.:gxcBt 'WxcB 26.'Wb4 'Wc7
27.'WbBt 'WxbB 2B.:gxbBt <j:rh7 Black has
reasonable chances to hold the rook ending.
White can insist on a queen trade with
24.'Wc4, but after 24 . . . :gaB 25 .'Wxe6 :gxe6
26.:gc4 the black rook is much better
on e6 than on a7 where it ends up in the
game.
24 ... :gcB 25 .:ge5 :gdc6!
Black is living dangerously, but he may
have just enough counterplay to stay in the
game.
26.h3
After 26.:gxe6 :gel t 27.'Wd l :gxd l t 2B.:gxd l
fxe6 29.:gal a5 Black has good drawing
chances as White's rook is passive.
There is also 26.:gn :gc4 27.:gxe6 :gxa4
2B.:gxe4 :ga2 when Black is a pawn down,
but his doubled rooks on the second rank
will tie White up considerably.
26 . . . 'Wd6 27.'Wb4 'Wf6 2B.:gn :gel
Black is still alive.
24 .. J:hd7 25.�c5!
White blocks the a-pawn as early as possible.
On the fifth rank the rook can also threaten to
snatch the e4-pawn.
25 ... �a7
Black would prefer to defend his a-pawn
from the sixth rank, but after 25 . . . :ge6?
26.:gbBt <;t>h7 27.h4! he has to give up mate­
rial to avoid being mated on the back rank.
Another possible line is 25 . . . :gd6 26.:gb7 :ga8
27.:ga5 :gcB 2B.h4 :gf6 29.:ga2 :gaB 30.:ge7
:ge6 (30 . . . a5 3 1 .:gxe4) 3 1 .:gxe6 fxe6 32.:ga5
when White wins as he has blocked the a-pawn
early.
26.�a5 c;t>f8
Once again 26 . . . :ge6 27.:gbBt <j:rh7 2B.h4!
wins.
27.�b6
Black cannot contemplate glVlng up the
a6-pawn and trying to draw a rook ending
with four pawns versus five, as White's passed
d-pawn is protected and the e4-pawn is
vulnerable.
1 989
1 09
Anatoly Karpov - Ljubomir Ljubojevic
27. . J�ea8
27 . . J�e6 leads to a race which White is
just fast enough to win: 28Jhe6 fxe6 29.h4
me7 30.mh2 'i!?d6 3 1 .mg3 mc6 32.mf4 mb6
33.l'!a2 a5 34.'i!?e5 a4 35.'i!?xe6 'i!?b5 36.d5
Black can resign.
32.�c5t!
Once again Karpov prefers to send his
opponent's king back rather than going after
the e4-pawn.
32 ... <j.Jb8 33.�a2!
Karpov makes sure Black cannot push his
a-pawn.
33 ... �e7 34.i>f4 i>b7 35.�b2t i>a7
a
b
c
d
e
f
g
h
28.h4!
Having paralysed Black's rooks, Karpov
opens the way for his king.
28... <j.Je7 29.<j.Jh2!
Karpov continues his plan, rather than
getting distracted by a loose pawn. If 29.l'!e5t
md7 30.l'!b4 (Or 30.l'!xe4 a5 when Black gets
counterplay as Karpov pointed out.) 30 . . . a5
3 1 .l'!a4 l'!h8 32.l'!xe4 l'!h5 33. 'i!?fI mc6 Black
keeps some chances.
29 ... <j.Jd7 30.<j.Jg3 <j.Jc7 3 1 .�b2 �b7
8
a
b
c
d
e
f
g
h
36.�c6!
Another strong move; the threat is l'!bb6.
36 ... �h8
Black resigns himself to the loss of the
a-pawn. If 36 . . . l'!b7 37.l'!xa6t! 'i!?xa6 38.l'!a2t
wins.
37.�a2!
If 37.g3 l'!h5 ! Black remains in the game.
37... a5
Black cannot take the h4-pawn: 37 . . . l'!xh4t?
38.'i!?g3 l'!h5 39.l'!cxa6t mb8 40.l'!a8t mc7
4 1 .l'!2a7t md6 42.l'!d8t White wins, as
42 . . . 'i!?e6 43.l'!a6t leads to mate.
1-- //,,-/,
7
6
5
4
38.�xa5t i>b7 39.�ca6 �xh4t
Just for a moment Black restores the material
balance, but he remains defenceless against
White's threats.
3
2
a
b
c
d
e
f
g
h
The Prime Years
1 10
40.�g3 �h5 41 .�a7t �c6 42.�5a6t �b5
Black did well not to lose a rook, but now
he can do nothing to save his pawns from
falling.
a
b
c
d
e
f
g
h
43.�xe7 �g5t 44.�h2 �xa6 45.�xf7
1-0
Ljubojevic resigned as the endgame is
trivial.
In Round 7 Hjartarson made a mistake on
the white side of a complex Zaitsev position,
and Karpov punished him expertly. After a
bye in Round 8, Karpov scored a good win
against Portisch in a rook ending. In the next
game he held Ivanchuk on the black side of
a Zaitsev variation, and in the final round he
tried to press for a win against Timman but
had to settle for a draw. Karpov's final score
of 7/ 1 0 was decent, but it was only good
enough for second place, half a point behind
Ivanchuk.
Karpov's next event was a four-game match
against Ulf Andersson in Marostica in
northern Italy. In the first game he lost a
pawn for nothing but managed to hold a
draw. In the second it was Karpov who won a
pawn, but was unable to convert it. The third
game was a quick draw, and in the fourth
Andersson blundered in an equal position
before the time control. Thus Karpov won
the match 2Y2- 1 Y2, but the quality of play
was well below par for both of these super­
grandmasters. The database entries do not
specifY the time control, but I would assume
it was faster than normal .
The next event in the calendar was Karpov's
third World Cup tournament; this one
took place in Rotterdam. Karpov started
with a quick draw against Vaganian, then
he beat Nogueiras, although for a while he
was in trouble on the white side of a French
middlegame. In Round 3 Karpov won against
Seirawan after the American blundered in an
equal position, and then he defeated Timman
in a game that can be found in the notes to
Game 1 1 (Karpov - Kasparov, Belfort 1 988) .
Karpov's fine form continued; he held Andrei
Sokolov then defeated Portisch, although
he did not quite manage to capitalize on a
big advantage against Short. In the next five
rounds he drew with Ehlvest, defeated Yusupov,
drew with Van der Wiel, scored yet another
win against Hjartarson, and then defeated
Sax.
At this stage in the tournament Karpov
had a superb score of 9Y2/ 1 2, and looked to
have good chances to catch Kasparov in the
overall World Cup. But then, seemingly out of
nowhere, he suffered an incredible collapse and
lost his final three games. First he overpressed
against Sax, then he failed to make the most
of an excellent position against Ljubojevic and
went down. In the final round he was doing
fine against Nunn but got outplayed there
too.
Karpov's final score of 9Y2/ 1 5 was still good
enough for second place behind Timman, but
it could have been so much better. It looks
as though he simply ran out of energy after
playing too many tournaments. This had
occurred in a couple of his matches, most
notably against Korchnoi in 1 974 and the
1 989
111
Anatoly Karpov - Jaan Ehlvest
unfinished Kasparov match in 1 984/85, but he
had never suffered such a dramatic collapse in a
tournament.
Karpov's next tournament was the last in the
World Cup series. He drew his first five games,
including in Round 2 against Kasparov, who
was pressing for a win on the black side of a
King's Indian. The other four draws, against
Sax, Salov, Portisch and Nunn, were all
peaceful affairs.
In Round 6 Karpov met Jaan Ehlvest.
The Estonian grandmaster faced the world
champions fifty times. He won two of those
games, drew thirty one and lost seventeen. Up
to this point he had scored two draws and one
loss against Karpov. Their lifetime score stands
at four wins for Karpov, with eight draws and
no defeats.
1 Game 161
Anatoly Karpov - Jaan Ehlvest
Skelleftea 1 989
l.d4llJf6 2.e4 e6 3.llJB b6 4.g3 J.a6 5.'� a4
Karpov surprises his opponent by playing a
line he had never used before in a regular game,
although he often employed it in rapid games.
Karpov faced it from Black's side a few times
as well. At the time of writing, this variation
has returned to grandmaster tournaments,
showing that it is not without venom.
5.. ib? 6.ig2 e5 ? dxe5 i.xe5
Later both Polugaevsky and Judit Polgar
took back with the pawn; Karpov beat them
as well.
.
8.0-00-0 9.llJc3 ie? 10J�dl llJ a6
According to the database this natural
developing move was a novelty at the time this
game was played.
1 1 .ie3
The main line is 1 1..� f4. Karpov faced this
move from Black's side on four occasions,
scoring two wins and two draws. He also
used it with White and drew against Leko in
the third game of their 2006 rapid match in
Miskolc.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 1 . . .'�· e8 12.gad tLl e5 13.'� e2 llJ ee4
Black exchanges in order to create more
room for his pieces. Karpov mentions 1 3 . . . d5
and attaches "!?", but hardly anyone has tried
this suspicious-looking move. Perhaps there
was a typing error and he intended to suggest
1 3 . . . d6.
14.J.d4 B! d8
If 1 4 . . . Wxc4?? 1 5 .ltJxe4 wins.
15J�d3 tLlxc3
There is nothing wrong with this, although
it is not necessary to exchange yet and 1 5 . . . d6!?
is fine as well.
16.'� xc3 We?
Black is behind in development, so it not
logical for Black to open the position: 1 6 . . . d5
1 7.ltJe5 dxc4 1 8.Wxc4 ixg2 1 9.<;f;>xg2 Wxc4
( I 9 . . . Wb7t?! 20.Wc6) 20.1"1xc4 ic5 2 1 .<;t>f3
h6 22.1"1d3 White keeps a small but nagging
edge.
112
The Prime Years
Karpov prevents . . . b5, and in some positions
a4-a5 might be useful.
22 ... ttJ e8
22 . . . h4 would have been more consistent.
a
b
c
d
e
f
g
h
17.b3
Karpov effectively admits that he has not
obtained any advantage, so he settles for a
small improving move.
17.. J�ac8
1 7 . . . ttJ e4 1 8.�b2 i.f6 is also fine for Black.
18.'lWb2 i.e4 19.ttJel
a
b
c
d
e
f
g
h
23J�d3!
Karpov did not achieve anything significant
in the opening, nevertheless he continues to
find ways to strengthen his position.
23 ...i.f6 24J:!cdl a6 25.'lWd2
Karpov piles up his heavy pieces on the
d-file, yet the queen move has another more
subtle purpose.
8
7
6
5
a
b
c
d
e
f
g
h
Karpov exchanges the bishop in order to
create some possibilities in the centre.
19 ... i.xg2 20.ttJ xg2 'lWb7 21.llJe3 h5
This is not a bad move, although Karpov
prefers 2 1 . . .a6 intending . . . b5.
22.a4
4
3
2
1
a
25 ... i.e7?
b
c
d
e
f
g
h
An ato ly Karpov - Jaan Ehlvest
1 989
The Estonian grandmaster decides to keep
the bishops on the board, but he has overlooked
a crafty threat.
25 . . . Vfic7 was possible, although after 26.a5
bxa5 27.:B:a1 :B:b8 28.ixf6 ltJxf6 29.Vfixa5
White has some initiative on the queenside.
Black's best chance was to look for
counterplay:
25 . . . b5! 26.Vfib4!?
26.cxb5 axb5 27.a5 ixd4 (27 . . . d5?! 28.ib6!
l"i:d7 29.:B:c l is awkward) 28.:B:xd4 d5 29.b4
tiJf6 Black is in the game. (But note that
Black should avoid 29 . . . ltJ d6? because of
30.ltJxd5!.)
Karpov mentions 26.a5!?, so it is possible
that he wanted to play like this, but after
26 . . . bxc4 27.bxc4 ixd4 28.:B:xd4 ltJ f6 29.f3
l"i:c5 30.:B:d6 Black should be all right.
26 ... l"i:c7
113
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
26.a5!
Amazingly, this witty move wins material by
force.
26 bxa5
Black also loses material after 26 . . . b5 27.ib6
or 26 . . . ic5 27.axb6.
•.•
27.'lWxa5 ltJf6
The threat was 28.ib6.
8
7
28.hf6 hf6 29.'lWxh5
White picks up a pawn for no compensation.
At this level Black has little chance of surviving
such a loss. Nevertheless it is instructive to
observe how Karpov sets about converting his
advantage.
6
5
4
3
2
a
b
c
d
e
f
g
h
27.cxb5
27.c5 Vfic6 28.axb5 axb5 29.ltJc2 ixd4
30.ltJxd4 Vfixc5 3 1 .Vfixb5 Vfixb5 32.ltJxb5
l"i:b7 Black has solved all his problems.
27 ... axb5 28.a5
28.axb5?! ixd4 29.:B:xd4 :B:b8 is fine for
Black.
28 ...ixd4 29.Vfixd4 d5
The passed a-pawn makes White's position
somewhat preferable, but Black has decent
chances to resist.
8
7
6
a
b
c
d
e
f
g
h
114
The Prime Years
29 ...i.e7 30.�a5 d6 3 1 .b4
Karpov gains space.
40 ... gbc6?!
Ehlvest's last move before the time control is
an unfortunate one which eases Karpov's task.
3 1 ..J:!e8 32.�a4 @f8 33.h4
Stopping . . . j,g5 .
33 .. J"k6 34J:!b l gb6 35.gdb3 ga8
35 . . . l"IbS would probably have met with the
same response.
40 . . . g6 would have offered more resistance,
although after 4 l .l"Ib3 j,g7 (4 1 . . .�bS 42.b5)
42.b5 axb5 43.l"Ixb5 White should eventually
win with the extra pawn.
41 .b5!
Karpov opens the queenside, not to create a
passed pawn, but to invade with his rooks.
41 ...gc5
If 4 1 . . .axb5 42.l"Ixb5 �a7 43.l"IbS l"IcS
44.l"ISb7 Black is in trouble.
42.bxa6 �xa6 43.gb8 gc8 44.g l b6 �a4
45.gxc8
Exchanging one rook enables White to
invade with the second one.
a
b
c
d
e
f
g
h
36.�a2!
Karpov transfers his queen to the centre.
36 ... @g8 37.�c2 gc8 38.�d3
Now the queen is nicely centralized.
38 ...i.f8 39.@h2 gc7 40.g3b2
It looks like Karpov just made a move to pass
the time control.
45 ... gxc8 46.gb5 ga8 47.�e4
47.�b l intending l"Ib7 looks strong as well.
47...�a2?
Ehlvest misses a chance to reduce the number
of the pawns: 47 . . . d5! It looks rather risky, but
White has no direct win on the seventh rank.
a
b
c
d
e
f
g
h
4S.�f3 Planning l"Ib7. (Alternatively after
4S.�f4 �a6 [4S . . . dxc4 49.l"Ib7] 49.l"Ib3
dxc4 50.lLJxc4 Black faces a tough defence,
a
b
c
d
e
f
g
h
Anatoly Karpov - Jaan Ehlvest
1989
but it is not hopeless.) 48 . . .1.Wa6 49.:B:b2 dxc4
(49 . . . d4 50.tLlg4!) 50.tLlxc4 :B:a7 White has to
settle for an endgame with four pawns versus
three on the kingside. His winning chances are
quite high, as the knight will be stronger than
the bishop with the pawns all on one side.
48.YNf3
Karpov goes after the f7 -pawn. White has
another strong continuation as well: 48.tLlg4!?
fie7 (48 ... :B:e8 49.:B:g5! !i.e7 50.:B:h5 g6 5 1 .:B:b5!
wins) 49.:B:h5 g6 50.:B:b5 White will soon open
Black's kingside.
48 ... :B:a7
4B . . Wa6 49.:B:b7 is winning.
.
49.:B:b8
Black stopped the invasion on the seventh
rank, but the pin along the eighth is equally
powerful.
49 ...YNdl
8
7
6
115
50 'lWd4 5 1 .'lWc6 d5?
Ehlvest blunders the bishop in a lost
position.
•••
5 1 . . .Wc5 would have lasted longer, without
offering Black any real hope: 52.WeB d5
53.:B:cB (53.cxd5 exd5 54.tLle3 d4 5 5 . tLl g4
We7 56.tLle5 wins as well) 53 . . . :B:c7 (53 . . . We7
54.Wxe7 :B:xe7 5 5 .c5 wins, as Karpov pointed
out) 54.:B:xc7 Wxc7 5 5 .cxd5 exd5 56.tLlf4 Wc5
57.\t>g2 d4 5B.h5 White wins.
52.'lWd6
1-0
In Round 7 Karpov made a quick draw
with Vaganian, then defeated Seirawan in
an endgame masterpiece. A quick draw
with Ribli was followed by a good win over
Nikolic. Karpov then made four draws in a
row, including two quick draws, but then
he managed to beat Andersson in the last
round. This enabled Karpov to draw level with
Kasparov and share first place with a score of
9l/2/ 1 5 . Kasparov took first prize in the World
Cup overall, and Karpov was the only player
close to him.
World Championship Semi-Final
5
4
3
2
a
b
c
d
e
f
g
h
50.ttJg2!
Karpov covers his king while also setting up
ideas of tLl f4 and tLlxe6 or tLl g6.
After the less incisive 50.Wc6?! Wxe2 5 1 .WeB
�xf2t 52.'it>h3 'it>h7 53.WxfB Wf3 Black is still
alive.
Karpov's next event was a crucial one, as he
faced Artur Yusupov in the candidates semi­
final on the road to a future showdown with
Kasparov. The match took place in London,
and was scheduled for eight games. Although
Karpov was the heavy favourite, the match
turned out to be a lot more competitive than
most people anticipated.
Karpov had the black pieces in Game 1 , and
held a 4.g3 !i.a6 Queen's Indian without too
much trouble. In the second game he got into
trouble on the white side of a 4.Wc2 Nimzo­
Indian, but managed to salvage a draw. Game 3
The Prime Years
1 16
was a 4.a3 Nimzo-Indian in which Yusupov
sacrificed an exchange for an attack. Karpov
defended well and reached a winning position,
but in the time scramble he blundered and
allowed Yusupov the chance to win, but the
latter missed his opportunity and Karpov won
the ending. In the next game Yusupov equalized
with the Lasker Variation against the Queen's
Gambit, and outplayed Karpov in a simplified
position, but was unable to capitalize on his
advantage. Thus at the halfWay stage Karpov
was leading, but he had been in varying degrees
of trouble in three of the four games.
In Game 5 Yusupov played a Torre Attack
and obtained an advantage as early as move
5, as Karpov played an inaccurate move
order. Yusupov built his position skilfully and
eventually broke through just before the time
control to level the match. In the sixth game
Yusupov repeated the Lasker Defence. Once
again he outplayed Karpov, but missed a win
in the endgame and had to settle for a draw. In
Game 7 Karpov more or less equalized against
the Torre, and later sacrificed a pawn to reach
a drawn position with opposite-coloured
bishops, which he held comfortably.
In the eighth and final game, Yusupov
repeated the Lasker Variation which had
served him so well. This time, however, Karpov
was ready for it and played a much more
challenging counter. Yusupov was unable to
solve his problems and Karpov punished him
to win the game and the match.
***
After his successful though somewhat shaky
match performance, Karpov joined the Soviet
team at the second World Team Championship,
where he played on board one. His first game of
the event took place in Round 3, when he held
Ljubojevic with the black pieces. In Round 5
Karpov won a nice game against Short, and in
Round 8 he met Andras Adorjan of Hungary.
Adorjan faced all the world champions from
Smyslov to Kramnik, with the exception of
Fischer. He scored one win, twenty four draws
and six losses. Up to this point he had scored
one win, seven wins and two losses against
Karpov. This was the last time they met over
the board.
1 Game 111
Anatoly Karpov - Andras Adorjan
Lucerne 1 989
l .d4 lLl f6 2.c4 e6 3.lLlo b6 4.g3 i.a6
This is one of Adorjan's favourite lines.
Karpov also played it many times with both
colours.
5.b3 i.b4t 6.i.dl i.e7 7 ..ig2 c6 8.0-0 d5
9.i.c3 0-0 10.lLle5 lLlfd7 1 1 .lLlxd7 lLlxd7
12.lLldl E:c8 13.e4 b5
Adorjan deviates from his previous games.
In 1 985 he played 1 3 . . . c5 and drew with
Chernin, and in 1 986 he preferred 1 3 . . . dxe4
but lost to Ftacnik.
14.E:el dxe4
According to the database this move had
only been seen in one previous game, when
Timman used it to draw against Chernin in
1 987. Nowadays well over a hundred games
have been played with it.
1 5.Le4
Four years later Karpov attempted to improve
with 1 5 .c5?, but his novelty was refuted and has
never been repeated: 1 5 . . . f5 1 6.f3 b4! 1 7.j,xb4
liJ e5 ! 1 8.j,c3 liJ d3 1 9.fxe4 liJxe l 20.Wxe l e5
Black successfully converted his advantage in
Karpov - Timman, Netherlands (2) 1 993.
1 5 ... bxc4 1 6.bxc4 lLl b6
Adorjan is a rather dynamic player, so it is
117
Anatoly Karpov - Andras Adorjan
1989
bit surprising that he opted for a solid but
somewhat passive line such as this.
a
Karpov improves his knight, as exchanging
it would result in a truly miserable position for
Black.
1 6 ... cS is a common alternative.
20 ...�c7
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
17.c5!
Karpov does not mind giving away the dS­
square as he will gain numerous things in
return. The c6-pawn becomes a fixed weakness
and Black's pieces - with the exception of his
knight lack good prospects. After this game
this position did not occur again until 200S ;
presumably Queen's Indian players were put
off by the course of the present encounter.
-
17... tiJd5 18.�c2 g6?
Modern
tournament
practice
has
demonstrated that I S . . . h6! is stronger, as after
1 9.ttJf3 tiJ f6 Black can take the bishop without
having to worry about his dark squares.
19.ttJf3 J.f6
I found one other game from this position:
19 ... ttJf6 20.tiJeS lLlxe4 2 1 .'lWxe4 'lWdS 22.'lWxdS
cxdS 23.:gab l :gfeS 24.:gb3 f6 2S.lLlg4 :gc6
26j�eb l mf7 27.:ga3 :gaS 2S . .id2 .ics 29.:gf3
White obtained a clear advantage and went
on to win in P. H. Nielsen - Palac, Turin (01)
2006.
1
a
b
c
d
e
f
g
h
21 .h4!
Having built up a strong central position,
Karpov prepares to soften Black's kingside.
2 1 ..JHd8 22.J.d2 J.g7
Black is unable to get away with the cheeky
22 . . . lLl b4?? 23 . .ixb4 :gxd4 due to 24.lLlg4!, but
with his last move Adorjan prepares to launch
the knight.
23J�adl!
By defending the d-pawn White prevents
the . . . lLl b4 idea.
23 ... tiJe7
Adorjan fortifies his kingside in anticipation
of hS.
24.�c3
Now Black must worry about 'lWa3.
24 ...J.b5 25.J.c2!
Karpov attaches an exclamation mark to this
move, which stops any counterplay based on
. . . .ia4.
The Prime Years
118
2S ... lL'lf5?
Adorjan allows the bishop to come to f4 with
strong effect. He should have preferred one of
the following alternatives:
25 . . . tLld5 26.Wf3 f6 (26 . . . a5 27.a3 1':1bB
2B.h5 White remains in control) 27.tLlg4
Wf7 2B . .tb3 1':1eB 29.tLlh6t Black's position is
a bit unpleasant, but for the time being he is
surviving.
25 . . . f6 26.tLlc4 (26.tLlg4 Wd7 27.h5 is also
promising) 26 . . . tLld5 27.Wb3 .txc4 2B .Wxc4
Wf7 29.Wg2 f5 30 . .tb3 White certainly stands
better, but it will not be easy to crack Black's
position.
8
7
27...i.a6 28.i.e4 Wla8
Other moves were gloomy as well. For
instance, if 2B . . . h6 29.Wg2 Black can hardly
move, while after 2B . . . h5 29.1':1b l Was
(29 . . . We7 30 . .tg5) 30.Wa5 .tb7 3 1 .1':1xb7
Wxb7 32.tLlxc6 White wins.
29.hS!
Karpov prepares to land another heavy blow,
this time from the right side.
29 ... gS
Adorjan tries to fight back with a punch of
his own, but only succeeds in leaving himself
open for a haymaker.
Other moves were also dismal, for instance
29 . . . gxh5 30.Wf3 or 29 . . .1':1fB 30.Wa5 and
White is winning.
30.i.xgS! f6
6
5
4
3
2
a
b
c
d
e
f
g
h
26.i.f4!
Karpov activates his bishop. His play
from here on is reminiscent of a boxer's final
combination of punches which knocks out his
weary opponent.
26 ...Wlb7
If 26 . . . We7 27.a4 .ta6 2B.Wa5 .tb7 29 . .txf5
gxf5 30.Wxa7 Black has nothing for the pawn,
and 30 . . . 1':1aB 3 1 .Wb6 1':1xa4 is refuted by
32 . .tg5 f6 33.tLlxc6 winning.
27.a4!
Karpov knocks his opponent back with a
sharp left hook.
a
b
c
d
e
f
g
h
3 1 .h6!
Karpov exchanges pieces around Black's
king, so that soon there will be nothing left to
defend it.
3 1 . ..lL'lxh6
Both 3 1 . . . .thB 32.Wb3 and
32.hxg7 were equally hopeless.
32.i.xh6 i.xh6 33.Wlb3!
3 1 . . .fxg5
Anatoly Karpov - Andras Adorjan
1989
Threatening to invade from the centre. In
this game Karpov has skilfully created threats
all over the board.
33 .l:J:e8 34.�g4 'it>g7
..
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
35 � g2!
.
Even the king contributes to the massacre,
by opening the first rank for the rook to get
to the h-file.
35 5 36.�xh6 'it>xh6 37.Lf5!
...
A small sacrifice to end the game.
37 . exf5 38.Wff7
1-0
..
Black resigned before he was checkmated.
1 19
In the final round Karpov drew from the black
side of a Catalan in a long but always balanced
game against Nogueiras. Thus he finished with
a personal score of two wins and two draws,
and the Soviet team took the gold medal.
Karpov's final tournament of the year and the
decade was in Reggio Emilia. His first three
games against Andersson, Kiril Georgiev and
Mikhail Gurevich, were all drawn, although
each game was hard-fought. In Round 4 he
scored his first win after Beliavsky blundered in
an equal endgame. There followed quick draws
with Ehlvest and Portisch, a fighting draw
against Ribli and a Zaitsev ltJg5 repetition draw
with De Firmian. In the penultimate round
Karpov was able to grind down Petursson, and
in the final one he drew with Ivanchuk. His
final score of 6/ 1 0 was enough for third place
behind Ehlvest and Ivanchuk.
It seems that by this stage Karpov's age may
have started to become a factor. At thirty
eight he was far from ancient, but he would
not have had the energy reserves of a young
man, which may explain why he took more
quick draws than he had done previously.
During some parts of 1 989 Karpov was still
the same almost invincible tournament player
from previous years, but during some other
periods he dropped to the level of a 'mere' top
grandmaster.
1 20
1989 Summary
Candidates quarter-final match versus Hjartarson, Seattle: Won 3Yz- I Yz (+2 =3 -0)
Linares (2nd place) : 7/ 1 1 (+4 =6 - 1 )
Match versus Andersson: Won 2Y2-1 Yz ( + 1 =3 -0)
World Cup, Rotterdam (2nd place) : 9Y2/ 1 5 (+7 =5 -3)
World Cup, Skelleftea ( l st-2nd place) : 9Y2/ 1 5 (+4 = 1 1 -0)
Candidates semi-final versus Yusupov, London: Won 4Yz-3Y2 (+2 = 5 - 1 )
World Team Championship, Lucerne (Board one) : 3/4 (+2 = 2 -0)
Reggio Emilia (3rd place) : 6/ 1 0 (+2 =8 -0)
Total 63.2% (+24 =43 -5)
I!I Wins • Draws • Losses
1990
Rating 2730 (2 in the world)
Karpov began the year with the candidates final match against Jan Timman, a clash which would
determine the next challenger for Kasparov's crown. The match took place in the Malaysian
capital of Kuala Lumpur, and was scheduled for twelve games. The Dutch grandmaster was rated
2680 at the time - not quite at Karpov's level, but close enough to be competitive. Timman
was already an experienced match player and had proved himself capable of winning super­
tournaments. Karpov was certainly the stronger player when at his best, but his recent form had
dipped below his usual high standards, and he had looked especially vulnerable in the Yusupov
match of 1 989.
In the first game Timman introduced a bad novelty early in the Zaitsev, and immediately had
to struggle to stay in the game. Was it poor home analysis or an unfortunate piece of over-the­
board improvisation? Either way the outcome was unfortunate for him, and a further inaccuracy
allowed Karpov to press home his advantage and make a dream start to the match. Game 2
was a quiet draw in a symmetrical Fianchetto Griinfeld. In Game 3 Timman switched to l .d4
and opted for 4.f3 against the Nimzo-Indian. Karpov outplayed him, but Timman managed to
survive the endgame a pawn down. Game 4 was another Fianchetto Griinfeld, culminating in a
remarkable endgame in which both sides had connected passed pawns on opposite flanks. In the
end it went Karpov's way, and the former champion took a commanding two-point lead.
In Game 5 Karpov almost increased his lead after outplaying Timman in another Zaitsev, but
the latter escaped with a perpetual. In the next game Timman tried a Modern Benoni in an effort
to shake things up, but he lost a pawn and only managed to hold the ending by the skin of his
teeth after Karpov uncharacteristically missed a win. (An extract from the game can be found in
the notes to Game 1 9 below.) In Game 7 Karpov drew almost effortlessly with the Zaitsev. In
Game 8 Timman went for a complex variation in the Queen's Indian. For a while he was fine, but
he erred in the middlegame and Karpov punished him. Game 9 was a tense and complex Zaitsev.
For a long time it was dynamically balanced, but eventually Timman slipped up in the endgame
and Karpov scored another fine win.
With the score at 6Y2-2Y2 in Karpov's favour, the match was over after just nine of the scheduled
rwelve games. It was a resounding victory over a top contender, although Karpov was still probably
not completely satisfied as he failed to convert a couple of winning positions. Nevertheless he
proved beyond any doubt that he was once again the most worthy challenger to Kasparov's
crown.
***
1 22
The Prime Years
Just as in previous years involving a world
title match, Karpov did not take part in many
tournaments. His next event was in Haninge,
Sweden. He started with a draw with white
against Polugaevsky, then lost to Seirawan after
some risky opening play against the English.
In Round 3 Karpov was unable to make
any headway against Ehlvest, but in Round
4 he managed to grind out a win against
Wojtkiewicz. Ironically he did it from the black
side of a symmetrical Fianchetto Griinfeld.
Karpov's opponent in Round 5 was a familiar
foe, Ulf Andersson.
I Garne l8 1
Anatoly Karpov utf Andersson
-
Haninge 1 990
l.d411Jf6 2.e4 e6 3.lLlc3 i.b4 4.'iMe2
Karpov played the white side of this variation
sixteen times, scoring four victories, eleven
draws and one loss.
4 ... 0-0 5.a3 .hc3t 6.'iMxc3 b6 7.i.g5 i.b7
8.£3 d6
When Karpov played this line from the
opposite side he usually inserted 8 . . . h6 here,
and followed up with . . . d5 rather than . . . d6.
9.e3
Karpov deviates from the last game he played
against Andersson, which continued: 9.e4
c5 1 O.dxc5 bxc5 1 l .:8:d l tt:l c6 1 2.tt:lh3 tt:l d4
1 3. tt:l f2 h6 1 4 .�e3 e5 1 5 .�d3 a5 1 6.b3 �c6
1 7.0-0 Wb6 Black equalized and the game
was later drawn, Karpov - Andersson, Reggio
Emilia 1 989. It looks like this was enough to
persuade Karpov that he would do better not
to give up the d4-square at this stage of the
opening.
9 ... lLl bd7
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
10.llJh3!?
Karpov introduces a novelty. 1 0.�d3 is the
main line, but Karpov wants to keep the d-file
clear. At this stage in his career Karpov was
devoting more time to his opening preparation
than when he was the World Champion.
10 ... e5
Black can also play 1O . . . :8:c8 with the idea of
using the rook to recapture on c5 .
H.dxe5 bxe5 12.i.e2
Karpov aims to catch up on development,
with a view to exploiting his bishop pair in the
middlegame.
12 ... a5!?
Karpov would later choose a different path
when he encountered this position from Black's
side: 1 2 . . . Wb6 1 3 .0-0 d5 1 4.:8:ad 1 �c6!?
Black utilizes his bishop in an interesting way.
1 5 .tt:lf2 h6 1 6.�h4 �a4 1 7.:8:d2 �b3 1 8.tt:lg4
tt:lxg4 1 9.fxg4 :8:ab8 Kasparov - Karpov, Las
Palmas 1 996. At this stage Black had a decent
position, but he eventually lost after a time­
trouble error.
13J!:dl
Karpov plays his second novelty in the same
game! It sounds like a contradiction, but the
position before this move had previously been
reached via a different move order.
1 990
13 ... h6
1 3 . . . 1Mfb6 and 1 3 . . .l::\ a6 were reasonable
alternatives, which would have given Black
the option of keeping his pawn on d6, but
Andersson was obviously happy to advance it.
(After 1 8.Wfc3 ?! d4 1 9.1Mfb3 E!:b8! Black
takes over the initiative.) 1 8 . . . 1Mfe7! ( l 8 . . . g5 ?
1 9.itJxg5!) 1 9.E!:f2!
a
a
b
c
d
e
1 23
Anatoly Karpov - Ulf Andersson
f
g
h
14.if4!?
Karpov provokes his opponent into changing
the pawn structure in the centre.
14.ih4 allows a thematic simplification:
14 liJd5! 1 5 .ixd8 ( l 5 .1Mfxg7t? 'it>xg7 1 6.ixd8
...
i2lxe3 1 7.ie7 E!:fe8 Black wins material.)
15 . liJxc3 1 6.bxc3 E!:fxd8 1 7.E!:xd6 liJe5 Black
stands well as he will surely win back his pawn,
Zaitsev - Kimelfeld, USSR 1 989.
b
c
d
e
f
g
h
1 9 . . . E!:ad8! ( l 9 . . . g5 is well met by 20.1Mff5!;
19 . . . 1Mfxe3 20.1Mfxe3 E!:xe3 2 1 .cxd5 White has
a small edge as Black's queenside pawns are
weak.) 2o.if1 ! liJh5!? Black has other plans as
well. He should refrain from taking his pawn
back for the moment, as the queen exchange
would help White. 2 1 .1Mff5 liJ f6 22.E!:e2 ia6
Black keeps a playable position.
8
7
..
14 .. d5
14 ... e5!? looks playable as well. After 1 5 .ig3
both 1 5 . . . E!:a6 and 1 5 . . . 1Mfb6 are possible, and
Black will be able to play on the b-file more
freely than in the game.
.
15.0-0 Wfb6
By controlling the d6-square, Black sets
up the idea of . . . liJh5 to eliminate the dark­
squared bishop.
15 . e5!? This interesting pawn sacrifice was
analysed by Igor Zaitsev, Karpov's long-time
second. 1 6.ixe5 liJxe5 1 7.1Mfxe5 E!:e8 1 8 .1Mff4!
.
6
5
4
3
2
1
a
b
c
d
e
f
g
h
16.ig3!
Karpov preserves his important bishop. For
the moment Black cannot organize any plan
on the b-file, as the b8-square is covered.
.
16 ...ia6
It was worth considering 1 6 . . . liJh5!? in order
1 24
The Prime Years
to put a rook on the b-file: 1 7.ih4 �abB I B.g4
lLlhf6 1 9.ig3 e5 20.cxd5 lLlxd5 Black's activity
should be enough to counter White's bishop
pair and the slight weakness of the c5-pawn.
17 J�f2 ga7?!
This method of repositioning the rook is
artificial and takes too much time. Black
should have preferred one of the following
alternatives.
22.�c2 Wb3! works out well for Black.
22 . . . lLl d3! 23.ixb6 lLlxc l 24.�xc l lLlxb6
25.�ec2
White is fractionally better, but Black should
be able to hold the position without any special
effort.
1 7 . . . a4 I B.cxd5 lLlxd5 1 9.Wc l ixe2 20.�xe2
White is somewhat better as the c5-pawn will
soon become a target.
1 7 . . . �fcB!
It looks logical to bring the last piece into
play.
I B.cxd5
I B.ifI Wb7 1 9.�c2 lLl b6 Black's hanging
pawns are safely supported.
I B . . . lLlxd5 1 9 .Wc l ixe2
Also after 1 9 . . . c4! ? 20.e4 lLl 5f6 Black's
activity should compensate for his slightly
worse pawn structure.
20.�xe2 c4 2 1 .if2
a
b
c
d
e
f
g
h
1 8.cxd5!
Karpov chooses a good moment to clarifY
the situation in the centre.
The immediate I B. lLl f4?! would have allowed
I B . . . d4! 1 9.exd4 cxd4 20.Wxd4 Wxd4 2 1 .�xd4
e5 when White must give up an exchange.
1 8 ... c!tlxd5
The alternative was to go for a posItIon
with hanging pawns, in which Black is also
a bit worse: I B . . . exd5 1 9.ixa6 Wxa6 2o.lLlf4
Wc6 White is unable to hurt his opponent
immediately, nevertheless after something like
2 1 .b3 lLl b6 22.�c2 Black's position is a bit
unpleasant.
a
b
c
d
e
f
g
h
2 1 . . .lLle5!?
There was a second tactical solution available
in 2 1 . . .c3!? 22.e4 Wb3! 23.exd5 cxb2 24.Wb l
Wxd l t 25 .Wxd l �c l 26.�e l �xd l 27.�xd l
�cB 2B.ie3 �c3 with an unclear endgame.
22.e4
19.\Wc1 .be2
Exchanging pieces make it easier to ap­
proach the c-pawn. On the other hand after
1 9 . . . c4 Black's bishop is restricted, and White
maintains an edge: 20.e4 lLl 5f6 2 1 .Wc3 �cB
22.ifI lLlc5 (22 . . . �c6 23.\j;>h l lLl c5 24.�fd2
�aB 25 .ie5! Black has got his queens ide
125
Anatoly Karpov - Ulf Andersson
1990
together, but White begins to develop some
initiative on the kingside.) 23.E:d6 1.&b5
24.l"k2 Black is not much worse, nevertheless
his position would not be much fun to play.
24 . . . e5 25 .E:d6 ttJh5 (25 . . . c4 26.1.&e3) 26.ttJ g4
ttJxg3 27.hxg3 c4 2S.Wh2 Black is not
drastically worse, but he is passive and can do
little except sit and wait.
20.E:xe2 1.&b5 21 .�c2 �b7
The rook finally makes it to the b-file, but by
now White is ready to meet it.
Fixing the b2-pawn with 24 . . . a4!? was
possible, although Black remains worse here
too, for instance: 25.E:dc3 c4 26.E:xc4 E:xc4
27.E:xc4 1.&xb2 2S.1.&xb2 E:xb2 29. 'it>f1 Black
faces a difficult ending, as the a4-pawn is
vulnerable .
2z..�J£Z
Karpov wants to bring the knight to e4.
22... tLl5f6
After 22 . . . c4 23.ttJe4 E:b6 24.E:d4 Black
drops the c4-pawn.
8
7
6
8
5
7
4
6
3
5
2
4
1
3
a
2
a
b
c
d
e
f
g
h
23.e4!
Karpov restricts the enemy pieces. Black
faces the difficult problem of choosing between
several somewhat worse continuations, and it
is not easy to judge which one offers the best
chances.
23 ... �c8
1f 23 . . . c4 24.E:d4 ttJ b6 2 5 . ttJ d l E:d7 26.�f2
Elc8 27.ttJc3 all of White's pieces work well.
24.E:d3 llJb6?!
This leads to a worsening of Black's position.
He should have considered one of the following
alternatives:
b
c
d
e
f
g
h
25.b3!
Karpov restricts the black knight and
lures Black into a tempting but incorrect
continuation.
25 ... �d7 26.�dc3 c4?
Andersson takes the bait. It is understandable
that he wanted to rid himself of the weak
c-pawn and the suffering that goes with it.
The best chance was: 26 . . . a4! 27.b4 c4 2S.ttJ d l !
Transferring the knight to b2 increases the
pressure against both of Black's queens ide
pawns. 2S . . . ttJeS! (2S . . . E:cdS 29.ttJb2 threatens
�f2, and after 29 . . . ttJeS 30.�h4! f6 3 1 .�f2 ttJd6
32.�c5 Black's position is difficult.) 29.ttJb2
ttJ d6 Black does not have much freedom for
his pieces, but it is not easy for White to make
serious progress.
The Prime Years
126
32.�xa5
White picks up a second pawn, and the game
is as good as over.
27.bxc4 tvc5
32 ... tlJdc5 33.tvb l tlJb7 34.�b5 tlJd6
35.�b4 tlJc5
1-0
In this hopeless position Andersson
overstepped the time limit.
a
b
c
d
e
f
g
h
28.�f4!
The bishop switches to a better diagonal,
where it neutralizes Black's queenside play
and leaves White a pawn up for nothing. The
Swedish grandmaster may well have overlooked
this idea when making his 26th move.
28 ... tlJ a4 29.�b3 tvc6 30.�e3 �dd8
3 1 .�b5!
Karpov is not only a pawn up but also has
the more active pieces.
3 1 ...tlJd7
3 1 . . .l"i:a8 32.�xh6! gxh6 33.tvxh6 wins.
a
b
c
d
e
f
g
h
In Round 6 Karpov outplayed Hellers on the
black side of a Vienna Game, then he won
convincingly against Hector who ventured
the dubious Hennig-Schara Gambit. A quick
draw with Ftacnik was followed by a win over
Van der Wiel and draws against Karlsson and
Sax. Karpov's final score of 7Y2/ 1 1 was enough
to share second place with Ehlvest, behind
Seirawan who won the tournament by a full
point.
Karpov's second and last tournament of the
year before his championship match was
an eight-player, double-round-robin event
in Biel. For the first half of the tournament
Karpov followed a pattern of winning with the
white pieces and drawing as Black. His victims
in rounds 1 , 3 and 5 were Miles, Lautier and
Polugaevsky respectively, and the opponents in
the drawn games were Hort, Andersson and
Wahls.
In Round 7 Karpov faced Nick De Firmian of
the USA, who won his national championship
in 1 987, 1 995 and 1 998. De Firmian played
ten games against the world champions,
scoring two wins, three draws and five losses.
Prior to this event he had drawn twice with
Karpov, and after the final round of the present
event the two players never clashed again.
127
Anatoly Karpov - Nick De Firmian
1 990
The Benoni was a favourite opening of De
Firmian.
1 Game 191
Anatoly Karpov Nick De Firmian
-
Biel 1 990
l.d4 lLlf6 2.c4 e6
Earlier the same year Karpov played a fine
game against the Benoni, until the endgame
where he uncharacteristically allowed his
opponent to escape from a losing position.
2 ... c5 3.d5 e6 4.lLlc3 exd5 5.cxd5 d6 Timman
was already behind in the match and he
desperately needed to win, hence his somewhat
risky choice of opening. 6.e4 g6 7.ttJf3 �g7
8.1d3 0-0 9.h3 a6 1 O.a4 ttJ bd7 1 1 .0-0 1'Ific7
12.1f4 !"le8 1 3.!"le l !"lb8 1 4.�c4 ttJ b6 1 5 .�f1
IiJfd7
4.d5 exd5 5.cxd5 d6 6.llJc3 g6 7.h3
Karpov opts for a different line from the one
he used against Timman.
7 ... a6 8.a4 Wie7
Preventing e4.
9.i.g5
According to the database this move was
first played by Chernin in 1 990. It shows that
Karpov paid attention to the new moves his
contemporaries played.
9 ... llJ bd7 10.e3 i.g7 1 1 .i.e2 0-0 12.0-0 h6
13.i.h4 llJ e5
8
7
6
5
4
3
a
b
c
d
e
f
g
h
16.a5 After developing in a healthy way
Karpov pushes back the knight on the
queenside. 1 6 . . . ttJ a8 1 7. ttJ d2 b5 In order
to release
the grip Black must accept a
weakening on the queenside. 1 8.axb6 ttJ axb6
1 9.1xa6 Karpov grabs the pawn at once.
19 ... �xa6 20.!"lxa6 ttJe5 2 1 .1'Ifie2 c4 22.�xe5
1xe5 23.1'Ifie3 !"lb7 24. ttJ a4 ttJxa4 25.!"lxa4 c3
26.bxc3 1'Ifixc3 27.ttJ f3 Thanks to his strong
play White obtained a winning advantage
in Karpov - Timman, Kuala Lumpur (6)
1 990.
2
1
a
b
c
d
e
f
g
h
14.lLldl
White often retreats the knight to this
square voluntarily, but Karpov waited until his
opponent threatened to exchange it.
14 ... g5 1 5.i.g3 :Se8
Normally in this line Black either puts the
other rook on e8, or refrains from putting
either rook there. A common continuation is:
1 5 . . . �f5 1 6.a5 ( 1 6.e4 is also possible) 1 6 . . . !"lae8
White has tried several moves here, but overall
Black seems to be all right.
128
The Prime Years
16J3el 13b8?!
The rook will have no function on the b-file,
and Black soon feels the need to move it again.
16 ... .if5 was better.
a
17.a5 i.f5 18.e4
Karpov wins a tempo, and is unafraid of the
e-pawn becoming a target.
18 ...i.g6
Keeping the bishop on the other diagonal
with 1 8 . . . .id7 was insufficient to equalize:
1 9. 1Ll c4 lLlxc4 ( 1 9 . . . lLlh7 2o.lLl b6) 20 . .ixc4
The prospects of a breakthrough with e5 IS
very real.
19.13a4!
With the black bishop far away on g6, the
rook can use the a4-square without fear of
harassment. From this square it reinforces
the e4-pawn and inhibits the thematic
. . . c4. Note that defending the e4-pawn
with 1 9 . .ifl ?! would be less desirable due to
1 9 . . . lLlh5!.
1 9 ... tLl fd7 20.Y;Yc2 l3bc8 2 1 .b3
Karpov stops . . . c4. In some positions White
can consider lLl c4, intending to meet . . . lLl xc4
with bxc4, strengthening the centre and
opening the b-file.
b
c
d
e
f
g
h
2 1 ...Y;Ym
It may have been worthwhile for Black to
swallow his pride and admit his earlier error
with 2 1 . . .:gf8!?, for instance: 22.lLlfl f5 23.exf5
.ixf5 24.Wd2 :gce8 25.f4 (25 . lLl e3 .ig6 26.h4)
25 . . . gxf4 26.:gxf4 The position is unclear;
Black stands better than in the game, as his
major pieces have taken up good positions on
the open files.
22.tlJf1 f5
After 22 . . . lLl f6 23.lLle3 lLlh5 24 . .ixh5 .ixh5
2 5 . lLl f5 .ig6 26.Wd2 White will play f4 and
maintain an edge.
23.exf5 i.xf5
a
b
c
d
e
f
g
h
1 990
Anatoly Karpov
24.�d2
Karpov prepares to play on the kingside, where
Black has one less pawn to protect his king.
24... �f6
With 24 . . . c4!? Black can sacrifice a pawn to
improve the d7-knight: 2 5 .bxc4 � c5 26.Ela3
(26.E1b4 �g6) 26 . . . ltJ ed3 27.�xd3 �xd3
28.ltJe3 White keeps the advantage, but he
does not enjoy the same attacking prospects as
in the game.
-
129
Nick De Firmian
the present move does not combine well with
Black's previous one - if Black is going to
exchange on f4, then he should keep his strong
knight on e5.
27 ... Elc7! looks best, as the rook can play both
an active and a defensive role. 28.f5 (If White
opts for piece play, the active Black rook proves
useful: 28.�d3 �xd3 29.�xd3 ltJh5 30.fxg5
ltJ xg5 3 1 .ltJ f5 [3 1 . ltJ e2 Elce7 32.ltJg4 Wif7 is
playable for Black] 3 1 . . .Elxe 1 t 32.�xe 1 Elf7
Black's active pieces enable him to count on
a reasonable game.) 28 . . . �h7 29.g4 (29.�d l
Elce7) 29 . . . Elce7 30.\t>g2 ltJ d7 Black has some
pressure on the e-file and three of his minor
pieces are well-placed. These factors should
more or less compensate for the dismal position
of the h7 -bishop.
8
7
6
5
a
b
c
d
e
f
g
h
26.f4!
Karpov softens his opponent on the
kingside.
26... �f7
If 26 . . . gxf4 27.Elxf4 Elc7 28.Elefl White has
pressure on the f-file.
27.ifl
Defending the e3-knight is a useful
precaution. More importantly, White sets up
the idea of choking the g6-bishop with f5
followed by g4.
27... gxf4?
De Firmian must have been worried about
having his bishop locked away on h7, but
4
3
2
1
a
b
c
d
e
f
g
h
28.gxf4
Nine moves after its unusual development
on the fourth rank, the rook swings to the
kingside to spearhead White's attack.
28 ...�e7
Breaking the pin looks natural, but Black
will be left without enough pieces around his
king.
A better chance was 28 . . . Elc7! 29.Elfl ltJe5
30.ltJf5 Elf7 3 1 .�e3 when White keeps some
The Prime Years
1 30
advantage, but Black has better defensive
chances than in the game.
After 32 ... :B:f8 33.ig3 ttJe5 34.ig4 Black is
unlikely to last long.
29.ttJf5!
Karpov exchanges an important defensive
piece.
33.:B:fl :gf8
Black needs to watch the f6-square, for
instance after 33 . . . Wt'd7 34.ih4! 'kt>h8 35 .if6
White's attack should decide the game.
29 ... hf5 30Jhf5 fie7
The point of this move is not so much to
take the a5-pawn, but rather to sidestep the
potential pin from a bishop on h4.
30 .. .:1%f8 3 1..� h4 (3 l .id3 ttJ e5 32.ic2 is also
strong) 3 1 . . . ttJ e4 32.ttJxe4 Wt'xh4 (32 . . . Wt'xe4
33.ig4) 33.:B:f4 Wt'd8 34.ig4 :B:c7 35 .ie6
White's attack is dangerous.
8
7
6
5
4
3
2
8
1
7
a
6
b
c
d
e
f
g
h
34.i.el !?
Karpov has time to defend the a-pawn
before proceeding with his attack. In some
lines he may consider putting the bishop on
c3, especially in conjunction with a ttJxd6
trick.
5
4
3
2
1
a
b
c
d
e
f
g
h
3 1 .fif4!
De Firmian's queen just deserted the
kingside, which makes the arrival of Karpov's
heavy artillery all the more effective.
3 1 ... ttJ h7
If 3 1 . . .Wt'd8 32.ih4 ttJg5 33.ixg5 hxg5
34.:B:xg5 Black loses a pawn while his kingside
troubles remain.
32.lLl e4!
Karpov brings another piece into the attack.
32 ... lLl e5
34... b5
Black is losing in all variations, for example:
34 . . . :B:xf5 ?! 3 5 .Wt'xf5 :B:e8 36.ih5 :B:e7 37.ih4
wins.
34 . . . :B:cd8 3 5 .ic3 Black has no good defence
against ttJxd6, for instance 35 . . . 'kt>h8 36.ttJxd6
(36.Wt'g3 :B:de8) 36 . . . Wt'xd6 37.Wt'xe5! ixe5
38.ixe5t and White wins.
35.axb6
Although Karpov is winning on the kingside,
he still takes the opportunity to draw the black
queen further away, while also weakening
Black's pawn structure.
131
Anatoly Karpov - Nick De Firmian
1 990
35 .. JWxb6 36.<i!lhl
White has time to safeguard his king before
commencing the final assault.
36 ... ct!lh8 37.�g3 gxf5
If 37 .. .E!ce8 38."We3 (38.i.c3 is also strong)
38 ... Ei:e7 39.Ei:xf8t itJxf8 40.i.h4! White
wins.
38.Ei:xf5 gg8 39.�e3
8
a
7
b
c
d
e
f
g
h
41 .lLlxd6!
The d6-pawn has been a target for a long
time. Finally it drops, and Black's whole
position collapses.
6
5
4
41. �xd6 42.gxe5 �f8 43.ge6
1-0
Black decided not to prolong his suffering.
3
.•
2
a
b
c
d
e
f
g
h
39 ... a5
The American grandmaster is short of time
and defends against b4.
A slightly more resilient defence would have
been 39 . . J%f8, although White still wins
with 40.E:h5! (Instead after 40.E:xf8t itJxf8
4 1 .b4 itJfd7 Black is still breathing.) 40 . . . E:b8
(40 ... a5 4 1 .i.c3 'it>g8 42.i.xe5 dxe5 43.d6+-;
40 ... itJf7 4 1 .i.c3 E:e8 42.i.xg7t 'it>xg7
43.Ei:f5+-) 4 1 .i.c3 c4 (4 1 . . ."Wxb3 42.itJxd6+-)
42.�xb6 E:xb6 43.itJxd6 Black's position
collapses.
40.i.g3 gb8?
Black blunders a pawn but his position was
hopeless anyway, for instance 40 . . . Ei:e8 4 1 .i.f4
a4 42.bxa4 "Wb 1 t 43.'it>h2 itJg4t 44.i.xg4
1"1xe4 45 ."Wf3 with a decisive advantage.
In the second half of the tournament Karpov
slowed down and drew six consecutive games.
Three of them ended in under twenty moves,
and the other three were played out to the
endgame. In the final round he scored a second
win over De Firmian to finish with a score of
9Yz/ 14, a point and a half clear of Andersson
who was second.
World Championship Match
versus Kasparov
The fifth and, as it turns out, last Karpov
- Kasparov match was scheduled for the
customary twenty four games, the first half of
which took place in New York and the second
in Lyon. Kasparov's team consisted of Zurab
Azmaiparashvili, Sergey Dolmatov, Mikhail
Gurevich, Giorgi Georgadze and Alexander
Shakarov. Karpov was assisted by Ron Henley,
Lajos Portisch, Andrei Kharitonov, Alexei
Kuzmin, Mikhail Podgaets and Igor Zaitsev.
132
The Prime Years
Kasparov performed exceptionally well in
1 989, and by the time of the match his rating
had reached the magical 2800 level, seventy
points above Karpov. At twenty eight years
of age he was coming into his prime, whereas
Karpov was approaching forty. Many people
expected Kasparov to demolish Karpov or
at least beat him convincingly. Since their
previous match the two titans had contested
five games, from which Kasparov had scored
two victories to Karpov's one, with two
draws.
after an early queen exchange the game soon
simplified to a level ending. In Game 6 Karpov
temporarily abandoned the Zaitsev in favour
of the 9 . . tLJ d7 system, and obtained a fine
position. But later he took a pawn in a way
that allowed White to become very active, and
in the end Karpov was fortunate to draw. Game
7 was another King's Indian. This time Karpov
got a nice advantage and Kasparov blundered
badly, allowing Karpov to win a pawn which
he converted efficiently. The match was
all-square.
In the first game Kasparov played the Kings
Indian; it seems to me that choosing such a
combative opening showed a high level of
confidence. Karpov chose the ambitious
Samisch variation, but did not achieve
anything special and Kasparov drew without
much difficulty. In Game 2 Karpov repeated
the Zaitsev Variation which had brought
him so much success against Hjartarson and
Timman, but Kasparov was ready with an
excellent novelty, and after some powerful
follow-up play Karpov was demolished.
In Game 3 Kasparov went for another King's
Indian, and it soon became clear that he was
looking for a complicated fight as opposed to
merely equalizing with the black pieces. As early
as move 1 0 he sacrificed an exchange, and a few
moves later he transitioned to a position with
rook, bishop and pawn versus Karpov's queen.
The idea was fully sound and in the end it was
Karpov who had to fight for the eventual draw.
In Game 4 Karpov tried a different method of
handling the Zaitsev, and a wildly complicated
middlegame ensued. Kasparov failed to
make the most of his attacking chances, and
Karpov obtained a winning position with four
connected passed pawns on the queenside,
but just before the time control he erred and
allowed an instant perpetual.
In Game 5 Karpov once again failed to make
any headway against the King's Indian, and
In Game 8 Karpov stuck with the 9 . . . tLJd7 line
against the Ruy Lopez. Kasparov obtained a
strong initiative, but misplayed the position
and after some strong play Karpov emerged
with the advantage. He was close to winning
the endgame, but eventually Kasparov was
able to cling on for a draw. Game 9 was the
first Griinfeld of the match. The queens were
exchanged early and Karpov kept a slight plus
for most of the game. Towards the end he had
a chance to obtain a significant advantage, but
he faltered and dead drawn position ensued.
In the next game Karpov surprised Kaspatov
by switching to the Petroff, and drew in just
eighteen moves. In Game 1 1 Kasparov returned
to the King's Indian, and played a nice exchange
sacrifice which led to full compensation and
a subsequent draw by perpetual. In Game
1 2 Karpov reverted back to the 9 . . . tLJ d7 Ruy
Lopez. Kasparov was ready with a new idea
and obtained an edge, which subsequently
disappeared as the position simplified. Thus at
the halfWay point the match was tied at 6-6,
with one win apiece and eight draws.
.
Lyon
The players received a long rest period of sixteen
free days between games 1 2 and 1 3 . Karpovwas
unable to achieve much in the next Griinfeld,
and despite some minor inaccuracies on both
1 990
Anatoly Karpov - Garry Kasparov
sides, neither player had any real winning
chances and a draw was the natural outcome.
In his first white game of the Lyon part of the
match, Kasparov showed that he had been
hard at work during the break as he unveiled
a surprise opening, the Scotch. Karpov did
not shy away from complications, and a sharp
position with opposite-sided castling ensued.
Neither player could gain the upper hand, and
the game was eventually drawn.
Game 1 5 was another Griinfeld with an
early queen exchange. Karpov was able to gain
control but failed to make the most of his
chances and the players agreed a draw for the
eighth consecutive time. In Game 1 6 Karpov
tried a different line against the Scotch, but his
new idea was not particularly impressive and
he soon had to sacrifice a pawn for insufficient
compensation. Kasparov did not convert his
advantage in the most efficient way - perhaps
he was affected by the tension of the match.
However, he eventually succeeded in breaking
Karpov's fortress on the 1 02nd move.
Trailing by one point, with eight games
remaining, Karpov needed to strike back. He
was able to do it in the very next game.
I Game 20 I
Anatoly Karpov - Garry Kasparov
New York/Lyon ( 1 7) 1990
l.d4 tLlf6 2.c4 g6 3.lLlc3 d5
Kasparov sticks with the opening that has so
far brought him three draws out of three in the
present match.
4.cxd5 tLlxd5 5.e4 tLlxc3 6.bxc3 i.g7 7.i.e3
Karpov prepared this variation specially for
the match - he had never used it previously.
He continued to use it until 1 993, scoring a
total of five wins, three draws and one loss.
a
b
c
d
e
133
f
g
h
7 ... c5 8.'i;Ydl 0-0
The first time the players reached this
position, Kasparov exchanged in the centre
without delay: 8 . . . cxd4 9.cxd4 lbc6 1 O.2"i:dl
Wi'a5 1 1 .Wi'xa5 lbxa5 1 2.lbf3 0-0 1 3 .ie2 id7
1 4 .id2 b6 1 5 .0-0 2"i:fd8 1 6.2"i:cl ig4 1 7.d5
White was slightly better but in the end Black
was able to hold, Karpov - Kasparov, New
York/Lyon (9) 1 990.
Karpov subsequently faced 8 ... Wi'a5 three times;
he beat Timman twice but lost to Kamsky.
9.tLla i.g4
Kasparov tries to undermine the d4-square.
The position after 9 . . . Wi'a5 1 O.2"i:cl e6 occurred
in two previous games in the match:
I l .d5?! This ambitious move is somewhat
premature. 1 1 . . .exd5 1 2.exd5 2"i:e8 1 3 .ie2
if5 1 4.0-0 lb d7 1 5 .h3 lb b6 ( 1 5 . . . lb f6! was
stronger.) 1 6.g4 id7 1 7.c4 Wi'xd2 1 8.lbxd2
lba4 The position is roughly equal and the
game eventually ended in a draw, Karpov Kasparov, New York/Lyon ( 1 3) 1 990.
Two games later Karpov showed a much better
way to handle the position: 1 l .ih6 lbc6 1 2.h4
cxd4 1 3.ixg7 �xg7 1 4.cxd4 Wi'xd2t 1 5 .\t>xd2
2"i:d8 1 6. �e3 The king is ideally placed here.
1 6 . . . id7 1 7.2"i:b l 2"i:ab8 1 8 .id3 lb e7?! 1 9.h5
The Prime Years
134
f6 20.hxg6 hxg6 2 1 .�h2 White obtained a
promising attack but failed to make the most
of his chances and eventually drew, Karpov Kasparov, New York/Lyon ( 1 5) 1 990.
a
a
b
c
d
e
f
g
h
10.tLlg5!?
Karpov finds a nice way to safeguard the
knight from being exchanged. It was a new
idea at the time, and has since been repeated
in several dozen games.
b
c
d
e
f
g
h
14 ... tLl a5
Since Karpov's novelty on move 1 0,
both sides have played naturally, and most
subsequent games have proceeded in the same
fashion.
15.i.d3
1 5 .�e2 is another option, leading to some
subtle differences.
1 5 i.e6
Black had scored fairly well with 1 5 . . . tZlc4,
eliminating one of the enemy bishops. 1 6.�xc4
�xc4 1 7.0-0 b5 From this position 1 8.�fc1
Was is fine for Black, but 1 8.�h6!? deserves
attention.
..•
1 0 ... cxd4
1 O . . . h6 1 1 .h3 is not helping Black.
1 l .cxd4 lLl c6 12.h3 i.d7
Black cannot really take the pawn: 1 2 . . . i.xd4?!
1 3.�xd4 Wxd4 1 4.Wxd4 tLlxd4 1 5 .hxg4 tZl c2t
1 6.'tt> d2 tZlxa1 1 7.�d3 Black faces a depressing
endgame.
The other capture is even worse: 12 . . . tZlxd4?
1 3.hxg4 �c8 1 4.�d l ! tZl c2t ( 1 4 . . . �c2 1 5 .Wb4)
1 5 .'tt> e 2 Wc7 1 6.c;t>f3! �fd8 1 7.�d3 White has
a large advantage, as pointed out by Mikhail
Gurevich.
1 6.0-0 i.c4
If 1 6 . . . tZlc4?! 1 7.�xc4 �xc4 1 8.�fc l White
is a bit better.
17.gfdl
Karpov supports his d-pawn, with a view to
advancing it later.
17 b5
This move was condemned by some
commentators, but in my opinion the real
mistakes occurred later.
.•.
13J�bl gc8 14.tLla
Taking the pawn would have been a dreadful
mistake: 1 4.�xb7?? tZl xd4 1 5 .�xd4 �xd4
1 6.Wxd4 �c l t 1 7.'tt> d2 �d l t! 1 8.c;t>xd 1 �a4t
Black wins the queen.
If 1 7 . . . e6 1 8.�g5! is somewhat unpleasant.
1 990
17 ... b6 has been played a few times, and is
probably the safest continuation. Play may
continue l S . .ig5 (other options include l S.d5
and lS . .if4 'lWd7 1 9.�bc l ) lS ... 'lWd7 1 9.�bc l
with approximate equality.
However, the active 2 1 . . .tLl c4!? was worth
considering, for instance: 22 . .ixe7 (22.�b 1
'lWd7) 22 . . . �eS 23 . .ia3 tLl xa3 24.'lWxa3 �xe4
25 .d5 .ifS Black is not worse.
a
a
b
c
d
e
135
Anatoly Karpov - Garry Kasparov
f
g
h
18 .ig5!
Karpov forces Kasparov to keep an eye on
the e7-pawn.
.
18 a6 19J�bc1
Taking the pawn would squander White's
small advantage: 1 9 . .ixe7?! 'lWxe7 20.'lWxa5
i,xd3 2 1 .�xd3 'lWxe4 Black regains the pawn
and the endgame is equal.
...
b
c
d
e
f
g
h
22.�b 1
22.a4 bxa4 is nothing special for White.
22.�c l tLl c4 23.a4 e6 Black should be safe.
Finally, if 22.'lWa3 tLl c4 23.'lWxe7 'lWxe7
24 . .ixe7 �eS 25 . .ic5 �xe4 Black wins back
the pawn with a good position.
22 . . . �cS!
It is too early to insert 22 ... h6?! as 23 . .id2!
is strong.
23.'lWa3
19 hd3
In the event of the natural 1 9 . . . �eS, Karpov
may have been tempted to keep his light­
squared bishop with 20 . .ib 1 . It keeps the game
more complex, which Kasparov may not have
wanted due to the standing of the match.
.••
20.13xc8 �xc8 2 1 .�xd3 13e8?!
The rook is misplaced here. A sounder
alternative was:
2l ...Wfb7
From this square the queen can defend both
the a6- and the e7-pawns.
Instead 2 1 . . .'lWd7?! 22.'lWa3 is strong as
Zaitsev pointed out.
a
b
c
d
e
f
g
h
23 . . . h6!
After 23 . . . tLl c4 24.'lWxe7 'lWxe7 25 . .ixe7 �eS
26 . .ic5 �xe4 27.a4! Black loses a pawn.
24 . .ih4
24.'lWxa5 hxg5 25 .e5 �c4 26.a4 b4 The
position is a bit irregular, but Black should
be all right.
The Prime Years
136
24 . . . ltJc4 25 .1Wxe7 1Wxe7 26.i.xe7 �eB 27.i.c5
�xe4 2B.a4 bxa4
Black can hold this ending.
a
b
c
d
e
f
g
okay; 26.i.e3!? ltJxe3 27.1Wxe3 �cB 2B.ltJb3
i.fB 29.ltJa5 White is better as the knight
will be strong on c6.) 26 . . . ltJxd2 27.1Wxd2
b4 (if 27 . . . �cB 2B.�xcBt 1WxcB 29.d6 Black
is close to losing) 2B.�c6 a5 29.1Wc2 Black is
in trouble.
26.i.g3 ltJ xd2
26 . . . ltJe5 27.i.xe5 i.xe5 2B.ltJf3 is difficult
for Black.
27.�c7! 1Wb6 2B.1Wxd2
White keeps some advantage thanks to the
powerful rook on the seventh rank.
h
22Jkl!
Naturally Karpov occupies the open file.
22 ...Y«b7
Kasparov chooses to keep both of his
vulnerable pawns defended. 22 . . . 1Wd7 was
possible, although after 23.1Wa3 ltJ c4 24.1Wxa6
i.xd4 25.ltJxd4 1Wxd4 26.1Wxb5 ltJ d6 27.1Wd7
Black faces a difficult endgame.
23.d5
Karpov gains space, fixes the e7-pawn and
secures control over the c6-square.
23 ... � c4
Kasparov hopes to cover the c-file, but it
seems he overlooked a nice little tactic.
23 . . . h6
This was worth considering, but it also fails
to solve Black's problems.
24.i.f4 ltJ c4 25.ltJd2 g5
Compared with the game continuation,
taking on d2 would cost Black a tempo due
to the attack on the h6-pawn.
25 . . . e5 gives White a choice: 26.i.g3 (After
26.dxe6?! �xe6 27.ltJxc4 1Wxe4! Black is
a
b
c
d
e
f
g
h
24.�d2!
Karpov wants to exchange Black's best piece
and opens the c-file to invade.
24 ... �xd2
24 . . . h6 2 5 .i.f4 transposes to the note to
Black's previous move.
If 24 . . . ltJ d6 25 .�c6 �cB 26.1Wc2 White can
keep up the pressure.
Mikhail Gurevich mentions the reasonable
24 . . . ltJe5!?, although here too White can take
advantage of his control over the c-file. 25 .1Wc2
1Wd7 26.i.e3 f5 27.f4 ltJ f7 2B .1Wc6 �dB 29.i.b6
1Wxc6 30.dxc6 �cB 3 1 .c7 White keeps a clear
advantage.
25.,hd2 �c8
1 990
137
Anatoly Karpov - G arry Kasparov
28JWd4! f6
25 . . . b4 26J'k6 a5 27.\Wa6 is strong.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
26J:�c6!
Kasparov apparently missed this tactical
finesse.
29.ia5!
With this subtle positional move, White
increases his possibilities on the c-file.
26 ...ie5
If 26 . . . :gxc6 27.dxc6 \Wc7 (27 . . . \Wxc6?
28.Wi'dBt wins) 2B.\Wd7 ie5 29.ih6! \Wxd7
30.cxd7 ic7 3 1 .e5! White cages in Black's
king and wins by walking his king to the
queens ide.
29 ...id6 30.V9c3
Karpov could have won a pawn with
30.\Wb6, but after 30 . . . \WaB (30 . . . :gbB 3 1 .\Wxb7
:gxb7 32.:gxa6) 3 1 .:gxcBt \WxcB 32.\Wc6 \WfB
33.\Wxa6 \Wh6 Black gets some play.
27.ic3!
Karpov increases his domination of the
c-file.
27 ... ib8
If 27 . . . :gxc6? 2B.dxc6 \Wc7 29.ixe5 \Wxe5
30.Wi'dBt 'it>g7 3 1 .c7 \Wa l t 32.<;t>h2 \We5t
3 3.g3 \Wb2 34. <;t>g2 Black has no perpetual.
30 ... E:e8 3 1 .a3!
Karpov fixes the queenside without fearing
the bishop on d6.
3 1 ...'if7g7 32.g3 ie5 33.V9c5 h5?
Repelling the queen with 33 . . .id6 was
better, although after 34.\Wc I White still has
excellent winning chances.
Perhaps Black should have considered
27 . . . ixc3 2B.\Wxc3 :gxc6 29.dxc6 (If 29.\Wxc6
Wi'a7 30.e5 \Wd4 Black becomes active.)
29 ... \Wc7 30.e5 when he faces a dismal and
passive endgame, but nevertheless keeps some
chances to survive. When Kasparov gets a bad
position he generally avoids such scenarios in
favour of keeping more pieces on the board, in
the hope of drumming up counterplay.
a
b
c
d
e
f
g
h
The Prime Years
138
34.i.c7!
Karpov wins control over the c7 -square and
prepares the final invasion.
34...i.al 35.i.f4 Y;9'd7 36Jk7
Black could have resigned here, but he
struggles on for a few more moves.
36 ...Y;9'd8 37.d6 g5 38.d7 �f8 39.i.d2
This is not the only solution, but it is one of
the cleanest. Karpov prevents . . . Wa5 to avoid
giving Black the tiniest hint of counterplay.
39 ...i.e5
a
b
c
d
e
f
g
h
40.�b7
1 -0
The finish might have been 40 . . . h4 (40 . . . Wg6
4 1 .Wa7+-) 4 1..� a5! Wxa5 42.Wxe7t Wg6
(42 . . . Elf7 43.Wxf7t+-) 43.Wh7t! Karpov
must have enjoyed demonstrating his analysis;
he played so well in the earlier part of the game
so he earned it. 43 . . . <j;>xh7 44.d8=Wt The
reincarnated queen seals Black's fate.
With this win Karpov equalized the score, but
Kasparov struck back in the very next game,
a Ruy Lopez. Karpov played the 9 . . . ltJ d7
system in one game too many, and Kasparov
obtained a large advantage straight from the
opening. Karpov resisted for a long time,
but eventually found himself in a hopeless
endgame which Kasparov duly converted. In
Game 19 Kasparov switched back to the King's
Indian and gradually outplayed Karpov in a
semi-blocked position. Then on move 39, in a
clearly better position, he mysteriously offered
a draw which Karpov gladly accepted.
In Game 20 Karpov returned to the Zaitsev,
and a wildly complicated middlegame ensued.
Unfortunately for him the gamble did not
pay off, and Kasparov won with a violent
kingside attack. With this victory he took an
1 1 -9 lead with four games remaining. Karpov
desperately needed to win the next game, and
he almost managed it on the white side of a
Samisch King's Indian, but Kasparov narrowly
escaped with a draw after surviving a difficult
endgame. In Game 22 Kasparov deviated from
the previous game in the Zaitsev, but his idea
was unsuccessful and he had to struggle to
draw. Nevertheless he did it and thus reached
the magic score of twelve points, which meant
that he would retain his title. However, there
was still the matter of the prize fund: the
winner of the match would receive 5/8 of the
$3,000,000 prize fund, whereas a draw would
result in a 50/50 split. Money aside, there was
also a tremendous amount of pride at stake,
and in the next game Karpov showed that even
without the possibility of winning the title he
would still fight until the bitter end.
I Game II I
Anatoly Karpov - Garry Kasparov
New York/Lyon (23) 1 990
l .d4 ltJ f6 2.c4 g6 3.ltJc3 i.g7 4.e4 d6
After losing with the Griinfeld in Game 1 7,
Kasparov returned to the King's Indian and
stuck with it until the end of the match.
5 .0
1 990
1 39
Anatoly Karpov - G arry Kasparov
The Samisch was Karpov's most frequent
choice against the King's Indian. He lost
only one game with it (courtesy of a magical
performance by Kasparov at Linares 1 993) but
won fourteen and drew seven.
ltJxc4 1 3 .l"&c1 ltJa6 1 4.ltJdl ltJb6 1 5 .ltJh3 �d7
1 6.tDe3 f6 1 7.l"&hgl l"&ad8?! 1 8.b3 c6 1 9.dxc6
bxc6? 20.ltJf5! gxf5 2 1 .l"&xg7t Wxg7 22.l"&g l t
White obtained an excellent position and went
on to win, Karpov - Velimirovic, Skopje 1 976.
5 ... 0-0 6.i.e3 e5
Kasparov always played this move against
Karpov. The latter did well against the 6 . . . ltJ c6
variation, dropping just half a point from four
games.
10.0-0-0
Naturally Karpov has no interest in the
dubious pawn grab 1 O.g4?! ltJ f4 1 1 .�xf4?
( l l .ltJge2 is better) 1 1 . . . exf4 1 2.\Wxf4,
especially when facing such a formidable
attacking player.
7.d5
In 1 993 Karpov used 7.ltJge2 to beat
Kamsky and Dolmatov, but came unstuck
against Kasparov.
7... tLlh5
In their next Samisch battle in Reggio
Emilia 1 99 1 192, Kasparov opted for 7 . . . c6.
That double-edged encounter ended in a draw,
but for most of the game Karpov was fighting
to survive.
8.�d2 �h4t
Kasparov deviates from Game 2 1 , which
continued: 8 . . . f5 9.0-0-0 a6 1 O.�d3 c5 1 1 .dxc6
tiJ xc6 1 2.ltJd5 �e6 1 3 .�b6 \Wd7 1 4.ltJe2 l"&ac8
1 5 .Wb 1 \Wf7 1 6.l"&he 1 In this complex position
White's chances are somewhat higher, Karpov
- Kasparov, New York/Lyon (2 1 ) 1 990.
1 0 f5 1 l .exf5!
Karpov's strategy is to gain control over the
e4-square, and this is the first step on the way.
•••
1 1 . gx£5
.•
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
9.g3 �e7?!
Kasparov plays a very rare move, which may
have been inspired by Akopian who tried it
earlier the same year. The idea is rather risky,
but having already defended his title, Kasparov
may have wanted to make a statement by
demolishing Karpov by a heavy score line.
12.<�jh3!?
Karpov unveils a good novelty. The knight
is useful in a number of ways: it discourages
Black from playing . . .f4, supports a possible
advance of the f3-pawn to f4, and also prepares
the plan of ltJ f2 and g4 as seen in the game.
Karpov was among the trailblazers in the
9 . tD xg3 variation which leads to an unusual
material balance. Here is one of his early games
with it: 1 O.\Wf2 ltJxfl 1 1 .\Wxh4 ltJxe3 1 2.We2
A previous game continued 1 2 .�d3 a5
1 3.ltJge2 ltJ a6 1 4.f4 �d7 1 5 .fxe5 dxe5 1 6.l"&hfl
with some advantage for White, Vyzhmanavin
- Akopian, Lvov 1 990. Kasparov had surely
prepared an improvement over that game.
. .
140
The Prime Years
12 ... ltJ a6
If 1 2 . . . f4? 1 3.gxf4 .txh3 1 4 . .txh3 ttJxf4
( 1 4 . . . exf4 1 5 . .td4) 1 5 . .txf4 ( 1 5 . .tg4 is also
strong) 1 5 . . . E!:xf4 1 6.E!:dg 1 White gets strong
play on the g-file.
13J!gl !
Karpov prepares to play g4, in order to fight
for the crucial e4-square. If he can achieve that
goal he will have a good chance of condemning
Black's prized "King's Indian bishop" to a grim
defensive role.
13 ... ltJf6!
Kasparov brings the knight back from the
edge of the board to inhibit the g4-push.
Allowing it was rather unappealing for Black.
controls the e4-square and will soon launch a
kingside attack.
14.ltJfl
Karpov brings the knight into position to
support the advance of the g-pawn.
1 4 . .td3!? Karpov mistakenly gave this move
a double question mark in his analysis. It is
true that Black can win a piece, but White gets
massive compensation for it: 1 4 . . . f4 1 5 .gxf4
.txh3 1 6.fxe5! dxe5 ( 1 6 . . . Wxe5 1 7 . .th6 E!:f7
1 8 .E!:g5 White regains the piece with a winning
position.) 1 7.E!:g3 .td7 1 8 . .th6 E!:f7 1 9.E!:dg l
ttJe8 2 0 . .txg7 ttJxg7 2 1 .Wh6 Black i s dead
lost.
1 3 . . . ttJ c5 1 4.g4 fxg4 1 5 .fxg4 ttJ f4 1 6.ttJxf4
exf4 1 7 . .txf4 .txc3 1 8.bxc3 .td7 ( 1 8 . . . ttJe4
1 9.We3) 1 9.E!:e l Wh4 20 . .tg5 Wf2 2 1 ..te3
Black runs out of play for the sacrificed pawn.
1 3 . . . �h8 1 4.g4! fxg4 ( 1 4 . . . ttJ f6 1 5 .gxf5 .txf5
1 6.ttJg5 White achieved his objective and
controls the e4-square; Black's best chance
is probably to keep the position closed with
1 4 . . . f4, but after 1 5 . .tf2 ttJ f6 1 6 . .td3 White
controls the e4-square and the game.) 1 5 .fxg4
a
b
c
d
e
f
g
h
14 ... <bh8?!
This king move is not especially helpful to
Black.
a
b
c
d
e
f
g
h
1 5 . . . ttJ f6 (After 1 5 . . . ttJ f4 1 6.ttJxf4 exf4 1 7 . .txf4
Black has no play for the pawn.) 1 6. ttJ f2
Black has serious positional difficulties. White
1 4 . . . .td7
This was a more purposeful move, intending
to centralize the queen's rook.
1 5 . .te2
The ambitious 1 5 . .td3!? is also interesting,
for instance: 1 5 . . . E!:ae8 1 6.E!:de l Wf7 1 7.g4!
f4 ( 1 7 . . . e4 1 8.g5! exd3 1 9.9xf6 Wxf6
20.ttJxd3 �h8 2 1 .ttJ f4 Black is in trouble)
1 8 . .txa7 E!:a8 1 9.95 ttJh5 20 . .txh7t @h8
2 1 .ttJb5 Even if Black manages to win the
1990
141
Anatoly Karpov - Garry Kasparov
misplaced bishop o n a7, White will still have
excellent play.
1 5 .. .l'heS 1 6.i.g5
16.g4? f4 1 7.i.xa7 b6 traps the bishop.
However, White could consider switching
plans with 1 6.l'1:de l !? Wf7 1 7.f4 when his
central pressure gives him at least a slight
plus.
1 6 . . . Wf7 1 7.l"i:dfl lLl c5 I S.g4 fxg4 1 9.fxg4
\Wg6 20.h3
White's position is preferable, but Black's
problems are less severe than in the game.
15.i.e2
With this seemingly modest move Karpov
prepares to play g4.
1 5 .i.d3 was a good alternative, for instance:
1 5 . . . i.d7 ( 1 5 . . . lLl c5 1 6.i.c2 a5 1 7.g4 f4
I S.i.xc5 dxc5 1 9 .Wd3±) 1 6.We2 l"i:aeS I 7.i.b l
White's advantage is similar to the one Karpov
obtains in the game.
15 ... i.d7
Centralizing the knight at once would
encourage the advance of White's g-pawn:
1 5 . . . lLlc5?! 1 6.g4! f4 1 7.i.xc5 dxc5 I S.i.d3
White has a big advantage, as pointed out by
Karpov.
Karpov makes one more preparatory move
before playing g4. He could have done
it immediately, but in view of Kasparov's
tremendous abilities in dynamic positions, it
was a sensible idea to minimize any possible
chances for his opponent to sparkle.
16 lLl c5
...
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
17.g4!
Karpov carries out his plan to gain control
over the e4-square. His advantage is dear.
17 e4
Kasparov tries to complicate the game
by opening the long diagonal for his prized
bishop. It fails to achieve the desired effect, but
the alternatives would also have led to a bleak
situation for Black.
••.
1 7 . . . fxg4 I S.fxg4 a6 1 9.i.e3 White will soon
launch an attack with h4 and g5, and his
knights will obtain the use of the e4-square.
1 7 . . . a5 I S.h4 ( 1 S.gxf5 i.xf5 1 9 .1Ll d3
[ 1 9.lLlfe4!?] 1 9 . . . lLl a6 20.We3 is also good)
I S . . . f4 1 9.1Llfe4 White dominates the game.
1 7 . . .f4 may have offered the most resistance,
but after I S.Wc2 a5 1 9.1Lld3 lLla6 20.i.h4
Black has a depressing position with no
The Prime Years
142
counterplay - a nightmare scenario, especially
for a dynamic player like Kasparov.
1 8.fxe4 fxe4 19.�e3!
Karpov blocks the e-pawn and more
importantly frees the path of his g-pawn.
19 ... ttJ a4 20.g5!
Karpov does not mind incurring a doubling
of his queenside pawns, as White is streets
ahead in the attacking race, and the enemy
pieces are a long way from his king.
20.lLlxa4
This simple move also leads to a large
advantage for White.
20 . . . �xa4 2 1 .g5!
After 2 1 .:gde l lLl d7 Black is not in immediate
danger.
2 1 . . .lLl d7
2 1 . . .lLl g8 22.:gdf1 ±
22.b3 We5
24.Wxe4
White picks up a pawn and keeps good
prospects on the kingside.
24 . . . lLl c5
If 24 . . . :gxf2 25 .Wxe8t :gxe8 26.�xf2 :gxe2
27.:gd2 White should win the endgame.
25 .�xc5!?
White can also play the simple 25 .Wxe8
:gaxe8 26.'it>d2 with a clear extra pawn in the
endgame.
25 . . . dxc5 26.�d3 Wh5 27.Wg2
White's knight will go to e4 next. He keeps a
clear advantage, although Black's strong dark­
squared bishop gives him some chances to
resist.
20 ... ttJxc3 2 1 .bxc3
a
a
b
c
d
e
f
g
h
23 .Wd4!
Karpov misses this strong move in his Chess
Informant analysis.
23.Wc2? :gxf2 24.�xf2 Wf4t! (24 . . . e3? loses
to 2 5 .�e l ! - but not 25 .�g3? Wa l t 26.Wb l
Wc3t leading to a perpetual, which was the
only line mentioned by Karpov.) 25 .'it>b l
Wxf2 Black even takes over the initiative.
23 . . . We8
23 . . . Wxd4 24.:gxd4! wins for White.
b
c
d
e
f
g
h
2 1 ...ttJg8?!
Kasparov probably wanted to have a chance
to use the f8-rook in the attack, but it never
happens. The problem with the text move is
that Black's king is trapped in the corner.
2 1 . . .lLl e8
This was a better chance. Black keeps an
escape square for his king and defends the
d6-pawn, which might be useful if the
queenside opens up. Nevertheless Black is in
huge trouble here as well.
22.h4!
1 990
This direct attacking move is the strongest.
After 22.lLlg4 c5 23.h4 White's advantage is
clear, but not necessarily decisive yet.
22.ig4 c5 23.Wc2 (23.ixd7 Wxd7 24.lLlxe4
�a4 Black is still breathing thanks to his
queenside counterplay.) 23 . . . lLl c7 24.ixd7
�xd7 25.lLlxe4 b5 26.lLlf6 White should be
winning here roo, but the main line seems
clearest.
22 . . . We5
22 . . . b5 23.cxb5 E1b8 24.a4 a6 25.bxa6 ixa4
26.E1g4 wins.
22 . . . c5 23.h5 lLl c7 24.h6 ie5 25.lLlxe4
b5 26.lLlf6 also gives Black little chance to
survive.
a
b
c
d
e
143
Anatoly Karpov - Garry Kasparov
f
g
h
23.E1dfl !
White does not have to defend the c3-pawn,
and should instead concentrate on blasting
through the enemy kings ide.
Karpov only analyses 23.id4 e3 (23 . . . Wf4
24.lLlg4) 24.Wxe3 Wxe3t 25 .ixe3 ixc3
when White keeps the advantage, but Black
is still alive.
23 . . . c5
23 . . . Wxc3t 24.Wxc3 ixc3 25.lLlxe4 ig7
26.h5 Black has no good defence against
h6.
23 . . . if5 24.h5 Wxc3t 25 .Wxc3 ixc3
26.lLlxe4 also wins.
24.h5 lLl c7 25 .h6 Wxc3t 26.Wxc3 ixc3
27.lLlxe4 ie5 28.lLlf6
White is winning.
a
b
c
d
e
f
g
h
22.tlJg4!
Preventing . . . We5 . When reading Karpov's
annotations to this game, one gets the
impression he wanted to suggest that he had
found the one correct path through a labyrinth.
The twelfth World Champion is justifiably
revered for the quality of play, rather than that
of his published analysis.
22.h4!
This direct attacking move was also good
enough to secure White's victory.
22 . . . We5 23.�b2
23.id4?! is less accurate, although it can lead
to some interesting variations: 23 . . . e3?! (This
was the only move mentioned by Karpov,
but 23 . . . Wf4 is safer.) 24.Wd3! Karpov
failed to mention this move. (After 24.Wxe3
Wxe3t 25 .ixe3 E1ae8 26.<±>d2 ixc3t Black
is better, as Karpov pointed out.) 24 . . . We7
25.lLlg4 E1ae8 White is somewhat better.
23 . . . b5
a
b
c
d
e
144
The Prime Years
Karpov ends his analysis here, concluding
that Black has some initiative. However, a
closer inspection reveals that White has a
huge advantage.
24.cxb5
24.lLlg4 �xg4 2 5 .:B:xg4 bxc4 26.�d4 is also
powerful.
24 . . . :B:ab8 25.lLlg4 We8 26.a4 a6 27.�d4
Black is busted.
22 e5?!
This leads to a further deterioration of Black's
position; it was clearly not Kasparov's day. He
probably lost faith in his position and could
not summon the motivation to find the most
accurate way to lose.
.•.
22 . . . :B:ac8 puts more force behind Black's
queens ide play, despite being a bit slow. 23.h4
c6 (23 . . . �e8 24.lLlfl c6 25 .�g4 :B:c7 26.�e6
White's advantage is clear.) 24.h5 cxd5 25 .h6
�e5 26.lLlxe5 dxe5 27.Wxd5 �a4 28 .:B:d2
Black does not have enough play against
White's king.
a
b
c
d
e
f
g
h
25.cxd5!
Karpov has correctly judged that Black has
insufficient firepower to hurt his king.
25.h5!? Wa3t 26.'j{ b l Wxc3 (26 . . .�xc3
27.�d4t) 27.Wxc3 �xc3 28.cxd5 �d7 29.:B:c1
wins as well.
23.dxe6 Le6
23 . . . bxc6 24.Wxd6 wins.
25 ...Ld5
The sacrifice is unsound but other
continuations were equally hopeless, for
instance 25 . . . :B:ad8 26.d6 or 25 . . . Wa3t
26.Wb2.
24.h4!
Now Karpov threatens to destroy Black's
kingside with h5 followed by h6 or g6.
26.Wxd5 :B:ae8
No better is 26 . . . Wa3t 27.�b l Wxc3
28.�d4+-.
24 ... d5?!
This desperate pawn sacrifice hastens the
end, but Black's fate was already close to being
sealed.
27.Wd6 :B:xc3t 28.@bl Wffl 29 ..id4
1-0
The g7 -bishop will shortly be exchanged,
and Kasparov saw no point in fighting on.
The World Champion was not at his best,
nevertheless Karpov's play was strong and
instructive.
With this win Karpov maintained his hopes
to tie the match and obtain an equal share of
the prize fund, but defeating Kasparov with
the black pieces was an almost impossible task
if the champion adopted a risk-free approach
to the game. Sure enough, Kasparov opened
24 . . . :B:ad8 25.h5 b6 (25 . . . d5 26.h6 wins the
bishop.) 26.g6! hxg6 27.hxg6 Black will soon
be mated on the h-file.
24 . . . Wc7!? Improving the knight looks like
the best chance. Nevertheless after 25 .h5 lLl e7
26.h6 �e5 27.lLl xe5 dxe5 28.�c5 :B:f7 29.�d6
White has a huge advantage.
1 990
145
with l .ttJ f3 and obtained a slight plus in a quiet opening. Karpov tried to complicate the game
but made some mistakes. Kasparov obtained a winning position but offered a draw to secure the
match, and Karpov accepted.
The final result once again showed that Kasparov was the worthy champion, although the match
was always competitive and Karpov left no doubt that he was still an enormously strong player.
Kasparov once again dominated the opening battles. He took a risk by preparing only l .e4,
but it turned out to be a good decision as he won four games with it without suffering a loss.
Karpov scored one win against the Griinfeld with three draws, which was not a bad ratio. In his
remaining white games he scored two wins against the King's Indian, but this was nothing special
considering that one of them was the result of a blunder (admittedly in a clearly worse position)
from Kasparov, and the other occurred after Kasparov had retained his title.
***
1 990 turned out to be the final year in which Karpov and Kasparov would clash for the world
title. Over the course of their epic rivalry, these two titans faced each other a total of 1 44 times in
five matches over a seven-year period (not including tournaments) . There had never been a rivalry
like it in the chess world, and maybe there never will be again. Botvinnik and Smyslov played
three world championship matches, but even this was not really comparable.
Over the course of the five matches Kasparov scored twenty one wins to Karpov's nineteen,
with 1 04 draws. Kasparov has a plus score in the games lasting more than forty moves, having
won fifteen times versus Karpov's thirteen. This apparently surprising fact can be explained by
the fact that when Karpov had a lost position he would often wait until the adjournment before
resigning. Out of the games lasting more than forty five moves, each player scored six wins.
The match results leave no doubt as to who was the stronger player. Nevertheless every match was
competitive, and ironically their first 'marathon match', which lasted a gruelling forty eight games
before being aborted, was the only match which did not last for the full scheduled duration. Both
players experienced triumphs and frustrations. Karpov never won a match against Kasparov; he
was leading 5-0 in their first match but was unable to secure the elusive final win, and he was
leading 1 2-1 1 in Seville but lost the final game. On the other hand Kasparov, despite being the
stronger player, was never able to defeat Karpov by a truly resounding margin. The 1 990 match
was the only one in which he retained the title with at least a game to spare, but even then it
took until Game 24 for him to reach the magic score of 1 2Yz points, which was needed to secure
outright victory. Nevertheless my impression is that the gap in playing strength was larger than
the close match results would suggest.
146
1990 Summary
Candidates final match versus Timman, Kuala Lumpur: Won 6Yz-2Y2 (+4 = 5 -0)
Haninge (2nd-3rd place) 7Yz/ l l (+5 = 5 - 1 )
Biel ( l st place) : 9Y2/ 1 4 ( + 5 =9 -0)
World Championship match versus Kasparov: Lost l l Y2-1 2 Yz (+3 = 1 7 -4)
Total 60.3% (+ 1 7 =36 -5)
CI Wins • Draws • Losses
1991
Rating 2725 (2 in the world)
Karpov started the year in which he would turn forty at the Reggio Emilia tournament in
northern Italy, a seven-player, double-round-robin event. Karpov started with a quiet draw
against Andersson and had a bye in Round 2. In the next two games he drew with Ehlvest and
Ribli. Curiously his first three games were all Queen's Indians. In Round 5 Polugaevsky employed
the Tarrasch Defence, and Karpov was unable to make any headway. In the next game he made a
solid draw against Mikhail Gurevich.
After five consecutive draws Karpov finally scored his first win, on the white side of a Queen's
Indian against Kamsky. The young rising star sacrificed a pawn for insufficient compensation and
struggled hard, but eventually went down.
In the first game of the second half of the event, Karpov faced Andersson. Since Karpov's victory
in Haninge (see Game 1 8 in the previous chapter) the two players had drawn twice: in Biel 1 990,
and the first round of the present event.
I Game ll l
Anatoly Karpov - Ulf Andersson
Reggio Emilia 1 9 9 1
l.d4 ttJf6 2.c4 e6 3.ttJf3 .ib4t
Andersson played this opening a total of seven times against Karpov, but only achieved the
dismal total of two draws and five losses.
4.ttJbd2
This move brought Karpov a total of two wins and one draw against the Swede. Karpov also
played 4.id2 throughout his career, scoring a total of seven wins and four draws, including three
wins and one draw against Andersson.
4 ... b6 5.g3
Karpov deviates from two of their previous encounters, which continued as follows: 5 .a3
.txd2t 6.ixd2 ib7 7.ig5 d6 8.e3 tiJ bd7 9.ih4 c5 1 0.id3 0-0 1 1 .0-0 cxd4 1 2.exd4 d5 From
this position Karpov chose 1 3.�e 1 in the Skelleftea World Cup event of 1 989 and went on to
win an endgame with a microscopic advantage. Later in Biel 1 990 he varied with 1 3.cxd5 but
only managed to draw.
148
The Prime Years
5 ...i.a6 6.\Wc2
This position is often reached via the Queen's
Indian move order of 3 . . . b6 4.g3 ia6 5 . 4J bd2
(or 5 .Wc2 ib4t 6.4J bd2) 5 . . . ib4 6.Wc2.
12 tlJbd7
1 2 . . . e5? would be ill-advised due to 1 3 .dxe5
dxe5 1 4.ic3 4J c6 1 5 .4Jxe5! when White wins
a pawn.
6 i.b7 7.i.g2 i.e4
This is one of the soundest approaches for
Black, although Andersson subsequently
switched to 7 . . . c5.
13.i.c3 13fe8 14.13ac1 h6
The move is not a mistake in itself, but it
is indicative of the timidity with which the
Swedish grandmaster subsequently conducts
his position. A few other ideas deserved
attention.
.•.
8.\Wb3 ,bdlt 9.,bdl
Karpov secures the advantage of the two
bishops, although it will be tough to put
them to good effect in the immediate future.
My junior trainer and lifelong friend Karoly
Honfi told me that many players overestimate
the importance of the two bishops. I like to
teach my pupils that a pair of bishops is a
bit like chili or soy sauce. When the correct
ingredients are in place they can turn a decent
recipe into a fabulous one, but without those
other elements the effect is nothing special. Let
us see how Karpov sets about adding the right
ingredients to his recipe.
..•
14 . . . a5 seems generally useful for Black
although it does not change the position
much.
1 4 . . . e5!? 1 5 .dxe5 (After 1 5 .ih3 exd4 1 6.Wxd4
White's advantage is microscopic.) 1 5 . . . 4Jxe5
1 6.b3 a5 1 7.4Jxe5 ixg2 1 8 .'ifixg2 dxe5 White
only has a small advantage.
1 5 .b3 13ac8 16.i.h3!
Karpov wants to play 4J d2 and later e4,
and since White controls more space, the
avoidance of a bishop exchange is definitely in
his interest.
1 6 c6
Karpov mentions 1 6 . . . ib7, with the possible
plan of . . . 4J e4 and . . . f5 . Play may continue:
1 7.d5 !'i:cd8! Karpov does not mention this
subtle move, which threatens both the d5- and
e2-pawns. ( 1 7 . . . exd5? 1 8.Wxe7 !'i:xe7 1 9.ixf6
Black is in serious trouble; 1 7 . . . e5 1 8.4Jh4
White has the initiative.) 1 8.dxe6 fxe6 1 9.4Jd4
e5 Black has an acceptable position after
20.4Jf3 Wf7 or 20.4Jb5 4Jc5.
•.•
a
b
c
d
e
f
g
h
9 ... 0-0 10.0-0 d6 l l J3fdl
Now if Black ever plays ... c5, White will
have the option of gaining space with d5 .
1 1 ...\We7 12.\We3
Karpov brings the queen to the centre and
holds back . . . e5 for the moment.
Another playable alternative is 1 6 ... c5 when
Black looks to be all right.
Finally there is 16 . . . if5!? 1 7.ixf5 ( 1 7.ifl
a5) 1 7 . . . exf5 1 8 .Wxe7 !'i:xe7 1 9.e3 4J e4 when
Black's position is perfectly healthy.
1 99 1
149
Anatoly Karpov - Ulf Andersson
20 .. J�ec8 2 1 .i.a3!
Trying to provoke . . . c5.
2 1 ...ltJe8
By now it would have been difficult for Black
to advance a pawn even if he had wanted to do
so.
For instance, after 2 1 . . .c5?! 22.ttJc3 gd8
23.ib2 ttJ f8 24.Wd2 gcd7 25 .e4 White has a
clear advantage.
a
b
c
d
e
f
g
h
17.ib2!
At first glance it is hard to see the point
behind this subtle move. Karpov's idea is to
vacate the c3-square for the knight, while in
some positions a bishop transfer to a3 could be
a useful option.
17 ... gc7
Andersson's plan seems to be to avoid pushing
any of his pawns beyond his third rank. 17 . . . d5,
17 ... if5 and 17 ... a5 were all playable.
18.ttJd2 i.g6 19.1tJ b l
Th e position i s closed s o Karpov has time to
improve his pieces. By this time he had probably
picked up on the fact that his opponent was
not aiming for a sharp conflict, thus he takes
full advantage of the opportunity to improve
every piece.
19 ... a6
Gaining space with 1 9 . . . d5 is less favourable
here than it was before: 20.Wf4! (20.ttJc3?! dxc4
2 l .bxc4 Wb4 Black wins a pawn.) 20 . . . gec8
(Also after 20 . . . gcc8 2 1 .i.a3 White is better.)
2l .ia3 c5 22.ttJc3 Black's hanging pawns are
under pressure.
2o.ig2
Karpov recentralizes his bishop and takes
away the e4-square from the black pieces.
Alternatively there is 2 1 . . .b5?! 22.Wf4
ttJ e8 23.ttJc3 when Black cannot shake his
opponent's grip on the centre.
22.ltJc3 �d8
22 . . . c5? loses to 23.ttJd5.
23.�d2
Intending to advance in the centre.
23 .. J�b7
The immediate 23 . . . b5 would have reduced
White's options, although after 24.ib4 gb7
the first player can maintain some advantage
with 25.cxb5 or 25.ttJe4.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
24.e4!
When Black employs a queens ide fianchetto
against a queen's pawn opening, the execution
1 50
The Prime Years
of the e2-e4 advance usually represents a
significant achievement for White, especially
if he has avoided exchanging the light-squared
bishops, as is the case here.
Had Black left the queenside alone with moves
like 30 . . . lLl c7 or 30 . . . lLl fS, Karpov would
have played for a slow but powerful kingside
attack.
24... b5
Finally Black pushes one of his pawns beyond
his third rank.
3 1 .gal Vf!c7
3 1 . . .l"kbS!? would have given White one less
option, as 32.a3? allows 32 . . . a4! 33.axb4 axb3
when Black gets some counterplay. Instead
White can simply reply with 32.a4 as in the
game.
25.lLle2
Aside from gaining space, Karpov's previous
move also had the merit of vacating this retreat
square for his knight.
25 .. J�bb8
In the event of 25 . . . bxc4 26.bxc4 Black
gets the b-file but can do nothing with it:
26 .. .'IWc7 (26 . . . c5? Opening the position only
helps the bishops. 27.lLl f4 �h7 2S.e5! White is
dominating.) 27.lLl f4 �h7 2s.lLld3 White is in
full control.
26.lLlf4 �h7 27.c5!
Seizing space. A good alternative was 27.f3!
to restrict the enemy bishop.
27... d5 28.e5
Karpov goes for more even space. A different,
but equally promising approach was to open
the game: 2S.exd5!? exd5 (Or 2S . . . cxd5 29.c6
lLl df6 30.Wb4 and Black is desperately passive.)
29.�b4 Black is in trouble as White has good
chances to invade along the e-file.
28 ... a5!
Andersson stops �b4, thus enabling him to
keep the queens ide closed.
29.gel
Karpov ensures that Black will not be able to
loosen the grip by means of . . . f6.
29 ... ga8 30.�b2 b4
Black hopes to obtain some active play on
the queenside, but the plan soon backfires.
a
b
c
d
e
f
g
h
32.a4
Karpov elects to close the queenside in
order to develop his kingside attack without
distraction.
He could also have gone for active play on the
queenside: 32.a3!? Wb7 (32 . . . a4? is ineffective
here due to 33.axb4 axb3 34.b5! when White
creates a strong passed pawn.) 33.axb4 axb4
34.�fl White keeps a clear advantage here too,
but Karpov probably wanted to avoid giving
his opponent the opportunity to exchange the
rooks on the a-file.
32 ... bxa3?
Though it was a depressing prospect, Black
should have kept the position closed and
attempted to survive the inevitable kingside
assault. In a way it is a pity that we did not
Anatoly Karpov - Ulf Andersson
199 1
151
get to see how Karpov would have built his
attack.
7Y2/ 1 2 was enough for sole first place, half a
point ahead of Polugaevsky.
33J�xa3
Now Black simply loses the a-pawn by
force.
Karpov's next tournament was in Linares.
With hindsight, the 1 9 9 1 event is notable for
the participation of four young players who
would go on to become some of the greatest
in the world: Anand, Gelfand, Ivanchuk and
Kamsky (in alphabetical order) . Karpov started
with a black win over Ehlvest, although at one
point he made a serious mistake on which the
Estonian grandmaster failed to capitalize. Also
in Round 1 a twenty one year old Ivanchuk
scored a convincing win over Kasparov, which
must have made Karpov feel like he had a real
chance to win the event. However, in Round
2 he suffered a setback and lost to Anand after
making a couple of uncharacteristic positional
errors.
Karpov's tournament would soon go from
bad to worse. In Round 3 he outplayed
Ljubojevic on the black side of a Zaitsev, but
then spoiled a highly favourable position and
lost after some subsequent mistakes. In the next
game he got a significant advantage against
Yusupov, but failed to make the most of it and
only drew. Then in Round 5 Karpov was put
under uncomfortable pressure by the in-form
Ivanchuk, and eventually overstepped the
time limit after missing a chance to equalize.
The game reminded me of Botvinnik's win
over Alekhine at the famous AVRO 1 938
tournament, an event in which the Patriarch
defeated both Alekhine and Capablanca in two
fabulous and instructive games. Five rounds
into the Linares tournament, Ivanchuk had
repeated Botvinnik's trick of defeating the top
two players in the world at that time. Though
I cannot be certain, this may be the only time
that someone has defeated both Kasparov and
Karpov at the same tournament.
33 .. .'IWb7
36Jl:a4
a
llJc7
34 ..ic3
b
c
d
e
35 ..ixa5
f
g
llJb5
h
1-0
It is a bit early to resign, but Andersson
must have been disappointed with his
play, and he knew that the chances of
surviving such a position against Karpov
were almost nonexistent. White has a clear
positional advantage in addition to his extra
pawn, and Black does not have a hint of
coumerplay.
After a second bye in Round 9 Karpov
drew with Ehlvest, Ribli and Polugaevsky, in
games lasting twenty six, twelve and sixteen
moves respectively. Then in Round 1 3 he
used the Samisch against the King's Indian
to gain a slight plus against Gurevich, which
he eventually converted to a win in a long
endgame. In the final round Karpov was under
some pressure against Kamsky, but eventually
managed to survive an endgame with rook
versus rook and bishop, after eliminating his
opponent's last pawn. Karpov's final score of
After this disappointment Karpov met another
brilliant rising star in Boris Gelfand. At the
The Prime Years
152
time of writing the grandmaster from Minsk
has played 1 50 games against the world
champions, scoring seventeen wins, ninety
seven draws and thirty six losses. This was the
first time he played Karpov. Their lifetime
score stands at seven wins to five in Karpov's
favour, with seventeen draws.
Before looking at the game, I would like to
say something about these numbers. Although
Gelfand is a world-class player with extensive
experience at the top level, comparing his
statistics (in particular the 1 50 games against
the champions) with those of the older
generation of super-GMs is a bit misleading.
Nowadays most top-class tournaments involve
roughly the same group of players competing
against one another, which was not the case
in the seventies and eighties. Players such as
Geller, Portisch, Gligoric and Timman were
not less great - they merely did not get to play
as many of those types of tournaments. To take
another example of how times have changed,
John Nunn won two individual gold medals
at Olympiads, and these magical performances
deservedly earned him several invitations to
tournaments where he was able to face world
champions. In recent years Gabriel Sargissian
has performed incredibly well at Olympiads,
but at the time of writing the Armenian
grandmaster has not played a single regular
game against Anand, Kramnik or Topalov,
due to the exclusivity of the elite tournament
circuit.
4...0-0 5 ..ig2 c6
Gelfand had used most of the other main
lines as well, but this time he follows the
example of Kasparov.
6.lLlc3 d5 7.cxd5 cxd5 8.lLle5 .if5
Against the main line of 8 . . . e6 Karpov scored
four wins and seven draws without a loss. He
never defeated Kasparov in that line, but his
score against lesser players was impressive.
Gelfand subsequently switched to this move
in 1 995.
Karpov had never faced the text move before.
It occurred in a game between Botvinnik and
Smyslov, but is still regarded as a sideline.
9.0-0 lLl e4 10 ..ie3
Botvinnik exchanged on e4 here and the
game soon ended in a draw. Instead Karpov
develops and maintains the tension.
10 ... lLlxc3 1 1 .bxc3 lLl c6 12.�b3
Karpov deviates from the play of Larsen,
who preferred 1 2.lt'lxc6 bxc6 1 3.W1'a4 and went
on to win a long game against Benko in 1 963.
12 ... e6 13.lLlxc6 bxc6
8
7
6
5
I Game lJ I
4
Anatoly Karpov - Boris Gelfand
3
2
Linares 1 99 1
l .d4 lLlf6 2.c4 g6 3.lLlf3 oig7 4.g3
Karpov does not want to play a sharp King's
Indian where his opponent may be able to
create a strong attack.
1
a
b
c
d
e
f
g
h
14.�a3
Karpov takes away the a5-square from the
enemy queen, while also restraining the . . . c5
Anatoly Karpov - Boris Gelfand
199 1
advance. Despite Karpov's great strategic vision,
I suspect that even he did not anticipate how
important the latter point would eventually
become.
14 .. J:�e8
Black could have tried 1 4 . . . Wb6!? intending
to play on the b-file with . . . l'!fb8 followed by
... Wb2 or . . . Wb5 .
15.if4 e5
This frees Black's position to some extent;
nevertheless the potential power of the g2bishop increases, as the d5-pawn is less solidly
supported.
16.he5 Le5 17.dxe5 l'!xe5 1 8.e3 ie4
Gelfand exchanges the g2-bishop. It was
also possible to build his position without
simplifications, for instance: 1 8 . . . Wb6 1 9.1'!fc 1
2:d8 20.c4 dxc4 (20 . . . d4!?) 2 l .l'!xc4 c 5 22.l'!ac 1
2:d2 Black i s very active.
19.1'!fdl �f6
1 53
Ftom this point on, the future super­
grandmaster does not push any of his pawns
for another thirty one moves.
21 .h3!
Naturally Karpov has no intention of
allowing the pawn to advance to h3, where it
will create all sorts of threats.
2 1 ..J:�b8 22J3adl
Taking the pawn was our of question as
the black took would take up a dominating
position on the second rank. Therefore Karpov
prepares to respond to . . . l'!b2 by putting a rook
on d2.
22 ...Lg2
Gelfand decides to simplifY. A simple
improving move such as 22 . . . <;t>g7 would have
been risky, as White can gobble the a-pawn:
23.Wxa7 l'!b2 24.l'!4d2 l'!ee8 (23 . . . l'!e7 25 .Wd4
White consolidates his extra pawn) 24.Wc5
Black has some compensation, but he will have
to fight for a draw.
Defending the a7-pawn with 22 . . . l'!b7!? looks
like the most straightforward way for Black
to hold the position: 23.c4 l'!f5 24.l'!4d2
(24.f4 ixg2 2 5 .ct?xg2 dxc4 26.e4 l'!fb5
Black is very active, and White's kingside is
too open for him to have any real winning
chances.) 24 . . . ixg2 2 5 .ct?xg2 We6 (25 . . . dxc4
26.e4 l'!fb 5 27.l'!d6 We5 Black is not worse.)
26.cxd5 We4t 27.ct?h2 l'!xd5 Black will have
an isolani, but the chances of White winning
it are slim.
a
b
c
d
e
f
g
h
20J:�d4!
Karpov puts a bit of pressure on the e4bishop, but his main idea is to prepare c4 in
order to leave Black with a pawn weakness.
20 ... h5
23.@xg2
Karpov recaptures the bishop, while
conveniently making the kind of small
improving move which he liked to make
anyway.
The Prime Years
1 54
Gelfand keeps on exchanging but this is not
the best way of doing it. The problem with
exchanging queens in such positions is that the
opponent gets the opportunity to use his king
freely.
Swapping the rook by 23 . . . l'!e4! would have
enabled Black to get more active prospects
for his remaining pieces: 24.c4 (24.l'! 1 d2 l'!b 1
25 .Wi'xa7 h4 26.Wi'd7 [26.g4 l'!xe3] 26 . . . l'!xe3!
27.l'!f4 Wi'e6 28 .Wi'd8t Wi'e8 Black holds easily.)
24 . . . l'!xd4 25.l'!xd4 dxc4 26.Wi'c3 (26.l'!xc4
l'!b2 27.l'!f4 Wi'e5 Black is very active.)
26 . . . c5 27.l'!xc4 Wi'xc3 (Or 27 . . . Wi'c6t 28.e4
l'!c8 and Black is safe.) 28.l'!xc3 l'!b2 29.l'!a3
c4 Black can exchange all the queenside
pawns and reach a theoretically drawn
ending.
24.V;¥xe7 :gxe7
a
b
c
d
e
f
g
h
25.:g 1 d2
Karpov covers the second rank and defends
the a2-pawn, which he needs to keep in order
to preserve realistic winning chances.
25 :gb6
Benko recommended 25 . . . l'!eb7!? in order
to go after the a2-pawn as quickly as possible.
Black's most effective drawing plan is to get
rid of the queenside pawns. 26.c4 (Also after
.•.
26.�f3 l'!b2 27.l'!xb2 l'!xb2 28.l'!a4 l'!c2
Black holds.) 26 . . . dxc4 27.l'!xc4 l'!b2 28.l'!xb2
(28.l'!dc2 l'!xc2 29.l'!xc2 l'!b6 30.�f3 �f8
3 1 .�e4 �e7 32.�d4 �d6 Black is not even
slightly worse.) 28 . . . l'!xb2 29.l'!a4 c5 Black will
simplifY to a drawn ending with four pawns
versus three on the kingside.
26.e4! dxe4 27.:gxe4 �g7
27 . . . l'!e6? Defending the rook from the side
might seem like a good idea, but it gives up
the seventh rank. 28.l'!d7! l'!a6 29.l'!f4 l'!xa2
30.l'!fXf7 l'!xe3 3 1 .l'!g7t �f8 32.l'!df7t �e8
33.l'!b7 1t is possible that Black can survive this
position with perfect defence, but the situation
would have to be really desperate to try
it.
27 . . . �f8!?
Activating the king using this alternative
route deserved attention. It is not easy for
White to make progress.
28.e4!?
After 28 .g4 hxg4 29.hxg4 l'!e5 30.l'!d7 :ge7
Black holds.
If28.�f3 :ge5 29.:ga4 (29.:gd7 :gf5t) 29 . . . a5
Black should be able to hold this position,
as his rooks are much more active than in
the game.
28 . . . l'!e6 29.f4
29.:gd7 :gb2! is good enough.
29 . . . �e7 30.�f3 l'!d6
Black is still worse, but his drawing chances .
are higher than in the game.
28.:gde2 :ge7
From this square the rook defends all of
Black's weaknesses, but it becomes uncomfort­
ably passive.
28 . . . :ge6! was a better choice, for instance
29.l'!a4 :ge7 30.�f3 �f6 and Black should be
able to draw, although his position remains a
bit unpleasant.
155
Anatoly Karpov - Boris Gelfand
199 1
Black threatens to mobilize the c-pawn.
33.�d4t
33.gS? cS Black is fine.
33 . . . �e6 34.�cS
Once again 34.gS ? c5! frees Black's position.
34 . . . �d7
Black should exchange one of his opponent's
active rooks.
a
b
c
d
e
f
g
h
29.g4!
Karpov starts gaining space and his pawns
will eventually suffocate his opponent. With
the text move he gently puts a finger on his
opponent's throat, in a manner than can hardly
be felt. Later when the rest of the fingers follow,
Karpov's iron hand will crush him to death.
29 ...hxg4 30.hxg4 �f6
Black should have preferred 30 ... g5! to hold
back his opponent's space-gaining operation.
His position is not much fun, but this was
the lesser evil. 3 1 .�cS (3 1 . �g3 �f6 32.�cS
Ela6) 3 1 . . .f6 32.e4 (32.�f3 �a6 33.�e4 �e7t
34.�fS �a4) 32 . . . �e7 33.f3 �e6 Black has
good drawing chances.
There was another way to fight: Gelfand
suggests 30 . . . cS! to sacrifice a pawn in return
for activity. 3 1 .�xcS �xcS 32.�xc5 �a6 33.�c2
ElaS Black has decent drawing chances.
a
b
c
d
e
f
g
h
3S .�eSt!
If 3S .�xd7 �xd7 36.�aS a6 37.�f4 �d6
Black's king becomes active.
3S . . . �xeS 36.�xd7 �e6
There is also 36 . . . cS!? 37.�xf7 c4 (37 . . . �c6
38.f4t) 38.�c7 �dS when Black has good
drawing chances thanks to his strong passed
pawn.
37.�xa7 c5 38.aS �b2!
From the second rank the rook prevents the
white king from approaching the c-pawn.
3 1 .�g3 �e6 32.a4
Threatening as , when the reply . . . �a6 can be
met by �cS when the black rook is in a cage.
32 ... �d7?!
This passive move gives Karpov time to
improve his position. Much more resilient
was:
32 . . . �d6!
a
b
c
39.�c7 �c2 40.a6
d
e
f
g
h
The Prime Years
156
40.'tt> f3 c4 4 1 .a6 Ei:a2 42.a7 c3 leads to the
same outcome.
40 . . . Ei:a2 4 1 .a7 c4
The queenside pawns will soon be elimi­
nated, and Black can hold the resulting rook
endgame with two pawns versus three.
tries to cage the rook with 35 .Ei:c5 Ei:d7 36.a5?
Ei:d5! Then Black breaks the grip and solves his
problems.
However, White can improve with
36.Ei:dc4! c;t>d6 37.Wf4 when he maintains his
advantage.
35J:�c5! �b6
This costs time, but the rook was in danger
of being paralysed. For instance, 35 . . . Ei:d7?
is refuted by 36.a5! Ei:xd4 37.exd4 For the
moment Black is defending all his weaknesses,
but the white king can infiltrate easily:
37 . . . We7 38.c;t>f4 f6 (If 38 . . . We6 39.Ei:e5t
c;t>d6 40.c;t>e4 White pushes his pawn to f6 and
wins.) 39.gxf6t c;t>xf6 40. c;t>g4 c;t>f7 4 1 . c;t>g5
c;t>g7 42.f4 c;t>f7 43. c;t>h6 c;t>f6 44.Ei:g5 White is
winning.
a
b
c
d
e
f
g
h
33.g5!
Karpov gains more space and fixes the f7pawn. More abstractly, endgames can often
turn into pawn races, so advancing a pawn
closer to the promotion square is useful in
itself. Nevertheless 33. c;t>f4! was also strong.
33 J�a6?
Gelfand wants to become more active by
means of . . . c5, but he probably overlooked the
way in which Karpov counters his idea.
Black should have taken advantage of the
fact that the c6-pawn is no longer hanging in
order to free his rook: 33 . . . Ei:b l ! 34.Wf4 Ei:d l
Black still has decent drawing chances.
•.
34.Ei:d4t!
With this check and the next move Karpov
keeps the c-pawn fixed, which is a vital part of
his success.
34 'it>e8
According to Gelfand, Karpov mentioned
34 . . . c;t>e6!? which sets a small trap. If White
.•.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
36.'it>f4!
Karpov improves his king; soon its
domination will become unbearable for
Black. Instead 36.Ei:e5t was weaker because of
36 . . . Ei:e7.
36 �d7?
This final mistake proves fatal, as the rook
exchange allows White's king to take up a
dominating position. Generally four-rook
endgames give the weaker side better drawing
chances than two-rook endgames, as the
•••
1 99 1
Anatoly Karpov - Boris Gelfand
I S7
stronger side will have fewer opportunities to
use the king. Black could still have resisted,
although by now he did not have much margin
for error.
36 ... �e7? allows a quick finish: 37.l"i:eSt! �f8
38.l"i:d8t �g7 39.l"i:ee8+Activating the rook does not generate enough
play: 36 . . . l"i:b2 37.f3 l"i:b3 38.l"i:eSt l"i:e7
39.l"i:de4! l"i:xeS 40.l"i:xeS t �d7 4 1 .l"i:aS l"i:b7
42.�eS The position can be compared to that
reached in the game. White's dominant king
should be enough to decide the game.
36 ... �f8!?
The waiting approach looks like Black's best
chance. It is not easy for White to make
progress.
37.e4!?
White improves his pawn structure a bit.
Alternatives are not fully convincing:
37.l"i:d6 l"i:b4t 38.'tt> eS l"i:e7t 39.�f6 l"i:e8
40.l"i:cxc6 l"i:be4 Black is alive.
37.�eS l"i:e7t 38.<tt> d6 l"i:b2 39.l"i:xc6 (39.l"i:f4
l"i:bb7) 39 . . . l"i:xf2 40.l"i:c8t l"i:e8 4 1 .l"i:xe8t
�xe8 42.l"i:dS f6 (another idea is 42 . . . l"i:b2
43.l"i:bS l"i:d2t 44.c;f;>eS <tt> f8 and the game
goes on) 43.c;f;>e6 fxgS 44.l"i:xgS l"i:c2 Black
has reasonable drawing chances.
37 ... l"i:b2 38.f3 l"i:b3 39.l"i:eS
39.l"i:d8t c;f;>e7 does not help White.
39 ... l"i:b8
White will press and Black will suffer for a
long time to come, but the second player has
realistic chances to survive.
a
b
c
d
e
f
g
39.£4!
Having put his rook and king in dominating
positions, Karpov gains more space and gets
his pawns closer to the eighth rank.
39 .. J:�b4 40J3a5 13b7 41 .e4 13c7
In his excellent book My Most Memorable
Games, Gelfand mentions the following
variation, which he attributes to Karpov:
4 1 . . . l"i:d7 42.l"i:cS l"i:d6 43.fS gxfS 44.exfS f6t
4S.gxf6t l"i:xf6 46.l"i:aS White wins.
42.13c5
Karpov brings the rook back in order to
invade along the b-file.
42 ... 13c8 43.13c3 13e8 44.13c4 13c8 45.13b4
13c7
37Jhd7!
Naturally Karpov does not pass up an
opportunity to invade with his king.
37 c;!;>xd7 38.c;!;>e5! c;!;>e7
Black can only temporarily stop the further
invasion.
•••
a
b
c
d
e
f
g
h
The Prime Years
158
46.a5!
Karpov advances another pawn, creating the
possibility of a6 followed by 1':!:b7, as well as
getting closer to the promotion square.
46 @d7 47. .E:b3
Karpov decides to play for zugzwang. 47. c;t>f6
would have won as well.
.•.
47 @e7 48.a6!
Karpov makes every possible improving
move before eventually launching his winning
plan.
.•.
48 @d7 49.@f6
The king gets even stronger.
•••
49 @e8 50 .E:h3 1':!:d7
•••
•
a
b
c
d
e
f
g
h
5 1 .f5!
Finally it is time to win the game by creating
an unstoppable passed pawn.
5 1 ...gxf5 52.exfS e5
After thirty one moves Gelfand finally pushes
a pawn, but in the meantime Karpov earned a
massive head-start in the race.
53.1':!:c3 1':!:e7 54.g6 fxg6 55.fxg6 @d7 56.g7
.E:e8 57 . .E:g3!
1-0
Taking on c5 was also good enough, but
Black's rook could have resisted against the
queen for a while, whereas the text move
wins instantly. Gelfand was not at his best,
nevertheless it was a fantastic endgame display
by Karpov.
In Round 7 Timman played the 4.f3 line
against the Nimzo-Indian, and Karpov had to
work hard to draw. In the next round Karpov
met Kasparov, who equalized with a Semi­
Slav and drew comfortably. Then in Round 9
Speelman executed a strong pawn sacrifice in
the middlegame after a 4.e3 Nimzo-Indian,
and Karpov found himself with an exposed
king and the problems proved too much for
him.
After that loss Karpov got a nice positional
edge against Salov in a Slav, but uncharacter­
istically allowed a piece sacrifice which led to
perpetual check. But then in Round 1 1 he
managed to get the better of Kamsky in a long
manoeuvring game on the black side of the
Hubner Variation of the Nimzo-Indian. In the
next game he held the black side of a 4.'1M1'c2
Nimzo-Indian against Mikhail Gurevich in
a game that was always roughly balanced. In
the final round Karpov clawed his way back
to fifty percent against Beliavsky on the white
side of a Queen's Gambit Declined.
Despite his late resurgence, Karpov's final
score of 6Y2/ l 3 must have been disappointing,
and losing four games in a single tournament
was something that had never happened to him
during his entire professional career. Karpov
was approaching forty years of age, so energy
levels may have been a factor; furthermore his
openings did not carry enough bite against the
new wave of top grandmasters.
By the way Ivanchuk finished in first place
half a point ahead of Kasparov, who for the
first time in almost a decade had to be content
with less than first place or at least equal first
in a tournament.
199 1
Anatoly Karpov - Jan Timman
Karpov's next tournament was the Euwe
Memorial in Amsterdam, where he started
with three draws against Hjartarson, Korchnoi
and Ljubojevic. In Round 4 Karpov achieved
no advantage against Mikhail Gurevich's
Leningrad Dutch, but Gurevich made a serious
mistake in the middlegame and was unable
to recover. In the next round Karpov drew
solidly on the black side of an English against
Salov.
In Round 6 Karpov met Kasparov, and the
two players followed Game 20 of their New
York/Lyon match, a Zaitsev which resulted in
a crushing victory for White. This time Karpov
was ready with a novelty and got the better of
a hugely complicated position. But later the
endgame maestro made some shocking errors
which enabled Kasparov to draw the endgame
a piece down.
After this terrible disappointment, Karpov
had to face Jan Timman. Since their 1 979
meeting in Montreal (Game 52 of the first
volume) , Karpov established a heavy plus score
against the Dutchman by winning sixteen
games, drawing thirty one and losing just
two.
I Game l41
Anatoly Karpov Jan Timman
-
Amsterdam 1 9 9 1
I.d4 d5 2.c4 dxc4
Timman seldom played this move, but he
did it twice against Karpov. Considering the
number of times these two players faced one
another in tournaments (sixty five times, not
including matches) , it is natural that such an
adventurous player as the Dutch grandmaster
would have tried a variety of different openings
against Karpov.
3.e3
1 59
3.e4 was Karpov's usual choice, but the text
move leaves little room for surprises. Karpov's
overall score against the Queen's Gambit
Accepted is ten wins, thirteen draws and no
defeats.
3 ... e5 4 ..L:c4 exd4 5.exd4 lLl f6 6.ttJf3 i.e7
Timman was probably not expecting
Karpov to employ this variation, and thus
found himself in relatively unfamiliar territory,
which is why he plays it safe. 6 . . . id6 is more
active.
7.0-0 0-0 8.h3
Karpov makes sure Black will not be able to
pin the knight.
8 ... lLl bd7 9.ttJc3 ttJ b6 10.i.b3 c6
It is hard for Black to play without this
move.
l 1 .E:el ttJfd5
1 1 . . .ttJ bd5 is more common. One example
continued: 1 2.ig5 ie6 1 3.ttJe5 Ele8 1 4.Elcl
ttJ d7 1 5 .ixe7 Elxe7 1 6.ttJe4 f6 1 7.ttJd3 ttJ f8
1 8 .�f3 �c7 1 9.ttJg3 Eld8 20.ttJc5 White is
somewhat better, I. Sokolov - Hubner, Wijk
aan Zee 1 996.
12.ttJe4!?
Karpov avoids the exchange of knights,
which would relieve the congestion in Black's
camp.
12 ...i.f5
Timman introduces a novelty. It is especially
interesting since he had reached this position
previously from the opposite side of the
board. That game continued: 1 2 . . . Ele8
1 3.id2 if5 1 4.ttJg3 ( l 4.ttJe5!?) 14 . . .ie6
1 5 .ic2 ttJ d7 1 6.a3 ttJ f8 1 7.id3 g6 Black
has equalized although White later went on
to win in Timman - Panno, Mar del Plata
1 982.
The Prime Years
160
once said in an interview that the mistakes
that could sometimes be found in Kasparov's
analysis betray the fact that he had not really
done the analysis. But then who is responsible
for this piece-losing blunder?
a
b
c
d
e
f
g
h
13.lLle5!
If Black could safely bring his as-rook into
play while preserving his light-squared bishop,
then he would be fine. Therefore Karpov acts
quickly to prevent this from happening.
1 3.tLlc5 could be met by 1 3 .. .'IWc7 1 4.tLle5
ElaeS! (but not 1 4 . . J:'!:adS?! 1 5 .Wf3!) when
Black has a good position as he threatens to
take on c5 .
13 tlJd7
Timman wants to exchange pieces in the
centre, but it costs him valuable time.
a
b
c
d
e
f
g
h
1 5 .�d2! �xe4 1 6.�xa5 �xf3 1 7.tLlxd7 White
wins a piece at least.
14 . . . �xe4 can be compared with the note
to Black's previous move. Black gets a
reasonably stable position, but he is passive
and may suffer on the light squares later in the
game.
•••
13 . . .�xe4 Giving up the bishop would ease
the congestion in Black's position, but White
will dominate the light squares. 1 4.Elxe4 tLl f6
1 5 .Ele2 (White is also better after 1 5 .Ele3 tLl fd5
1 6.Wf3) 1 5 . . .�d6 ( l 5 . . . tLl fd5 1 6.Wfd3) 1 6.�g5
Black will have a hard time breaking White's
grip.
14.WfB!
Karpov's previous strong move enables his
queen to develop with gain of tempo.
14 lLlxe5
Timman exchanges the unpleasant knight.
1 4 . . . Wfa5 ?? was mentioned in Karpov's
annotations without any evaluation. Karpov
•••
14 . . . �e6!? may have been the best chance,
although Black does not equalize here either:
1 5 .�d2 Wc7 ( l 5 . . . tLl xe5 1 6.dxe5 a5 1 7.Elad l a4
[ l 7 . . . Wb6 I S.Wg3 ct?hS 1 9. tLl d6] I S.�c2 Wfb6
1 9 .�c l White has good attacking prospects,
for instance 1 9 . . . tLl b4? loses to 20.tLlf6t! �xf6
2 1 .�xh7t ct?xh7 22.exf6+-.) 1 6.tLlxd7 (White
can also play on the kingside by means of
1 6.Wfg3 ElfeS 1 7 .�h6 �fS I S .Elac l , with some
pressure.) 1 6 . . . Wxd7 1 7. tLl c5 �xc5 I S.dxc5
White maintains an edge thanks to his two
bishops.
1 5.dxe5
White reduces the number of pawn islands
and obtains a potential knight outpost on d6,
while in some lines the e5-pawn might play a
role in a kingside attack.
15 ...ig6
Timman opts to fortifY his kingside. The
alternative was 1 5 . . . ieG I G.ic2 (or I G.id2)
16 .. .'&bG 1 7.'1W g3 rJih8 1 8.lLlg5 when White
has a promising kingside initiative.
The knight is powerful here, but compared
with the game Black has more chances to
live with it.
1 8 . . . Ei:ad8 1 9 .Ei:ad l rJih8!
Black prepares to free himself with . . . fG.
a
a
b
c
d
e
161
Anatoly Karpov - Jan Timman
1 99 1
f
g
h
16.if4!
Karpov develops fluently. I G.Ei:d l would
allow Black to sacrifice a pawn for reasonable
compensation: I G . . . 1M! c7 1 7.ixd5 cxd5
18J�xd5 1M!cG ( 1 8 . . .Ei:fd8 1 9. 1Ll dG) 1 9.ie3
Black may not have full compensation,
but compared with the game continuation
his position would be somewhat easier to
handle.
16 ...�a5?
Timman faces a difficult choice regarding
which pieces to exchange and how to go about
exchanging them. The text move turns out to
be a poor solution.
16 . ..ltJxf4!
This exchange would have given Black better
equalizing chances.
17.'&xf4 1M!c7
Black should resist the temptation to enter
an opposite-coloured bishop position with
1 7 . . . ixe4?! 1 8.Ei:xe4 when f7 is weak and the
b3-bishop is tremendously powerful.
1 8.tiJ d 6
b
c
d
e
f
g
h
20.1M!g3!?
This is the most interesting move to analyse,
although it is hard to say if it is objectively
best.
20.Ei:d2 fG 2 1 .Ei:ed l fxe5 22.1M!xe5 Ei:d7 Black
is okay.
2o. lLl f5!? ic5 White can maintain some
pressure with 2 1 .g3 or 2 1 .lLlh4, but Black
has decent chances to hold.
20 . . . fG! 2 1 . lLl f7t!
It almost feels wrong that this incredible
tactical blow does not win by force.
2 1 . . .ixf7
2 1 . . .Ei:xf7 22.eG! is difficult for Black.
22.exfG 1M!xg3 23.fxe7
a
b
c
d
e
f
g
h
The Prime Years
162
23 . . . Wi'xf2t!
By returning the queen Black simplifies to
an endgame with an extra pawn.
After 23 . . . ixb3 24.:!'hdB Wi'xf2t 25 .�h2
Ei:gB 26.Ei:xgBt (if 26.eB=Wi' Wi'f6! Black is
okay) 26 . . . ixgB 27.eB=Wi' Wi'xb2 2B.Wi'bB!
White has good winning chances.
24.i>xf2 ixb3t 25 .�g3 ixd 1 26.Ei:xd 1 Ei:deB
27.exfB=Wi't Ei:xfB 2B.Ei:d7 Ei:bB
Despite the extra pawn Black is passive and
his position remains worse. Nevertheless he
should be able to hold the endgame.
17.lLl d6! ,bd6
The knight was tremendously strong on d6,
and Black would have had to exchange it sooner
or later. If 17 . . . Wi'b4 then 1 B.ih2 maintains a
clear advantage, while after 1 7 . . . Ei:abB 1 B.ig3
h6 1 9 .Ei:ad 1 Black is also in trouble.
l S.exd6
The strong passed pawn is too much for
Black to handle.
l S .. JUeS?
Timman commits a tactical error in a
strategically lost position. 1 B . . . Ei:adB was better,
but after 1 9.Ei:ad 1 or 1 9.ig3 White is in full
control and should win.
8
Karpov finds a forced win which requires
strong calculating ability.
19 ... 13xeS 20.,bd5!
Timman once wrote that one of the
advantages of having rwo bishops is that the
player who possesses them can often exchange
one of them for an enemy knight at a time
that suits him. Karpov's last move illustrates
that point perfectly. After removing the knight
White gets the use of the c7-square and his
d-pawn grows decisively strong.
20 ... cxd5
20 . . . Wi'xd5? loses trivially: 2 1 .Wi'xd5 cxd5
22.d7 Ei:dB 23.Ei:e1 f6 24.Ei:e7 �fB 25 .id6+2 1 .d7!
By utilizing the powerful d-pawn and
the weakness of Black's back rank, Karpov
demonstrates a forced win.
2 1 ...13e7
2 1 . . .Ei:dB
This loses as well.
22.Wi'e3!
The most clinical route to victory.
22 . . . h6
22 . . . Wi'b5 23.ic7 Wi'xd7 24.ixdB Wi'xd8
25.Ei:d1 wins.
22 . . . f6 23.Wi'e7 is also hopeless for Black.
7
6
5
4
3
2
1
a
19.13xeSt!
b
c
d
e
f
g
a
h
23.Wi'eBt!
b
c
d
e
f
g
h
199 1
Anatoly Karpov - Jan Timman
This is the simplest win.
23.l"kl is also good enough: 23 . . .Wla4
(23 . . . if5 24.WleBt mh7 25 .Wlxf7 ixd7
26.ie5 wins; 23 . . . Wlb5 24.Ek7 if5 25 .Wle5
ixd7 26.Wle7 Black loses his bishop.) 24.l"k7
if5 (24 . . . Wh7 25 .Wle7) 25 .ixh6! gxh6
26.\Wxh6 ig6 27.Wlg5 l"!xd7 2B.b3! Wlb5
29.a4 White diverts the queen and wins.
23...Wh7 24.b4
24.l"!c l is also good enough.
24 . . . \Wb6 25 .l"!cl if5
25 . . .f6 26.Wle7+26.ic7
White wins.
22J'kl !
Exploiting the weakness o f Black's back
rank.
22 ... l"!xd7 23.l"!e8t l"!d8
1 63
27.�xd5 h6 28.�xb7 :gel t 29.c;t;h2 :ge2
30.�xb6 :gxal 3 1 .�d4!
1-0
In the final two rounds Karpov drew with
Van der Wiel and Short, in two hard-fought
but always roughly equal games. His final
score of 5 Y2i9 enabled him to share third place
with Kasparov, half a point behind Salov and
Short.
Karpov's next event was a short four-game
match against Simen Agdestein, which took
place in the Norwegian town of Gjovik. In
the first game Karpov got the upper hand in
a 3.ltJc3 ltJ c6 French, but Agdestein managed
to reach a pawn-down queen ending with
good drawing chances. He defended well
and reached a drawn position, when Karpov
blundered badly and lost. In the second game he
bounced back with a good win in the 4.g3 ia6
Queen's Indian. In the third game Agdestein
repeated the same line of the French and was
able to draw a long game. In the final game the
Norwegian grandmaster employed an unusual
line against the 1 .c4 e5 English. Karpov got
into trouble but then fought back to reach a
winning endgame with an extra pawn, but he
squandered his advantage and drew both the
game and the match. Karpov was not at his
best, and in particular his endgame play was
well below his normal level.
***
a
b
c
d
e
f
g
h
24.b4!
This cute move is the only way to divert the
queen successfully. If 24.Wlc3? Wlb6 25 .ie3
(or 25 .Wlc5 Wlxc5 26.l"!xdBt WlfB) 25 . . . Wld6
26.\We5 Wld7 Black stays alive.
24 Wfb6 25.i.e7 :gxe8 26.hb6 axb6
Despite being close to material equality,
Black has no real drawing chances as he will
soon be a pawn down on the queens ide.
..•
The early nineties was a period in which many
of the next generation of super-grandmasters
were breaking through to the top echelons of
world chess. It was therefore fitting that in
the quarter-finals of the next candidates cycle,
the former champion Karpov would do battle
with the future champion Viswanathan Anand
in an eight-game match, which took place in
Brussels.
The Indian had the white pieces in the first
game, and chose 2.d3 against the Caro-Kann.
The Prime Years
1 64
The queens were exchanged early and Karpov
equalized, then he took over the initiative after
Anand made a bad decision. Karpov continued
to play well, but then he erred and allowed
Anand to salvage a draw. Game 2 was a mirror
image of the first; Anand equalized in a Semi­
Slav and later took over, but failed to make the
most of his advantage and allowed Karpov to
escape with a draw. In Game 3 Anand tried
a different sideline against the Caro-Kann.
Karpov equalized but then misplayed his
position and got into serious trouble. Anand
won two pawns and was clearly winning,
but then went wrong and Karpov narrowly
managed to draw.
It is rare that one can afford to waste two
golden opportunities against a player of
Karpov's quality, and sure enough in the next
game the former champion rediscovered his
form.
I Game l� I
Anatoly Karpov -Viswanathan Anand
Brussels (4) 1 9 9 1
l .d4 d 5 2.c4 c6 3.lLIe lilf6 4.lilc3 e6 5.e3
lil bd7 6.Wfc2
This has been Karpov's most common choice
against the Semi-Slav. He has used it to score
an impressive thirteen wins, sixteen draws and
one loss.
6 ...i.d6 7.i.e2
This variation became popular in the late
eighties.
1 3 .g3 E!:e8 1 4.a3 i.b7 1 5 .dxe5 ttJxe5 1 6.i.g5
ttJxf3t 1 7.i.xf3 i.e5 1 8.i.xf6 gxf6 1 9.i.g4
White got an edge and went on to win, Karpov
- Kramnik, Monte Carlo [rapidl I 998.) I l .h3
b5 1 2 .i.d3 c5 1 3 .ttJe4 c4 1 4.ttJxd6 Wxd6
1 5 .i.e2 i.b7 1 6.i.d2 E!:fc8 1 7.E!:dc l ttJe4
1 8.i.a5 f5 Black had an easy game, Karpov ­
Anand, Brussels (2) 1 99 1 .
Another important option is:
9 . . . e5
a
b
c
d
e
f
g
h
The most popular move has been 9 . . . b5, bur
that is another story.
Let me show you a fantastic and revolutionary
game which played a key role in the
popularity of this whole variation.
1 0.h3 exd4
1 0 . . . We7 has been the most popular move,
although 1 0 . . . h6!? has scored highly.
I I .exd4 ttJ b6 1 2.i.b3 h6?
After the superior 12 . . . ttJ bd5 Black is only
slightly worse.
7... 0-0 8.0-0 dxc4 9 ..hc4 Wfe7
Anand deviates from 9 . . . a6 which worked
well for him in the second game of the match.
In the late nineties Anand returned to this
move against Karpov and made two draws.
1 O.E!:d l We7 ( l O . . . b5 1 1 .i.e2 Wc7 1 2.e4 e5
a
b
c
d
e
f
g
h
1 99 1
165
Anatoly Karpov - Viswanathan Anand
1 3.ixh6!! gxh6 1 4.Wg6t 'ithS 1 5 .Wxh6t 4:J h7
1 6.liJe4 ie7
1 6 .. .f6 1 7.4:Jxd6 Wxd6 l s.ic2 :9:fl 1 9.1iJg5
wins.
1 7.liJf6! if5
1 7 . . . ixf6 l s.ic2 is the end.
1 8.liJh5 if6
1 8 . . . :9:gS loses to 1 9.1iJe5.
1O.h3
In the decisive final game of the match
Karpov switched to a different move: 1 0.a3!?
e5 1 1 .h3 .tc7 12 ..ta2 h6 1 3.liJh4 :9:eS 1 4.4:Jf5
WfS 1 5 .liJb5 .tbS 1 6 . .td2 a5 1 7.dxe5 .txe5
l S.f4 .tbS 1 9.4:Jc3 :9:dS
6
4
3
2
a
1
a
b
c
d
e
f
g
h
1 9.1iJg5! ixd4 20.4:Jxh7 ixh7 2 1 .:9:ad 1 f5 ?
2 1 .. .c5 22.:9:xd4 Wxd4 23.:9:d 1 We5 24.f4
Wlxb2 25.4:Jf6 wins.
The last chance was 2 1 . . .:9:gS! 22 . .tc2
(22.ixfl WfS!) 22 . . . :9:g6 23 . .txg6 fxg6
24.liJ f4 Wf6 25.:9:xd4! Wxd4 26.:9:e l ! Wg7
27.Wg5 White has excellent compensation
for the pawn, but the fight continues.
b
c
d
e
f
g
h
20 . .te l ! Karpov finds a nice way to improve
the bishop. 20 . . . 4:J h7?! 2 1 .ih4 liJ df6 22.:9:ad 1
:9:xd 1 23.:9:xd 1 .te6? A blunder i n a difficult
position. 24 ..txe6 fxe6 25 .Wb3! WeS 26.4:Jxg7!
Wfl 27.4:Jxe6 Black soon had to resign in
Karpov - Anand, Brussels (S) 1 99 1 .
8
7
6
5
4
3
2
1
a
a
b
c
d
e
f
g
h
22.:9:fe l !+Smagin - Monin, Pinsk 1 9S6. Black has no
good defence against :9:xd4 followed by :9:e7. In
the game he gave up his queen but soon had
to resign.
b
c
d
e
f
g
h
IO ... a6
In the next Meran game Anand switched
to 1 0 . . . c5 with a much better result: 1 1 .dxc5
.txc5 1 2.e4 .td6 1 3. liJ d4 liJe5 1 4 . .tb3 id7
1 5 . .te3 4:Jg6 1 6.:9:ad 1 :9:fdS 1 7. liJ f3 .tc6 Black
obtained a lovely position and went on to win
The Prime Years
166
a long game, Karpov - Anand, Brussels (6)
1 99 1 .
Since that time 1 O . . .e 5 has become the main
line.
1 I .e4 e5 12J:�dl b5
This move diverts the bishop from the f7pawn, but gives White a target on which to
bite.
a
b
c
d
e
f
g
h
1 5.a4!
Karpov wastes no time in undermining
Black's queenside structure. It will take some
time, but he never loses sight of this goal.
a
b
c
d
e
f
g
h
13.i.fI !
With this long retreat Karpov moves his
bishop out of the way of Black's potential
counterplay on the e-file. In certain positions
the knight may use the e2-square as well.
13 ... c5?!
The resulting change in
configuration favours White.
the
central
1 3 . . . ib7 1 4.dxe5 lLlxe5 1 5 .lLld4! is somewhat
unpleasant for Black.
It was worth considering improving the rook
with 1 3 . . . :1!e8!? or 1 3 . . . :1!d8!?
14.d5 c4
Anand follows up his previous move in a
consistent manner, hoping to develop queens ide
play. 1 4 . . . lLl h 5 was possible, although White's
plan involving a4 would be strong here too.
1 5 .. J�� b 8
1 5 . . . b4 1 6.lLle2 highlights one of the
advantages of the earlier bishop retreat to
fl instead of e2. Black's best looks to be
1 6 . . . lLl c5 1 7.lLlg3 c3!? 1 8 .bxc3 b3 1 9.�b 1 ,
but i t i s doubtful that his compensation for the
sacrificed pawn can be sufficient.
1 6.axb5 axb5
167
Anatoly Karpov - Viswanathan Anand
199 1
advance prematurely, before the second player
can consolidate with . . . lLl c5 and . . . �d7.
If 20 . . . h6 2 1 .�d2 White carries out his plan
unhindered.
17... b4 18.lLla4
Karpov mentions the alternative I B.lLlb5,
which entails an exchange sacrifice. 'The idea is
interesting, but after I B . . . WdB! 1 9.1Llxd6 Wxa5
20.Wxc4 Ei:aB 2 1 .�d2 �a6 22.Wxb4 Wa2 the
position is roughly balanced.
2 1 .tLlxc5 hc5 22J�al !
22.Ei:a4 runs into 22 . . . �d7! when the
exchange sacrifice does not bring any advantage:
23.Ei:xc4 �b5 24.lLlxe5 WeB 25.lLlc6 �xc4
26.Wxc4 lLlxe4 'The position is roughly equal.
18 ...Wd8 19.Ei:a7 b3!
1 9 . . . lLlb6 2o.lLlxb6 (Another option is
20.ie3 lLlxa4 2 1 .Ei:xa4 �d7 22.Ei:a7 c3 23.bxc3
bxc3 24.�d3 WcB and White is better here
as well.) 20 . . . Wxb6 2 1 .�e3 �c5 22.�xc5
�xc5 23.Ei:da l White can exploit his lead in
development, as after 23 . . . c3 24.Ei: 1 a5 b3
25.Wd3 Black is in trouble.
a
b
c
d
e
f
g
22 ... c3!
Anand
keeps
imagination.
playing
with
great
22 . . . Wd6
With this move Black defends his queenside,
but risks becoming too passive.
23.�d2
After 23.�g5 �d7 (23 . . . c3?! 24.bxc3 b2
25 .Ei:ab 1 is not helping Black) 24.lLld2
White exerts pressure on the c4-pawn, but
it is hard to say if it is enough for a serious
advantage.
23 . . . �a6 24.�c3
Karpov stops here, evaluating the position
as clearly better for White. 'The verdict is
more or less correct, but we can look a bit
further.
h
20.We2!!
With this great move Karpov intends to play
id2 followed by Wxc4, without allowing the
reply . . . Ei:b4.
Anand may have been expecting the immediate
20.Wxc4? Ei:b4! when Black gets good
counterplay as . . . lLlxe4 is coming next.
2o ... lLlc5!
Anand makes the best of his situation
and wittily creates play on the queenside.
a
b
c
d
e
f
g
h
24 . . . lLl d7 25.lLld2
25 .Ei:a5 Ei:fcB defends.
25 . . . Ei:fcB 26.Wf3 �b5 27.Ei:dc l �b6
White certainly has the advantage, but how
should he increase it?
The Prime Years
1 68
a
b
c
d
e
f
g
h
28.lMrg3!
28.tLlxb3 i.xf2t is not entirely convincing.
After the text move i.e2-g4 is a promising
plan, and White may also consider preparing
f4. Black faces a difficult defence.
a
b
c
d
e
f
g
h
24.Ei:d3!
Karpov brings the rook to the third rank
in order to eliminate the b3-pawn as soon as
possible. As long as Black's b3- and c2-pawns
remain on the board White can never relax.
The alternative was: 24.Ei:e l i.b7 (24 . . . Ei:e8
25.tLlc6 lMrb6 26.tLlxbB lMrxbB 27.lMrc4 lMrb6
2B.i.e3+-) 25 .lMrc4 lMrc7 26.tLlc6 lMrb6 27.i.e3
i.xe3 2B.Ei:xe3 i.xc6 29.dxc6 Ei:fcB 30J'ic3
White is better here too. He has an extra pawn
and his c6-pawn is also strong , but Black has
kept hold of his annoying queenside pawns.
a
b
c
d
e
f
g
h
23.liJxe5!
Karpov not only plays well strategically, but
also controls the tactical battle with expert
precision.
Less convincing is: 23.bxc3 b2! 24.i.xb2 Ei:xb2!
(After 24 . . . tLlxe4 25 .lMrxe4 Ei:xb2 26.Ei:d2 Ei:xd2
27.tLlxd2 i.xh3 2B.lMrxe5 White keeps his extra
pawn.) 25 .lMrxb2 tLlxe4 26.Ei:d2 tLlxd2 27.lMrxd2
e4 2B.tLle5 (If2B.tLlel e3! Black becomes active.)
2B . . . i.e6 29.c4 lMrd6 30.tLlc6 i.d7 Black has
decent compensation for the pawn.
23 ... c2
24...�e8?
Anand avoids the knight fork and keeps
both his bishops, but he has no time for it as
his queenside pawns are too vulnerable.
Sacrificing the exchange was not an option:
24 . . . Ei:eB? 25.tLlc6 lMrb6 26.tLlxbB Ei:xe4
(26 . . . tLlxe4 27.i.e3+-) 27.lMrf3+24 . . . i.d7!
Developing the bishop would have led
to an exciting fight, with a lot of tactical
possibilities on both sides. It would have
been hard even for these great players to
handle the ensuing positions correctly.
2 5 .lMrf3!
This looks like the best chance to preserve
1 69
Anatoly Karpov - Viswanathan Anand
199 1
White's advantage.
2S.<iJxd7 'lWxd7 26.ig5 Eib6 (26 . . . EifcB!?)
27.eS 'lWf5 2B .ie3 <iJ e4 The position is
complicated.
2S .l"k3 id4 (25 . . . ib5 26.'lWf3 'lWe7
27.Eixb3 'lWxeS 2B.if4 'lWd4 29.ie3 'lWxe4
30.'lWxe4 <iJxe4 3 1 .EixbS ixe3 32.fxe3 <iJ d2
33.Eicl <iJxfl Black should be able to hold.)
26.<iJxd7 <iJxd7 27.Eid3 'lWf6 The position is
double-edged.
25.tLlc6 :Sb6
2S . . . Eib7!? may have offered more stubborn
resistance than the game continuation, although
Black is still struggling: 26.eS (26.ie3 id6
27.Eid4 is also promising) 26 . . . <iJxd5 27.<iJ a5
Eid7 2B.<iJxb3 ib6 29.Eid2 White will soon
pick up the c2-pawn, but winning the game
will still require some effort.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
2S ... ibS
2S . . . EieB 26.<iJxd7! (26.if4 can be met by
26 . . . <iJxe4! [but not 26 . . . Eib4 27.Eixb3 Eixe4
2B.<iJxd7 'lWxd7 29.Eic3 Eixf4 30.'lWxf4 <iJxdS
3 1 .Eid3 when White keeps the material]
27.'lWxe4 f6 2B.d6 fxe5 when Black is by
no means worse.) 26 . . . 'lWxd7 27.Eixb3 Eixb3
2B.'lWxb3 <iJxe4 29.ie3 ixe3 30.'lWxe3 (30.
fxd g6) 30 ... 'lWxdS 3 1 .'lWe2 EicB 32.Eic l
White picks up the c2-pawn and has decent
winning chances.
26.Eixb3 ixfl 27.Eic3
27.c;t>xfl Eixb3 2B.'lWxb3 <iJxe4 29.<iJd3 ib6
If White has any advantage here, it is small.
27.<iJc6!? Eixb3 2B .'lWxb3 'lWd6 29.'lWxc2
reaches a complicated position in which
White's chances are slightly better.
27 ... ibS 2B.Eixc5 'lWe7 29.Eixc2 'lWxeS 30.if4
�xe4 3 1 .'lWxe4 <iJxe4 32.ixbB EixbB
White will press with his rook and two
pawns against Black's two minor pieces, but the
second player has reasonable drawing chances.
������
a
b
c
d
e
f
g
h
26.i.e3!
Neutralizing Black's strong bishop. Note that
26.eS? would be useless due to 26 . . . <iJxdS .
26 ... tLlxe4 27.i.xc5 tLlxc5 28.:Se3!
With this move White wins a huge tempo,
which gives him time to attack the b3-pawn.
28 'lWd7 29.'lWc4!
After this move Black's position is on the
verge of collapse.
.•.
29 ... :Sxc6!?
Anand finds the best practical try, but
ultimately it is insufficient.
29 . . . 'lWd6 loses without much of a fight:
30.<iJe7t (In his Chess Informant analysis
Karpov gave 30.Eic3 <iJ d7 3 1 .Eixb3, which
is presumably what he calculated during the
game. White is winning comfortably here,
but the main line is even more convincing.)
The Prime Years
170
30 . . . 'it>h8 3 1 .l':!:a8 White wins a piece and the
game.
33 . . . lt:l e4! 34.l':!:xd5 (34.'1Wb4? Wxc6 is not so
easy. The simplest win is 34.Wxc2!, although
Karpov does not mention it.) 34 . . . lt:lxc3
30.dxc6 Wdl
a
a
b
c
d
e
f
g
h
3 l J�el! Wd6 32.V9c3!
With these two cool-headed moves Karpov
decides the outcome of the game. By taking
away the d2-square and defending the e 1 rook, he enables the other rook to move along
the a-file.
b
c
d
e
f
g
h
35 .l':!:d3! (35 .l':!:c5 It:l a4! 36.l':!:b5 It:lxb2! 37.l':!:cl
[37.l':!:xb3?? It:l d l ! even wins for Black!]
37 . . . �e6 Black is still alive.) 35 . . . lt:l a2 36.l':!:xb3
�e6 (36 . . . cl =W 37.l':!:xc l It:lxc l 38.l':!:b8 is
winning for White.) 37.l':!:b6 c l =W 38.l':!:xcl
It:lxc l 39.c7 White should be winning the
ending, but the game continuation is a lot
more straightforward.
33,..i.b7
32,..V9d5
If 32 . . . �f5 33.We5! wins.
33.c7
Keeping a passed pawn on the seventh rank
is a useful insurance policy.
With 33.l':!:a5 White can also win the pawns,
but he must carry out the operation with care:
a
b
c
d
e
f
g
h
34J�a5
Now this move decides the game without
leaving any doubts.
34,..c!tJe4 35J3xd5
37J3xb3
a
b
c
d
e
f
g
h
c!tJxc3 36J3d3
c!tJ a2
Anatoly Karpov - Viswanathan Anand
199 1
171
Compared with the variation examined in the note to move 33, the attack on the b7-bishop
makes the win so much easier.
37...!c8 38.!c4!
Preparing to trap the knight.
38 ... g6 39.:aa3 cl=Wf 40.:axcl .!iJxcl
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
41 .:ac3
Karpov could have won by pushing the b-pawn, but the simplest way is to collect the trapped
knight.
41. .. :ae8 42.:axcl :ae7 43.!fl
1-0
In Game 5 Anand played Short's set-up in the Advance Caro-Kann. In an equal middlegame
Karpov played passively, and the Indian grandmaster outmanoeuvred him skilfully. But later
Anand missed several wins and the game was eventually drawn. Then in Game 6 Anand equalized
easily in the Semi-Slav and gradually outplayed Karpov. This time he made no mistake and
successfully converted his advantage in the endgame.
In Game 7 Karpov once again got into trouble in the Advance Caro-Kann, but he managed
to hold a tough endgame. This meant that everything would rest on the eighth and final game.
Perhaps affected by the tension, Anand misplayed the early middlegame in a Semi-Slav, and a
subsequent blunder sealed his fate. Thus Karpov advanced to the semi-final stage, but the match
was close and in several games he was struggling to draw.
Next in Karpov's calendar was the World Cup event in Reykjavik. He started with a fine win
over Speelman, involving a tactical combination followed by skilful conversion of his endgame
advantage in a position with queen and several pawns versus the Englishman's rook and two
bishops. In Round 2 he made a quick draw against Portisch. This was followed by a win over
Salov on the white side of a complicated Queen's Indian. In Round 4 Karpov was a pawn down
172
The Prime Years
against Beliavsky, but managed to hold the endgame thanks to his strong bishop pair.
Karpov won the next three games in succession. He defeated Hjartarson and Khalifman (see
page 462) with the white pieces, then Gulko with black. But then the magic ended in Round 8
when he tried too hard to win an equal ending against Ljubojevic and ended up losing.
Karpov recovered and held Ivanchuk from a slightly worse position in the Caro-Kann, then
won convincingly against Andersson on the white side of a Bogo-Indian. His next three games,
against Seirawan, Nikolic and Timman, were all drawn relatively quickly. In the penultimate
round Karpov outplayed Ehlvest on the white side of a Griinfeld and won in good style. In the
fifteenth and final round Karpov outplayed Chandler but the Englishman stubbornly defended
a pawn-down endgame and eventually, after missing a few wins, Karpov had to take a draw after
1 1 9 moves.
Karpov's final score of 1 OY2/ 1 5 was enough to share first place with Ivanchuk. The second
World Cup was aborted after this tournament, which was the first of the new series. It was a
success for Karpov, who showed that he could still play great chess.
***
Karpov's next tournament was a double-round-robin event in Tilburg. He started with a double
white, but could only draw with Bareev and Short. He drew the next game in seventeen moves
against Timman. In Round 4 Karpov built up a winning position against Korchnoi, but allowed
the crafty veteran to escape with an ingenious piece sacrifice in the ending. In Round 5 Karpov
made his fifth consecutive draw after neutralizing Kamsky with the Caro-Kann.
In Round 6 Karpov faced Anand, who once again played the Short System against the Caro­
Kann. A complicated game ensued. At one point Karpov had the edge, but then the game turned
against him and the Indian grandmaster was able to take a measure of revenge for his defeat
in their candidates match. In Round 7 Karpov faced Kasparov, who ventured a hair-raisingly
complicated piece sacrifice in a Classical King's Indian. Karpov responded well and missed a win
in the middlegame, though he kept the advantage. The game eventually settled in an intriguing
endgame where Karpov had a bishop and two knights versus Kasparov's rook. Despite being
tricky for the defender, it was a theoretical draw and Kasparov successfully held it. Thus Karpov
ended the first half of the tournament on "minus one" and without a win.
Amazingly, after six draws out of his first seven rounds, Karpov only drew one more game in
the rest of the tournament! In Round 8 he received a gift from Bareev, who blundered a piece in
the early middlegame and soon resigned. In the next game he held Short. In Round 1 0 Karpov
defeated Timman, and in Round 1 1 he got the better of Korchnoi in a complicated English.
Then he suffered a setback on the white side of a Griinfeld against Kamsky, blundering in a
difficult endgame. He almost suffered a second white defeat after being outplayed by Anand in
a Queen's Indian, but the Indian squandered his advantage and collapsed. In the fourteenth and
final round, Karpov was blown away by Kasparov's lethal preparation in the Scotch and although
he lasted until move 44, he was never in the game.
Karpov's final score of 7Y2/ 1 4 was good enough for fourth place behind Kasparov, Short and
Anand. It was not a disastrous result, but it was not a great one either.
***
199 1
173
Karpov's final event of the year was the 1 99 1 /92 Reggio Emilia tournament, which started in
December and finished in January. He started with a good win over Beliavsky, then he held
Salov who surprised him with a Scotch. In Round 3 he used his trusted Samisch Variation
against Kasparov's King's Indian, but got into serious trouble and was on the brink of defeat, but
eventually managed to hold the endgame. In Round 4 he played the Zaitsev against Khalifman,
but eventually lost after an extremely sharp struggle.
Karpov bounced back with a win over Mikhail Gurevich, then drew his final four games against
Anand, Polugaevsky, Ivanchuk and Gelfand. His final score of 5/9 was good enough for fourth
place. Interestingly the winner of the tournament was Anand, who scored a superb win with the
black pieces against Kasparov.
In 1 99 1 Karpov achieved some successes; he won two tournaments, and was victorious in his
candidates match against Anand. Nevertheless his overall results did not match the tremendous
level he had achieved during the eighties. It looks like he may have once again played in too many
tournaments and lacked the energy to perform at his best in all of them. 1 99 1 was the year in
which he turned forty, so his age may have been a factor, especially with so many talented young
super-GMs moving through the ranks.
174
1991 Summary
Reggio Emilia ( l st place) : 7Yz/ 1 2 (+3 =9 -0)
Linares (7th-8th place) : 6Y21 1 3 (+4 = 5 -4)
Euwe Memorial, Amsterdam (4th place) : 5 Y2/9 (+2 =7 -0)
Match versus Agdestein, Gjovik: Drew 2-2 (+ 1 =2 -1 )
Candidates quarter-final match versus Anand (Brussels) : Won 4Y2-3Yz (+2 = 5 - 1 )
World Cup, Reykjavik ( l st-2nd place) : I DYz/ I 5 (+7 =7 - 1 )
Tilburg (4th place) : 7Yz/ I 4 ( + 4 = 7 -3)
Reggio Emilia (4th place) : 5/9 (+2 =6 - 1 )
Total 5 8.3% (+25 = 4 8 -1 1 )
II Wins • Draws • Losses
1992
Rating 2725 (2 in the world)
Karpov's first tournament in 1 992 (apart from Reggio Emilia which began in December 1 99 1
and ran into 1 992) was the prestigious Linares super-GM event. Karpov started well, grinding
out a win against Ljubojevic on the black side of a 2.d3 Caro-Kann. In Round 2 he had the black
pieces again, but this time Kasparov took the initiative in a 4 . . . ttJ d7 Caro-Kann and never let
Karpov off the hook. In Round 3 he got no advantage against Anand's QGA and the game ended
in a draw.
In the next two rounds Karpov faced two Englishmen. Speelman opted for the Four Knights
opening and sacrificed a piece for a promising initiative, but Karpov managed to hold. Then in
Round 5 Short experimented with the Ilyin-Zhenevsky variation of the Dutch Defence, but
failed to equalize and Karpov won convincingly. In Round 6 Karpov moved to "plus two" with a
hard-fought win on the black side of a Scotch against Illescas.
In Round 7 Karpov faced Valery Salov. The Russian grandmaster contested fifty six games
against the world champions, winning eight, drawing twenty four, and losing thirteen. Hitherto
Karpov had played him eight times, with one win apiece and six draws. Their lifetime score is six
wins to four in favour of Karpov, with ten draws.
I Game l61
Anatoly Karpov - Valery Salov
Linares 1 992
l .d4 ttJf6 2.c4 e6 3.ttJf3 b6 4.g3
Karpov played the white side of this position seven times against Salov. He did well with it,
scoring five wins, one draw and one loss.
4... ia6
Previously Salov had favoured 4 . . . ib7, scoring a win, a draw and a loss against Karpov. The text
move was also a mainstay of his repertoire.
5.b3 ib7
Karpov faced this move seven times, and won them all. It is a wonderful record, although in
one of the games he was losing and only won after Anand misplayed his position.
6.ig2 .ib4t 7..idl a5
176
The Prime Years
Karpov faced 7 . . . c5 just once. The game
proceeded as follows: B.O-O 0-0 9.�c3 d5
1 O.cxd5 It'lxd5 1 1 .�b2 cxd4 1 2.'lWxd4 'lWf6?!
( 1 2 ... f6!?) 1 3.a3 'lWxd4 ( 1 3 ... �e7 1 4.'lWd2 'lWf5
1 5 .lt'l c3 'lWh5 1 6.lt'le5 l"i:dB 1 7.l"i:fd l White is
somewhat better here as well.) 1 4.lt'lxd4 �c5
1 5 .lt'lb5! It'l c6?! (Better was 1 5 . . . l"i:cB I 6.lt'l I c3
a6 1 7.lt'lxd5 �xd5 I B.�xd5 exd5 1 9.1t'lc3 l"i:eB
and Black is not much worse.) 1 6.b4! �e7
1 7.e4! It'l f6 I B.e5 lt'l d5
a
b
c
d
e
f
g
9.�c3
In each of the other games in which he
reached this position, Karpov favoured 9.'lWc2.
He faced several replies:
a) 9 . . . c5 1 0.l"i:d l It'la6 1 1 .�e3 a4 1 2.lt'lc3 axb3
1 3.axb3 d5 ( 1 3 . . . d6!?) 1 4.lt'la4 h6 1 5 .'lWb2 l"i:e8
( 1 5 . . . lt'lg4!?) 1 6.lt'le5 l"i:bB?! 1 7.lt'ld3! Karpov
creates additional pressure against Black's
hanging pawns. 1 7 . . . dxc4? I B .�xb7 l"i:xb7
1 9.bxc4 'lWe7
h
1 9.1t'l l c3! After gaining space with his pawns,
Karpov exchanges some pieces in order to
invade. 1 9 . . . lt'lxc3 20.�xc3 It'l dB 2 1 .�xb7
It'lxb7 22.l"i:fd l l"i:fdB 23.l"i:xdBt �xdB 24.l"i:dl
a6 25 .l"i:d7 axb5 26.l"i:xb7 White obtained a
clear advantage and went on to win, Karpov Korchnoi, Biel 1 992.
8.0-0 0-0
8
a
b
c
d
e
f
g
h
20.lt'laxc5 ! bxc5 2 1 .l"i:xa6 White won a pawn
and later the game, Karpov - Nikolic, Tilburg
1 9BB.
b) 9 . . . h6 1 O.a3 (At Tilburg the previous year
Karpov chose 1 O.lt'lc3 against Anand, but he
was outplayed in the middlegame and was
fortunate to win in the end.) 1 O . . . �e7 1 1 .ctJc3
d5 1 2.l"i:fd l 'lWcB 1 3.cxd5 exd5 1 4.�f4 l"i:dB
8
7
7
6
6
5
5
4
4
3
3
2
2
1
a
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
1 992
177
Anatoly Karpov - Valery Salov
1 5.lLlh4! i.fS 1 6.lLlf5 E1d7? A mistake in an
already difficult position. 1 7.lLJe3 'lMrdS l S.i.e5
c6 19.i.h3 E1e7 20.i.xf6 gxf6 2 1 .lLJf5 Karpov
has already outplayed his opponent strategically
and he converted his advantage convincingly,
Karpov - Lobron, Baden-Baden 1 992.
c) 9 ... d6 1 O.lLJc3 lLJ bd7 1 1 .E1ad l E1eS?! 1 2.E1fe 1 !
White threatens to assume full control over
the centre with e4. 1 2 . . . i.xc3 1 3.i.xc3 i.e4
14.'Wc 1 a4
9 ... d5
The main line is 9 . . . d6, but Salov prefers a
different approach.
1O.�c2
Later in the same year Karpov drew quickly
from the other side of the board: 1 0.cxd5 i.xc3
1 1 .i.xc3 i.xd5 1 2. lLJ e 1 lLJ a6 1 3.f3 c5 1 4.dxc5
lLJxc5 1 5 .lLJ d3 i.b7 1 6.lLJxc5 bxc5 1 7.'lMrxdS
Y2-Y2 Chernin - Karpov, Tilburg 1 992.
1 0 ... tlJ a6 l 1 .�adl
Karpov decides to postpone any ideas of
taking on d5 or kicking the enemy bishop
with a3, but he does not forget about these
possibilities.
1 1 ...h6
a
b
c
d
e
f
g
h
1 5.i.h3! Karpov wants to gain space in the
centre without exchanging the light-squared
bishops. 1 5 . . . axb3 1 6.axb3 E1a2 1 7.E1d2 'lMraS
1 8.E1xa2 'lMrxa2 1 9. 1LJ d2 b5 20.f3 i.c6 2 1 .e4
bxc4 22.bxc4 e5 23.'lMral E1aS 24.'lMrxa2 E1xa2
25.d5 i.b7 26.E1b 1 i.a6
8
7
a
6
b
c
d
e
f
g
h
12.tlJe5!
Karpov exerts pressure on Black's centre.
5
4
12 ... �c8
Defending the b7-bishop is a useful
precaution for Black.
3
2
a
b
c
d
e
f
g
h
27.c5! dxc5 2S.f4! With the last two powerful
moves White obtained a clear advantage
which he converted smoothly in Karpov Romanishin, Tilburg 1 993.
13.i.h3 �e8
Salov wants to have the option to recapture
on d5 with the e-pawn.
14.cxd5 exd5
178
The Prime Years
This is not a bad move, although there is
always a risk that Black's central pawns will
come under unpleasant pressure, as occurs
later in the game.
1 4 . . . j,xc3!?
By giving away one half of his bishop pair,
Black avoids creating a structural target.
The note to White's tenth move reveals
that Karpov employed the same idea when
Chernin took on d5 against him. It is even
possible that he got the idea when analysing
the present game.
1 5 .Wi'xc3
After 1 5 .j,xc3 ltJxd5 1 6.j,a l ltJ ab4 Black
pieces are active enough.
1 5 . . . ltJxd5
1 5 . . . j,xd5?! is well met by 1 6.f3!
1 6.Wi'c4
1 6.Wi'b2 Wi'e7 is solid enough.
a
b
c
d
e
f
g
h
1 6 . . . c5!
Piece play is not enough to solve Black's
problems: 1 6 . . . ltJ ab4 1 7.a3 ltJ c6 ( 1 7 . . . j,a6
I B .Wi'c l f6 1 9.1tJc4 White gets the upper
hand in the centre.) I B.ltJxc6 Wi'xc6 1 9.Wi'xc6
j,xc6 20 J"k 1 White has a long-lasting
initiative on the c-file.
1 7.e4 ltJ db4 I B.a3 ltJ c6 1 9.1tJxc6 Wi'xc6 20.j,g2
cxd4 2 1 .Wi'xd4 ltJ c5
Black pieces are active, so he can live with
White's bishops.
1 5.i.f4
It was possible to play on the kings ide with
1 5 .f4, but ceding an outpost in the centre is
hardly Karpov's style.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 5 .. J3d8
It was worth considering 1 5 . . . j,d6, playing
against the knight immediately and preparing
. . . ltJh5. 1 6. ltJ d3 ( 1 6.j,g2 Wi'e6 and 1 6.1"i:fe1
ltJh5 are both okay for Black.) 1 6 . . . ltJb4
1 7.Wi'd2 ltJxd3 I B.j,xd6 cxd6 1 9.Wi'xd3 White
is a little better, but Black should be able to live
with his slight pawn weaknesses.
1 6.Wfc1 i.c8
Black decides to lose a tempo in order to
drive the enemy bishop back. There were
several other ideas available.
1 6 . . . Wi'e7?! 1 7.ltJb5 1"i:feB I B.a3 j,d6 1 9.1tJxd6
cxd6 20.ltJc6 White is clearly better.
1 6 . . . ltJh5 1 7.j,e3 ltJ f6? (The superior 1 7 . . . j,d6
gives Black a playable position.) I B.j,xh6! gxh6
( 1 B . . . j,xc3 1 9.j,xg7+-) 1 9.Wi'xh6 j,e7 20.j,f5
j,cB 2 1 .g4 White's attack is very strong.
1 6 . . . j,d6!? seems simplest: 1 7.ltJd3 ltJ b4 (There
is also 1 7 . . . c5!? I B.dxc5 bxc5 when Black is
okay.) I B.j,xd6 ltJxd3 1 9.1"i:xd3 1"i:xd6 Black
should be able to withstand White's pressure
in the centre.
Anatoly Karpov
1992
17.ig2 ib7 1 8.a3 id6 19.1iJc4!
Karpov exploits the pin on the long diagonal
to transfer the knight to e3. It is a strong albeit
well-known motif in this opening.
19 .ixf4 20.V;¥xf4 ic8?!
Salov transfers the bishop to e6, which
often happens in the related Tartakower­
Bondarevsky-Makogonov vanatlon.
The
idea is playable, but there was something
better.
..
20 ... a4! This clever move allows Black to
maintain a dynamically equal game: 2 1 .lLle3
(2 1 .bxa4?? 'lWc6 wins a piece; 2 1 .lLlxa4 bS
22.ttJaS bxa4 23.lLlxb7 E1b8 Black becomes
very active.) 2 1 . . .axb3 22.E1b 1 (22.lLlfS cS;
22.�fS c6) 22 ... lLl e4 23. lLl cxdS (23.E1fc 1
f5) 23 . . . ixdS 24.lLlxdS E1xdS 2S.ixe4 E1xd4
26.ih7t 'tt> xh7 27.'lWxd4 lLl cS Black is an
exchange down, but he is very much in the
game thanks to the mighty protected passed
pawn on b3.
21.tlJe3 ie6
-
179
Valery Salov
22 . . . 'tt> h 7?!
This cheeky move is not really effective.
23.e4!?
Attempting to refute Black's play.
The point behind Black's previous move is
that 23.lLlxh6?! can be met by 23 . . . 'tt> g6!
with unclear play.
However, White can always just improve his
position with moves like 23.h3 or 23.E1c 1 ,
when he keeps a nice advantage.
23 . . . dxe4 24.lLlxh6 lLlhS
Now 24 ... 'tt> g6? is not working as 2S.lLlxe4
lLlxe4 26.ixe4t fS 27.lLlxfS! ixfS 28.'lWg4t
wins.
2S .'lWgS gxh6 26.'lWxhS fS 27.'lWe2 ixb3
28.'lWxa6 'IWc6 29.E1c1 ic4 30.'lWa7 E1a8 3 1 .dS
White saves his queen and keeps a clear
advantage.
23.V;¥xfS
8 n-����7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
22.tlJf5!
This virtually forces Black to exchange his
light-squared bishop.
23 ... V;¥e6?
By defending the dS-pawn in this unfortunate
way, Salov allows his a6-knight to be driven
into a bad position.
Abandoning the d-pawn was not really an
option: 23 . . . 'lWe7? 24.lLlxdS lLlxdS 2S .ixdS
'lWxa3 26.ic4 White's bishop is clearly superior
to Black's knight.
22 ...ixfS
The best chance was:
a
b
c
d
e
f
g
h
The Prime Years
180
23 . . . c6!
This move has the advantage of using a less
powerful unit to defend a weakness, but its
real value lies in the fact that it vacates the
c7-square in order to facilitate the transfer of
the a6-knight to the centre.
24.e4!?
Opening the centre is a principled reaction,
which brings White some activity. Black
must take care not to be blown away.
Quieter alternatives do not lead anywhere
special, for instance: 24.\M!d3 CiJ c7 2S J''k 1
(2S . b4 �a8) 2S . . . CiJ e6!? This implies a
pawn sacrifice. (Black has a safer alternative
in 2S . . . Wd7 26.CiJa4 �bS 27.Wc3 CiJbS
[27 . . . CiJ e6] 2S.Wb2 �fcS when all his pieces
stand well.) 26.CiJa4 �bS 27.e3 (27.�c3
Wd7) 27 . . . Wd7 2S.Wa6 Wd6 29.CiJxb6 cS
Black gets promising play for the pawn.
24 . . . dxe4
24 . . . We6 does not solve all Black's problems:
2S .Wxe6 fxe6 26.eS CiJ d7 27.f4 CiJc7 2S.�c1
White is somewhat better.
2S.CiJxe4 CiJxe4 26.ixe4 g6 27.Wf4
After 27.Wf3 �d6 2s.Wf4 Wd7 29.Wxh6
CiJ c7 Black has enough compensation to
maintain the balance.
27 . . . CiJ c7 2S.�fe l CiJ e6 (2S . . . Wd7? 29.Wxh6
CiJbS 30.dS cS 3 l .d6!+-) 29.Wxh6 �d6
3D.dS WdS 3 l .b4 axb4 32.axb4 cxdS Black
should be able to hold.
2s.ixc6 Wxa3 29.Wxh6 Wxb3 30.�b l We6
3 l .dS Wf6 32.�xb6 CiJ b4
Black survives.
24.W!d3!
Karpov wants more than an endgame with
a small advantage. After 24.Wxe6 fxe6 2S .e4
dxe4 (if 2S . . . c6 26.eS CiJ d7 27.ih3 White can
press for long time to come) 26.CiJxe4 CiJxe4
27.ixe4 CiJ bS 2S.dS c;t>f7 Black should be able
to survive with a knight against a bishop.
24... lLl b8
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
25.b4!
Karpov gains space on the queenside and
prepares to fix Black's c-pawn.
a
b
c
d
e
f
g
h
27 . . . We7
Two other moves are playable:
27 . . . <;t>g7 2S.�fe l Wd7 29.dS cxdS 30.ixdS
�feS Black should be able to live with his
small disadvantage.
25 ... axb4 26.axb4 �c6 27.b5 � a5
Salov closes the a-file and makes Karpov pay
attention to the c4-square.
The alternative was:
27 . . . CiJ e7
The knight has better prospects in the centre,
on the other hand it fails to block the a-file
from here.
1 992
Anatoly Karpov - Valery Salov
28.E1al ttJc8
28 . . . E1d7 29.l"Ifc 1 and 28 . . . l"Ia8 29.l"Ia6 are
both difficult for Black.
28 . . . c5?! is not a good idea, as the weakness
on b6 is easier to attack than the one on c7.
29.bxc6 l"1c8 30.e3 l"1xc6 3 1 .l"Ifb l White has
excellent chances to squeeze the full point.
29.E1fc1 ttJ d6 30.e3 l"Ia8 3 1 .l"Ia6
White takes over the a-file and remains in
full control.
18 1
on the c-file now: 3 1 .l"Ic2 l"Ia8 32.l"Iec 1 ttJ b3
33.l"Ib l ttJ a5 White has made no progress.)
3 1 . . .l"Ie7 32.ii.f1 'lWd6 33.'lWa2 (33 .ii.d3 'lWb4)
33 . . . g6 34.ii.d3 �g7 3 5 . ttJ e2 �h7 (35 . . . g5
36.l"Ic2) 36.�g2 �g7 37.ttJf4 ttJ f6 White
keeps the advantage, but he is a long way
from winning. It would have been interesting
to see how Karpov would have tried to break
through.
30.�c2! �e7 3 1 .e3 g6 32.�ecl <i!.>g7 33.� e2
After improving h is rooks Karpov unblocks
the c-file and sends his knight towards a better
home.
33 ... �a8 34.�f4 �a7
8
7
6
5
a
b
c
d
e
f
g
h
28.l"Ifel!?
It looks like Karpov j ust wanted to tire
his opponent by making him calculate the
consequences of e2-e4.
28 ... l"Ife8 29.�cl
Karpov starts directing his heavy artillery
against the c7-pawn. 29.e4? would squander
a large part of White's advantage, and after
29 ... dxe4 30.ttJxe4 ttJxe4 3 1 .l"Ixe4 'lWc4 Black
should be all right.
29 ... Yfid6
Salov places his queen on a nice square and
reinforces the c7 -pawn.
It was worth considering a more active
defence involving the acquisition of extra
kingside space: 29 . . . ttJe4 30.e3 f5!? 3 1 .'lWc2
(It is harder for White to arrange his rooks
4
3
2
1
a
b
c
d
e
f
g
h
35.�a2!
At first it looks like Karpov is trying to play
on the a-file, but in fact he has something more
subtle in mind. The real idea is to exchange
Black's best defensive piece.
35 ... �d7?!
Salov may have either overlooked or
underestimated Karpov's plan. It must
be said that Black's alternatives were also
unappealing.
It is worth pointing out that 35 . . . �h7? is
refuted by 36.l"Ixa5! l"Ixa5 37.l"Ic6.
The Prime Years
182
35 . . . Ela8 avoids the queen exchange, on the
other hand White can still swap a pair of rooks
which will also soften Black's position. 36.Elc6
(36.�a3 �d8 37. ltJ d3 ltJ e4 38.ltJe5 is also
unpleasant for Black.) 36 . . . ltJxc6 37.Elxa8 ltJ a5
38 .h4 ltJe4 39.�c2 Black's position is difficult
as ideas of �a2 or even .txe4 are in the air.
35 . . . �d8 36.�a3!? White clears the d3-square
for his knight. (36.�b l Ele8 37.ltJd3 is also
promising.) 36 . . . Ele8 37.Elac2 i>g8 (37 . . . Ele7??
38.ltJxd5 wins.) 38.ltJd3 White remains in
control, and his knight will stand superbly on
the e5-square.
Black can no longer keep his queen on the
board.
37.�xa3
8
7
6
5
4
3
2
35 . . . g5!? Maybe Black should have pushed his
defensive line forward and chased the knight
away. 36.ltJe2
a
b
c
d
e
f
g
h
36 . . . �e6! (36 . . . ltJe4? allows White to exploit
the weakness of the f5-square with 37.f3 ltJ f6
38.e4 or 38.g4 with ltJ g3-f5 to follow.) 37.ltJc3
ltJe4 38.�b l (38.Ela3 f5 39.ltJa2 �d6 Black
is in the game) 38 . . . f5 39.Elac2 �f7 40.�a2
Eld7 Black remains worse, but he continues to
resist.
36.V9a3!
Black's queen is his best defensive piece.
Once it disappears, Black will have serious
problems defending his weaknesses.
36 ...V9xa3
a
b
c
d
e
f
g
h
37... i>fS
Salov brings the king to the centre. This
causes a subsequent traffic jam in his position,
but there was no choice as his pieces are unable
to hold the position together by themselves.
Here is an illustrative line: 37 . . . Eld8 38.ltJd3
Eld7 39.ltJe5 White nicely improves his knight.
39 . . . Eld6 (39 . . . Ele7 40.Elxa5!!+-) 40.Elac3
ltJ e8 4 1 .h4 h5 42 . .th3 f6 (42 . . . i>f8 43 . .td7)
43. ltJ d3 White's simple play has resulted in the
creation of a third weakness on g6, which is
too much for the defence to bear.
38.liJd3!
Karpov finds an even more effective square
for the knight.
38 ... i>e8
Black keeps the e7-square for his rook.
38 . . . i>e7? runs into 39.ltJe5 Eld6 (39 . . . Eld8
40.Elxa5! Elxa5 4 1 .ltJc6t +-) 40.Elac3 ltJe8
4 1 .h4 Eld8 42.h5 g5 43.f4 gxf4 44.exf4 and
White has a huge advantage.
39.liJ b4! h5
Salov knows that at some point Karpov
1 992
1 83
Anatoly Karpov - Valery Salov
will look to push him back with his kingside
pawns, so he prepares for it.
40J�ac3 �d8?
On the last move before the time control,
Salov blunders in a difficult position. He had
to try one of the following continuations.
Sitting and waiting would have been a
depressing prospect: 40 . . .'tt> f8 4 1 . tLl a6 tLl e8
42.if3 <j;>g7!? By vacating the centre the black
king avoids certain tactical motifs. Nevertheless
he is so paralysed that one would expect Karpov
to find a way to win. (42 . . . E!:b7? 43 .tLlb4 tLl f6
44.liJ c6 White soon invades; 42 . . . <j;>e7 43. tLl b4
tiJ f6 44.E!:c6 tLl b3 45.E!: l c2 tLl a5 46.i.g2 Black
can hardly move.) 43.h3 Advancing on the
kingside is a logical way to improve White's
position and squeeze the defence even more.
43 ... <j;>h6 44.<j;>g2 <j;>g7 45.g4 hxg4 46.hxg4
It>h6 47.<j;>g3 Black is in trouble.
The best chance looks to be: 40 . . . tLl e4!? 4 1 .E!:3c2
(White could also consider: 4 1 .i.xe4!? dxe4
42.lt>g2 <j;>d8 43.f3 exf3t 44.<j;>xf3 f6 45.<j;>f4
Karpov won a rook ending with a somewhat
similar pawn structure against Portisch in
Linares 1 989.) 4 1 . . .f5!? 42.tLld3 (There is also
42.f3 tLlf6 43.<j;>f2 <j;>f7 44. tLl d3 and Black is
living dangerously.) 42 . . . tLl b3 43.E!:d l tLl a5
44.LiJe5 E!:g7 45 .i.fl White maintains a great
positional advantage.
4 1 . Ek6!
This sweet move finally breaks Black's
resistance.
41 ...llJe4
After 4 1 . . . <j;>e7 the rook still cannot be taken,
and 42.i.h3! wins easily.
42.�h3! gd6
Black saves the exchange, but loses his d5pawn. 42 ... E!:e7 would have been refuted by
the pretty 43.E!:xb6!.
43.gxd6t llJxd6 44.�f1.!
Karpov defends his own pawn before he
takes his opponent's one.
44 ... llJ e4
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
45.llJxd5
Finally the d-pawn falls, along with any
realistic hopes for Black to save the game.
Salov struggles on for a while longer, but he is
fighting a lost cause and the rest of the game is
not particularly interesting.
45 ... tlJ b3 46.gc2 g5 47.�e2 gal t
47 . . . g4 48.f3 wins further material.
48.�g2 llJcl 49.6 llJd6 50.�fl g4 5 1 .fxg4
hxg4 52.llJf6 ga2 53.gil gxflt 54.�xfl
a
b
c
d
e
f
g
h
The Prime Years
1 84
lLla2 55.e4 lLle3 56.'iil e3 lLl cxb5 57.d5 e5
58.me6 lLla7 59.e5 lLlde8 60.e7t <t!?xe7
61 .i.e4 lLle6 62.lLlxg4
1-0
This win raised Karpov's score to an
impressive 5/7, but he was unable to maintain
his hot run of form. In Round 8 he lost to
Timman after mutual blunders just before the
time control. In the next two rounds Karpov
drew solidly with Ivanchuk and Gelfand. Then
in Round 1 1 Bareev surprised him with the
3.tt:lc3 e5!? variation of the Slav. Karpov got
no advantage, and later he played an unsound
pawn sacrifice which led to an eventual defeat.
He managed to recover with a good win
against Beliavsky, then drew in the last round
against Yusupov.
Karpov finished fourth with a score of
7Y21 1 3, half a point behind Ivanchuk and
Timman. Kasparov won the tournament by a
massive two-point margin ahead of those two
players. Interestingly Nigel Short finished in
equal last place, but Karpov would soon find
out that the English grandmaster was not a
player who should be underestimated.
***
Karpov's opponent in the candidates semi­
final match was Nigel Short, who had defeated
Gelfand and Speelman to make it this far.
Although these were impressive achievements,
he entered the match as the heavy underdog
against Karpov.
The first game underlined Karpov's
superiority. Short unveiled a surprise opening
in the Budapest Gambit, but Karpov was
unfazed and effortlessly obtained an opening
advantage. Later he won a pawn, and converted
his advantage without any problems. In Game
2 Short used his own patented system against
the Caro-Kann, but Karpov equalized and
drew without difficulty. In Game 3 Short
wisely abandoned the Budapest and switched
to a QGA. He equalized and went on to
outplay the former champion, but made some
mistakes and allowed Karpov to escape with
a draw from what had seemed like a hopeless
situation.
Game 4 was a turning point in the match,
as Short brushed aside his disappointment
from the previous game and levelled the score.
Karpov sacrificed an exchange to reach a drawish
endgame with three pawns versus Short's two
on the kingside, but the Englishman persevered
and eventually won a long endgame. In Game
5 Karpov tried a different line in the QGA,
but Short was ready with a new idea. Karpov
got into serious trouble in the endgame and
was probably lost, but he managed to salvage
a draw.
In Game 6 Karpov switched to 1 . . .e5 but
Short was ready with another surprise, the
Worrall System in the Ruy Lopez. Karpov
got a good position but then lost his way
and blundered horribly after the position
had already turned against him. In Game 7
Short changed openings again and went for an
Orthodox Queen's Gambit, but failed to react
correctly when Karpov unleashed a novelty.
Karpov won a pawn and converted it smoothly
to draw level in the match.
In Game 8 Short repeated the Worrall.
Karpov was ready with a different defensive
set-up, but he did not manage to equalize and
Short won a fine game. With just two games
remaining, Karpov's situation was becoming
desperate.
In Game 9 Short reverted back to the QGA.
Karpov got a slight advantage but was unable
to do anything with it, and Short gradually
took over the initiative. He won a pawn, but
Karpov managed to defend and keep his hopes
alive.
In the tenth and final game, Karpov was
faced with the almighty challenge of having
to win with the black pieces. He abandoned
his solid openings in favour of a Classical
1992
Anatoly Karpov - Joel Lautier
Sicilian. Short was not interested in playing
timidly for a draw, and instead opted for a
Rauzer set-up with opposite-sided castling.
Karpov achieved his goal of reaching a double­
edged game, but unfortunately he was not at
home in the ensuing middlegame and Short
soon built up a winning advantage. At one
point Karpov could have equalized, but he
missed his chance and had to resign just after
the time control as the endgame was hopeless
for him.
This was the first time Karpov had lost a match
to anyone other than Kasparov. He was not at
his best, although credit must go to Short who
played well and chose his openings cleverly.
Karpov's disappointing result may have served
as a wake-up call, as he played great chess in his
next few tournaments and looked motivated
to rediscover his form from the seventies and
eighties.
***
185
slight advantage against Yudasin and pressed
for a long time in a rook ending, but the
Russian grandmaster eventually succeeded in
holding a draw.
Despite the calibre of the tournament being
slightly below Karpov's usual level, his final
score of 7Y2/9 was highly impressive, and gave
him first prize by a two-point margin ahead of
Judit Polgar and Epishin.
Karpov's next tournament was a double­
round-robin event in Biel. His opponent
in Round 1 was Joel Lautier. The French
grandmaster played all the world champions
from Karpov to Anand, scoring fourteen wins,
thirty seven draws and twenty six losses. Prior
to this game he had scored one draw and one
loss against Karpov. Their lifetime score is nine
wins to four in Karpov's favour, with eleven
draws.
I Game l1 1
Anatoly Karpov - Joel Laurier
Karpov's next tournament was in Madrid. The
event featured some top grandmasters, some
local players and two of the Polgar sisters.
Karpov got off to a flying start, scoring five
straight wins against San Segundo, Salov,
Magem Badals, Zsuzsa Polgar and Granda
Zuniga. Although his opponents made a few
mistakes, Karpov played convincingly.
In Round 6 he took a quick draw with
Epishin - it looks like neither of them were
in a mood for a fight. In the next round
Karpov took a similar route with Judit Polgar,
drawing on the black side of a Fort Knox
French in eighteen moves. In the next game
Romero Holmes cheekily played the Centre
Game, and later played an interesting piece
sacrifice to obtain an attack. At one point he
could have created some dangerous threats,
but he faltered and Karpov never let him off
the hook. In the final round Karpov got a
Biel l 992
l .d4 d5 2.c4 c6 3.tLJc3 tLJf6 4.e3 e6 5.tLJa
tLJ bd7 6.Wfc2
This was Karpov's usual choice against the
Semi-Slav. Lautier had hardly ever played
this defence before the present game, so he
probably judged it to be one of the softer
points of Karpov's repertoire.
6 ... i.d6 7.i.e2 0-0 8.0-0 ge8
Hitherto Karpov had not faced this move.
He encountered it on four subsequent
occasions, and scored a total of three wins and
two draws.
9J;dl Wfe7 10.h3
In 1 996 Karpov chose l O.a3 and drew
against Kamsky.
The Prime Years
1 86
Karpov feels he needs to exchange bishops in
order to make inroads.
a
b
c
d
e
Avoiding exchanges with 1 4.tLle5 was not
really dangerous: 1 4 . . . f5 ( 1 4 . . . tLl f6 1 5 .'iWh4
c5 1 6 .�g5 h6 1 7.�xf6 'iWxf6 1 8.'iWxf6 gxf6
1 9.tLld7 Wg7 20.dxc5 �xc5 2 1 .tLlxc5 bxc5
Black should be able to live with his small
disadvantage.) 1 5 .'iWe3 tLl f6 ( 1 5 . . . tLlxe5
1 6.dxe5 �c7 1 7.�f3 White is just a fraction
better.) 1 6.�d2 c5 1 7.�c3 E':ad8 Black has a
playable position.
f
g
h
10 ... b6
Lautier goes for one of the main lines.
Developing the bishop to b 7 is certainly a
natural idea.
Karpov also encountered 1 O . . . h6 1 1 .a3 in two
games:
a) 1 1 . . .a6 1 2.b3 dxc4 1 3.bxc4 e5 1 4.tLlh4 tLl f8
1 5 .tLlf5 �xf5 1 6.'iWxf5 exd4 1 7.exd4 c5 1 8.�e3
cxd4 1 9.�xd4 �e5 20.�xe5 'iWxe5 2 1 .'iWxe5
E':xe5 22.E':ab 1 Karpov soon won a pawn and
later the game, Karpov - Illescas Cordoba,
Wijk aan Zee 1 993.
b) Three years later Lautier tried his luck in
the Semi-Slav again, and opted for a different
approach: 1 1 . .. dxc4 1 2.�xc4 e5 1 3.tLlh4
tLl f8 1 4.dxe5 �xe5 1 5 .�d2 �c7 1 6.tLlf5 'iWe5
1 7.tLlg3 �e6 1 8 .�d3 E':ad8 1 9.tLlce2 �b6
20.b4 tLl g6 At this stage Black's position was
all right although White eventually managed
to win, Karpov - Lautier, Dortmund 1 99 5 .
14 .. J:'Iad8
Black can choose from many moves, but
none of them equalize. Laurier's choice did not
become popular, but it is probably no worse
than the alternatives.
1 4 . . . c5 1 5 .�xd6 ( 1 5 .'iWxb7 �xf4 1 6.dxc5
tLlxc5 1 7.'iWxe7 E':xe7 Black is okay) 1 5 . . . �xe4
1 6.�xe7 E':xe7 1 7.dxc5 tLlxc5 1 8.b4 tLla4
1 9.E':ac 1 White controls the d-file, which gives
him an edge.
1 4 . . . tLl f6 1 5 .'iWe3 �xf4 1 6.'iWxf4 c5 1 7.tLle5
White has a typical slight plus.
14 . . . �xf4 1 5 .'iWxf4 c5 1 6.'iWc7 E':ab8 1 7.dxc5
tLlxc5 Black is slightly worse, but he should be
able to hold the position.
1 5.i.xd6
Karpov proceeds with his plan of exchanging
bishops.
1 l .e4
If White first develops his bishop to the b2square, it will allow Black to swap the dark­
squared bishops from a3 .
If he avoids it, then Black will play . . . c5
with good effect, for instance: 1 5 .tLle5 tLl f6
1 6.'iWc2 ( 1 6.'iWe3 c5! 1 7.dxc5 �xc5 1 8.'iWb3
tLl e4 Black has good chances to take over the
initiative.) 1 6 . . . c5 Black has done well from
this position.
1 1 ...l£J xe4 12.tLlxe4 dxe4 13.�xe4 �b7
14.1.£4
1 5 ...�xd6
I S7
Anatoly Karpov - Joel Lautier
1992
18 ...ia8
a
b
c
d
e
f
g
h
16.lLle5
Karpov offers to exchange knights, after
which he will endeavour to gain control over
the d-file.
16 ... lLlxe5
Lautier is unafraid.
The ugly 16 . . . f6 does not equalize: 1 7.lLlxd7
�xd7 I S.c5 ( I S.We3!?) I S . . . Wd5 1 9.We3
Black will struggle to find a role for his passive
bishop.
Maybe Black's best is 1 6 . . . ttJ f6!? when play
might continue: 1 7.We3 c5 ( I 7 . . . Wc7 I S.if3
c5 1 9.dxc5 bxc5 20.ixb7 Wxb7 2 1 .b3 Wc7
22.lLld3 Black should be able to hold this
position, but it will be not easy as the c5-pawn
is weak and White will control the d-file.)
1 8.dxc5 Wxc5 1 9.Wxc5 bxc5 20.f3 Karpov
evaluates the position as slightly better for
White.
17.dxe5 Wfc7
White has obtained a small space advantage
and he has the more active bishop, but it is
not easy to hurt his opponent as he has no
structural weakness.
18.iO!
Karpov prevents Black from freeing his
bishop with . . . c5 .
a
b
c
d
e
f
g
h
19J�!xd8!
Karpov finds a nice idea. Many players
would only think of trying to put a rook on
d6 and winning control over the d-file, but
Karpov actually gives it up. The point is that
without a rook Black cannot defend his light­
squared bishop.
1 9.Wh4!?
Karpov
mentioned
this
interesting
alternative, which controls the dS-square
and thus prevents Black from exchanging
both rooks. Black has to play carefully.
1 9 . . . h6
1 9 . . J::\ e 7? 20J'1:xdSt WxdS 2 1 .Ei:d l Wc7
22.Wd4 White dominates the d-file.
1 9 . . . Ei:xd I t 20.Ei:xd i Wxe5 2 1 .Ei:dS (2 1 .Ei:d7
h6 defends) 2 1 . . .'jr fS 22.Ei:xeSt mxes
23.Wxh7 me7 24.WhS ib7 25 .g3 White
keeps some advantage in the endgame as
Black's bishop is a problem.
20.Wg3
Planning Ei:d6.
20 . . . Ei:xd l t!
Black must exchange a pair of rooks while he
has the chance.
20 . . . c5 2 1 .ixaS Ei:xaS? (It is not too late for
Black to correct his mistake with 2 1 . . .Ei:xd 1 t!)
22.Ei:d6 Ei:adS 23.Ei:ad l With the rooks on
The Prime Years
188
the board, a passed pawn on d6 should be
enough to decide the game.
2 1 .Elxd i c5 22.�xa8 Elxa8 23.Eld6 Eld8 24.'1Wd3
Elxd6 25.exd6 1.Wd7
Black is passive, but it is hard to suggest a
winning plan for White.
19 .. Jhd8 20J�dl !
Karpov follows his plan and exchanges the
other rook as well.
20.. Jhdlt
Obviously Black could move the rook
somewhere along the back rank in order to
play . . . c5, but giving up the d-file would be a
serious concession.
The main alternative was to wait with 22 . . . a5,
when White has a couple of ideas.
a) He can go for a slightly better queen ending,
although it seems a pity to release the black
bishop from its cage: 23.1.We3 c5 24.�xa8
1.Wxa8 25 .1.Wd3 The ending is a bit unpleasant
for Black, but he should be able to hold it. It
is worth comparing the present position with a
typical endgame resulting form a 4 . . . �f5 Caro­
Kann. Consider the following:
2 1 .i.xdl �d8
a
b
c
d
e
f
g
h
This posltlon occurred after White's 28th
move in the game Adorjan - Orso, Budapest
1 977. The game was eventually drawn, but
White missed a win along the way. Compared
with the present situation, the pawn on h5 is a
big help to White.
a
b
c
d
e
f
g
h
22.i.f3!
This was the idea Karpov had in mind when
deciding on his 1 9th move - the bishop on a8
will not be able to play a role in the game any
time soon. The whole plan required remarkable
vision and imagination; In fact I cannot recall
seeing it in any other game involving this pawn
structure.
22 ...�d2
This queen invasion is obviously critical, and
Karpov must have planned his response on
move 1 9.
b) The alternative IS: 23.g3 1.Wd2 (Worse is
23 . . . �b7?! 24.1.We3 [24.b3!?] 24 . . . g6 [24 . . . 1.Wc7
25 .c5!] 25 .c5 b5 Black is really passive.)
24.'if?g2 h6 25 .1.We2! 1.Wxe2 26.�xe2 c5t
[26 . . . 'if?f8 27.�f3] 27.�f3 �xf3t 28.mxf3 mfS
29.'if?e4 White has some winning chances as
the queenside is open for an invasion.
23.h3!?
Karpov prefers to give up the a-pawn instead
of its next-door-neighbour.
After the feeble 23.1.We2? 1.Wxe2 24.�xe2 c5
White has virtually no advantage.
1 992
In the event of 23.g3 Black should probably
follow suit: 23 . . . g6! The queen is well-placed
on d2, where it controls several important
squares and stops b4, so Black should not
move it yet. (23 .. .'IWxb2? is too risky: 24.Wld3!
�b4 25 .Wld7 g6 26.Wlxa7 [26.Wlc7!?J 26 . . . Wla5
27.'lWxa5 bxa5 28.'kfif1 The bishop endgame
is probably lost for Black.) 24.Wlh4!? White
intends to invade to e7. 24 . . . Wld7 25.b4 ib7
26.'lWe4 ia6 Black has better chances to draw
than White has to win.
a
b
c
d
1 89
Anatoly Karpov - Joel Lautier
e
f
g
h
23 ...Wlxa2!?
Lautier takes the pawn and gives up the d-file.
It is a risky policy, but Black still possesses
enough defensive resources.
A waiting policy would have given good
chances to defend:
23 . . . ib7 24.g3 g6 25.'kfig2
25 . . . h5
25 ... Wlxa2? 26.Wld3 'kfig7 27.Wld7 Wla6 28.b4
b5 29.c5 Black is desperately passive for the
sake of a mere pawn.
25 . . . Wg7!? 26.a3 (26.Wlh4 Wld7 Black
defends.) 26 . . . Wlc3 27.b4 Wlxa3 28.b5 Wlc5
29.bxc6 ia8! Black survives, as White
will have to give up the c-pawn for Black's
a-pawn.
26.a3 Wlc3 27.b4 ia6 28.Wlxc6 Wlxc4 29.Wlxc4
ixc4 30.ic6
Black is passive but I think he should be able
to hold by using his bishop to prevent White's
king from penetrating.
24.b4
Karpov follows through his surprising plan
to exploit the pin on the diagonal. If White
occupies the d-file with 24.Wld4, Black's
queen can return to the defence in time:
24 . . . Wla3! 25 .Wld7 (25 .Wld8t Wlf8 26.Wlc7 c5
Black manages to free his bishop.) 25 . . . Wla 1 t!
26.'kfih2 (26.id 1 h6 27.c5 ib7 Black escapes.)
26 ... Wlxe5t 27.g3 Wlb8! Black returns the pawn
and holds the position.
24 WlaI t 2S.i>h2
. . .
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
c
d
e
f
g
h
2S Wla6
25 . . . Wlb2!? Going for the f2-pawn leads to
a frightening position, but Black can get away
•..
a
b
The Prime Years
190
with it. 26.b5 'Wxf2 27.'Wd3 'Wc5 28 . .ixc6
.ixc6 29.bxc6 g6 (29 . . . 'Wxe5t?? 30.'Wg3+-)
30.'Wd6 'Wxc4 3 1 .c7 <j;lg7 White cannot force
his pawn through as Black will get a perpetual.
It is understandable that Lautier did not wish
to take such a narrow path to a draw, when the
tiniest mistake or oversight would have led to
an instant defeat.
30 . . . g6 3 1 .h4 White continues to press.)
30 ..ie4 g6 (30 . . . a6?? loses to 3 1 ..ixh7t) . Black
still has fair chances to survive, bur he faces an
unpleasant squeezing.
29.�e7
Karpov improves his queen a bit and returns
the ball to Laurier's side.
29 ... g6
Sooner or later Black had to make an escape
square for his king. If 29 . . . .ib7 then White
replies 30.h4, improving his position on the
kingside.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
26.'�M4!
Karpov keeps up the pressure. After 26.b5
'Wc8 27.bxc6 h6 28.g3 a5 Black is safe as he
will soon divert the queen and get rid of the
annoying passed pawn.
26 ...�c8 27.c5!
Ensuring that the bishop on a8 will remain
passive for a while longer.
27 ...bxc5 28.�xc5 a6
Laurier stops the threat of b 5 . Black has an
extra pawn and no serious weaknesses, yet he
still faces problems due to his passive pieces,
especially his bishop.
Karpov mentions 28 . . . 'Wb8!? without any
further analysis. It is playable, bur fails to
cur the Gordian Knot after 29.'Wd6! 'Wf8
(29 . . . 'Wxd6?! 30.exd6 �f8 3 1 .b5 cxb5
32 . .ixa8 Black faces an uphill struggle to draw
this ending; 29 . . . 'Wc8 30 . .ie4 [30.'We7 'Wb8]
a
b
c
d
e
f
g
h
30.h4!
Karpov creates a nasty threat of pushing the
pawn to h6. In fact 30.<j;lg3! would have been
even more accurate as after 30 . . . 'Wb8 White
can play 3 1 .<j;lf4.
30 ... h5
By committing himself to a rigid structure
with all his pawns on the same colour squares
as White's bishop, Black runs an obvious risk.
On the other hand he had to do something
about the advancing h-pawn.
30 . . . 'Wb8!? deserved attention. The attack
on the e5-pawn at least temporarily distracts
White from the kingside, although the first
1 992
191
Anatoly Karpov - Joel Lautier
player can maintain his grip with 3 1 .1Mfc5!
followed by advancing his king.
It is worth mentioning that other moves do not
work for White, for instance: 3 1 .g3 (3 1 .1Mfd6??
1iNxd6 32.exd6 �f8 simply loses for White.)
3 1 ...c5! This energetic strike liberates Black's
position. 32 . .ixa8 1Mfxa8 33.bxc5 1Mff3 34.�gl
1iNd I t 35.�g2 1Mfd5t 36.�h2 1Mff3 37.1Mfd8t
�g7 38.�gl 1Mfc3 39.1Mfd6 a5 Black is fast
enough.
31.�g3!
Karpov activates his king.
32. �g3!! Eke8 33. �f4! .ic8 34. �g5! 1-0 Short
- Timman, Tilburg 1 99 1 . Timman resigned as
there is nothing he can do to prevent the white
king from walking in and helping the queen
to deliver mate. This brilliancy was shown
throughout the world's chess media. It is
possible that the level of attention received by
Short's masterpiece may have caused Lautier
to overestimate the strength of the king march
in the present game. Perhaps he was already
dreading being on the losing side of another
game that would receive worldwide attention,
and was desperate to steer the game away from
the potentially headline-grabbing finale.
Instead he should have played:
3 1 . . . 1Mfb8!
Intending to activate the queen.
32.�f4
Karpov evaluates the position as clearly
better for White, but it is not so simple.
32.1Mfd6 1Mfb6 leaves White struggling to
make progress.
32 . . . 1Mfb6!
8
7
a
b
c
d
e
f
g
h
31...1Mfb7??
Lautier makes a losing blunder, which may
have been influenced by a famous game that
took place the previous year:
8
7 �/ ....� ,,�
6
5
4
3
2
b
5
4
3
2
a
'""" F"'"
a
6
c
d
e
f
g
h
b
c
d
e
f
g
h
33.�g5
Taking the bishop leaves the white king
vulnerable to checks: 33.1Mfe8t�g7 34.1Mfxa8
1Mfd4t 35 .�g3 1Mfxe5t 36.�h3 1Mff5t Black
can give a perpetual.
33 . . . 1Mfxf2
White has no time to catch Black's king, as
the nimble queen can get back in time.
34.1Mfd8t �g7
The Prime Years
192
Black can even try to win with 34 . . .'j;>h7!? but
after 35 .1Mfxa8 c;t>g7 36.'j;>f4! White escapes
from the danger zone in time. 36 . . . 1Mfxh4t
37.i>e3 1Mfxb4 The endgame should be a
draw.
After the text move White should take a
draw. Let us see what happens if he tries to
be too ambitious.
35 .i>f4??
Correct was 35 .1Mff6t, intending to take a
perpetual.
8
7
Karpov simplifies to a winning bishop
ending. Same-coloured bishop endings often
prove difficult for the defending side to hold.
In the present position the prospect of White
being left with an h-pawn and the wrong
bishop makes little difference, as there are
plenty of other pawns that have the potential
to promote.
32 hb7 33.<;t>f4
White's superior pieces count for much
more than Black's extra pawn, which has no
real prospect of advancing.
.•.
33 ... <;t>f8
After 33 . . . c;t>g7 White has the luxury of
choosing between 34.c;t>e3 and 34.i>g5 , both
of which win comfortably.
6
5
4
3
34. <;t>g5 <;t>e7 35.i.e4 i.a8
2
a
b
c
d
e
f
g
8
h
35 . . . c5!! 36.1Mfxa8
36.1Mff6t i>g8 37.1Mfd8t c;t>h7 38.1Mff6 1Mfd4t
39.c;t>g3 c;t>g8 40.ixa8 cxb4 wins.
36 . . . 1Mfxh4t 37.c;t>e3 1Mfd4t 38.c;t>e2 1Mfxe5t
Black wins as he has too many pawns for the
bishop.
7
6
5
4
3
2
8
1
7
a
6
b
c
d
e
f
g
h
36.£3!
Karpov prepares to create a passed pawn.
5
4
3
2
1
a
32.�xb7!
b
c
d
e
f
g
h
36 ...i.b7
The plan of f3 and g4 was not difficult
to find, so it is hard to guess what caused
Laurier to misevaluate the endgame. Perhaps
he thought he could sacrifice the bishop with
36 . . . c5 37.ixa8 cxb4, bur this soon turns out
to be hopeless after 38 .ic6! b3 39.ia4 b2
4o.ic2 with a simple win.
1 992
Anatoly Karpov - Joel Lautier
37.g4!
Karpov continues his plan. It reminded me
of one of Portisch's wins from more than two
decades before.
a
b
c
d
e
f
g
h
39.g4! White gets either a passed pawn or wins
the pawn on h5. He went on to win in Portisch
- Reshevsky, Palma de Mallorca 1 970.
37 J.a8
If 37 . . . hxg4 38.fxg4 White advances the
h-pawn and penetrates with his king to f6,
winning easily.
•••
38.gxh5 gxh5 39.£4 J.b7 40.i.f3 J.a8
41 .I!;>xh5
1-0
In the second round Karpov held a solid draw
against Miles in a Spanish Exchange Variation.
Karpov's next two games were Queen's Indians:
he won with White against Korchnoi, then
drew with Curt Hansen. In Round 5 Karpov
outplayed Beliavsky in a Ruy Lopez Breyer,
then scored a victory with the black pieces
over a young Alexei Shirov in a Nimzo-Indian
(see page 280) . Karpov's next game ended
tragically; he lost on time in a completely
winning queen ending against Kiril Georgiev,
as he did not pay attention to the tournament
regulations and mistakenly believed he would
receive more time on the clock.
In the first game of the second half of the
1 93
tournament, Lautier tried the Scotch Four
Knights but Karpov drew solidly. Then in
Round 9 he returned to his winning ways
against Miles; the game is referenced in the
note to Black's first move in Game 58 of
the first volume. In the next round Karpov
received a slice of good fortune to make up
for the Georgiev game. Korchnoi played well
and built up a winning position, but allowed
Karpov to obtain some counterplay before
blundering and losing.
In Round 1 1 Karpov met Curt Hansen
who tried the Pirc. The Danish grandmaster
managed to equalize, but Karpov persevered
and eventually managed to decide the game
by sacrificing a piece for three pawns in
the endgame. He made a quick draw with
Beliavsky, then won a nice game on the white
side of a Semi-Slav against Shirov. In the final
round he drew quickly with Georgiev, to finish
on l OY2/ 1 4 , a point and a halfclear of Georgiev
who was second.
After a couple of rapidplay events, Karpov
took part in the Tilburg Knockout, where
he suffered a shock defeat in the first round
against Alexander Chernin. In their first game
Karpov got some advantage on the white side
of a Semi-Slav, but Chernin was able to hold an
inferior ending. The second game was a quick
draw, which meant that the contest would be
decided by a rapid tie-break. In the first rapid
game Karpov was winning but blundered and
lost. In the second he blundered an exchange,
and Chernin later forced a perpetual in a
winning position to ensure his passage to the
next round.
Karpov's next tournament was the Alekhine
Memorial in Moscow. He started with a good
win over Timman, but then suffered defeats
against Gelfand and Anand in succession.
Karpov managed to steady his ship and drew
with Yusupov and Shirov, before facing Gata
The Prime Years
1 94
Kamsky in Round 6. At the time of writing
Kamsky has played 1 06 games against world
champions, scoring eighteen wins, fifty three
draws and thirty five losses. Before this game
Karpov was leading their head-to-head score
by two wins to one, with two draws. Their
lifetime score is eleven wins to Karpov, five
wins to Kamsky and sixteen draws.
I Game l8 1
Anatoly Karpov - Gata Kamsky
Moscow 1 992
l .d4 liJf6 2.c4 g6
In their first meeting Kamsky played the
Queen's Indian and Karpov nicely beat him.
3.liJa i.g7 4.g3
The second time Karpov had the white pieces
against Kamsky, he played the main exchange
variation but lost.
4 ... c6 5.i.g2 d5 6.cxd5 cxd5 7.liJ c3 0-0
8.liJe5 e6 9.0-0 liJfd7 10.f4 liJc6 1 1 .i.e3
liJ b6
Kasparov used this move to score two draws
in the 1 987 Seville match. Karpov also faced
1 1 . . . ttJdxe5 and 1 1 . . . f6 in other games, scoring
two wins and two draws.
12.i.f2 i.d7
Kamsky follows the first game of the Seville
match. In the third game Kasparov switched
to 1 2 . . . ttJe7.
13.e4
This move either gains space or opens the
position. Without it, it will be hard for White
to play for a win.
13 ... liJ e7
Taking on e4 would only help White.
13 . . . ie8!? is an interesting alternative though.
It has been played much less frequently, but
based on the games appearing in the database
it appears quite reliable.
14.liJxd7
Karpov gets the advantage of the two bishops
and gains space. Exchanging the opponent's
bad bishop could have certain drawbacks, on
the other hand if the position eventually opens
up then White's light-squared bishop will have
great potential.
14...�xd7 1 5.e5 �ac8
The Seville game continued: 1 5 . . . l'!fc8
1 6.l'!c1 if8 1 7.if3 l'!c7 1 8.b3 l'!ac8 1 9.'1Wd2
ttJ c6 20.Wb2 a6 2 1 .ie2 We7 22.ttJ b 1 ttJb4
Black has equalized and the game was soon
drawn, Karpov - Kasparov, Seville ( 1 ) 1 987.
Kamsky moves the other rook to the c-file,
but it makes little difference as Black will soon
double his rooks in any case.
1 6.�cl a6
Black can also delay this move as Kasparov
did, but the pawn move is useful and will
generally be played at some point.
17.b3 �k7 18.�d2 �fc8
8
7
6
5
4
3
2
1
a
19.94!
b
c
d
e
f
g
h
1992
195
Anatoly Karpov - Gata Kamsky
Karpov starts playing on the kingside. Against
Kasparov he tried to play on the queens ide but
got nowhere. It would be interesting to know
if Karpov considered the more ambitious plan
in the Kasparov game. He may have preferred
not to risk it against a rival who played so well
in complicated situations, especially in the first
game of the match when a loss with the white
pieces would have made for a dreadful start.
In any case, by the time of the present game
Karpov must have analysed the idea carefully
at home.
19 i£8 20.Wfe3
White can also exchange rooks with 20.tLl e2,
for instance 20 . . . l:!xc l 2 1 .l:!xc l l:!xc l t 22.l2Jxc l
I2lc6 23.l2Jd3 when he maintains a slight edge
thanks to his space advantage and bishop pair.
Karpov's approach is more ambitious.
•.•
20 l2Jc6
Kamsky transfers his pieces to the queenside
in order to exert pressure there; on the other
hand he leaves his king a bit vulnerable.
••.
20 ... lt>h8!? vacates the g8-square and allows
Black to keep his knight close to the king,
while still enabling the bishop to go to the
queenside. 2 1 .ih4 l2J g8 22.l2J e2 l:!xc l Black is
close to equalizing and eventually drew after a
long fight, Nalbandian - Khachiyan, Yerevan
1 99 9 .
21.£5
Although the pawn assault looks menacing,
Black's fortress is not easy to breach, and
White will have to show a good deal of skill
and finesse to achieve anything.
21 ...ia3
2l . . .exf5 is playable but risky, and after
22.gxf5 �xf5 White can play 23.�h l or
23.1Wg3 with promising compensation for the
sacrificed pawn.
However, it should be noted that 23.ih3?!
is less effective, and after 23 . . . �h5 24.ixc8
ih6 25 .�h3 ixc l 26.�xh5 gxh5 Black is not
worse.
22.l:!cdl ttl b4
Kamsky raises the stakes by bringing
yet another piece to the queenside, while
provocatively leaving his king with almost no
defenders.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
23.Wfh6
Karpov brings the queen into the attack,
while exploiting the fact that his knight cannot
be taken due to mate in two.
If 23.l2J b l then 23 . . . l2Jc2! is strong.
23 Wfe8
This move turns out to be a loss of time.
Please do not forget that Karpov and Kamsky
were the trailblazers for this position, so it is no
surprise that subsequent analysis highlighted
a few subtle refinements. Three other moves
deserve consideration, and two of them have
been tested in subsequent games.
•••
23 . . . �e7!? has not been played but deserves
attention. If the players follow the same
path as in the game, then Black will end up
saving a tempo. 24.l2J b l ib2 25 .ih4 (25.a3
The Prime Years
196
It'l c2 26.l'!d3 'lWfB 27.'lWh4 It'lxa3 2B.l'!h3 'lWg7
29.lt'lxa3 ixa3 30.ie3 ie l Black is defending
himself.) 25 . . . 'lWfB 26.'lWd2 l'!c2 27.'lWe3
White's pieces look menacing, yet it is not
clear whether he can hurt Black.
23 . . . lt'l d3 24.lt'lxd5 It'lxd5 2 5 .l'!xd3 ifB 26.'lWg5
h6?! (26 . . . l'!c2! looks like an improvement, and
leads to a complicated position where Black has
some compensation.) 27.fXe6! 'lWxe6 2B.ixd5
'lWxd5 29.'lWe3 l'!c2 30.a4 ib4 3 1 .'lWf3 White
is a safe pawn up, Maherramzade - Vydeslaver,
Halle 1 995.
23 . . . lt'l c2 24.lt'le2 (Perhaps White should
deviate with 24.lt'l b l !? ifB 2 5 .'lWf4 It'l b4
26.a3 It'l c2 27.a4 when he keeps a slight plus.)
Karpov evaluates the position as clearly better
for White, but this seems too optimistic, and
in the following game Black did okay: 24 . . . ifB
25 .'lWf4 It'l b4 26.a4 a5 27.'it>h 1 l'!c2 This
unclear position eventually resulted in a draw,
Onoda - Soegaard, e-mail 1 99B.
24. .!tJ b l
Karpov drags the bishop t o b 2 . I f 24.lt'le2
then 24 . . . l'!c2 25.lt'lf4 ie l !? is an interesting
way to slow White's play.
24 ...i.b2
a
b
c
d
e
f
g
h
25.\Wd2!?
Karpov wittily retreats the queen. The idea is
to draw the knight to the c2-square, where it
blocks Black's rook invasion on the queens ide.
25 ... tlJc2
Kamsky accepts the invitation. Thanks to
some unusual tactics Black had other options
as well.
25 . . . a5!? 26.a3
26.'lWxb2? l'!c2 27.'lWa3 l'!xa2 wins the queen.
26 . . . l'!c2
If 26 . . . lt'l c2 27.'lWxa5 White takes a pawn for
insufficient compensation.
27.'lWe 1
a
b
c
d
e
f
g
h
27 . . . 'lWb5!?
If 27 . . . lt'l a2 2B.'lWxa5 It'l d7 29.ie3 l'!e2
30.l'!d2 Black has some play for the pawn,
but overall White's position is preferable.
27 . . . lt'l c6 2B.l'!d2 l'!el 29.'lWe2 l'!xfl t 3o.ixfl
iel 3 1 .l'!c2 ig5 32.'lWb5! Suddenly Black's
queens ide is a bit loose; it is remarkable that
White can obtain the initiative on the side
where Black had so many pieces.
2B.axb4 l'!e2 29.'lWxe2 'lWxe2 30.bxa5 'lWxg4
30 . . . lt'l d7?? 3 1 .l'!d2 wins as Karpov pointed
out.
3 1 .fXe6 fXe6 32.axb6 l'!c2 33.l'!d2 l'!xd2
34.lt'lxd2 ixd4 3 5 . lt'l f3
White is better, as he has a lot of material for
the queen.
1 992
1 97
Anatoly Karpov - Gata Kamsky
There is a second way for Black to leave the
bishop en prise:
25 ... �e7!?
when both 3 1 .Wf2 and 3 1 .�h l force Black
to take a perpetual. The main line should
not affect the outcome, although it does set
a small trap.
29.j,xg5 �xg5 30.Wxb4 Ei:xg2t
The attempt to win with 30 . . . Ei:cc2?! might
backfire: 3 1 . ltJ d2 Ei:xd2 32.Ei:xd2 �xd2
(32 . . . Ei:xd2 33.Ei:f3) 33.�xd2 Ei:xd2 34.fxe6
fxe6 3 5 .Ei:f6 White has slightly better
chances.
3 1 .Wxg2 �xg4t
8
a
b
c
d
e
f
g
7
h
26.�xb2
This should lead to a draw.
26.�h l ? Ei:c2 Black gets good play.
26.f6 Wf8 defends conveniently, and White
has lost the option of opening the f-file.
26.j,e l ltJ c2 27.j,f2 amusingly gives Black
more than one way to repeat moves, with
27 . . . ltJ b4 or 27 . . . Wb4 28.�f4 �e7.
If White wants to continue the fight then
he may have to try 26.a3!?, although after
26 . . . ltJ c2 27.a4 ltJ b4 (27 . . . �b4!?) 28.a5
�c2 29.We l ltJ d7 Black is very much in the
game.
26 . . . Ei:c2 27.�a3 Ei:xa2 28 .j,h4!
6
5
4
3
2
a
b
c
d
e
f
g
h
32.Wf2!
As a result of Black sacrificing his g-pawn on
move 28, White cannot play 32.�h l ? due
to 32 . . . �e4t 33.Wgl �h8! with a strong
attack. The text move easily secures a draw
though.
32 . . . �f4t 33.We2 We4t
White is unable to hide from the checks.
8
7
6
5
4
a
b
c
d
e
f
g
h
28 . . . g5!?
Black can force a draw immediately with
28 . . . �xh4 29.�xb4 Ei:xg2t 30.�xg2 Wxg4t
3
2
1
a
b
c
d
e
f
g
h
198
The Prime Years
26.ct!?hl!
Karpov hits upon a clever prophylactic plan
which enables him to tidy up his kingside,
safeguard his monarch and open the f-file for
his rooks.
26 ...�e7
Now Black's queen and knight will have the
option of using the b4-square.
ambitious he should play 29.'lMff4, leading to
a tense situation where Black's position should
be playable.
28J'H'3
Improving the rook is the logical follow-up
to White's previous two moves.
8
7
27.�gl !
Karpov completes his regrouping.
6
5
4
3
2
1
a
a
b
c
d
e
f
g
h
27... llJd7
Neither side has a direct threat, so Kamsky
decides to improve his worst-placed piece. It is
a sensible decision, although some other ideas
were playable as well.
b
c
d
e
f
g
h
28 ...�b4
The position remains in an odd state of
dynamic equilibrium; both sides have serious
advantages in one area or another, but
ultimately they should cancel each other out.
Apart from Kamsky's move there were a few
alternatives worth considering.
2s . . . lLl b4?! 29.a3 lLlc2
27 . . . lLl b4!? is witty, but perhaps Black is being
too clever for his own good. 2S.a3 (If2S.'lMfxb2?
E!:c2 the queen will be caught.) 2s . . . lLl c2 29.E!:f3
�xa3 (29 . . . lLl d7!? is another idea) 30.E!:h3
'lMffS 3 1 .lLlxa3 lLlxa3 32.E!:fl Black was able to
snatch a pawn but he is in real danger on the
kingside.
27 . . . 'lMfb4! ? 2S.'lMfh6 'lMffS Black challenges
his opponent to find something better than
a repetition of moves, as a queen exchange
would not be in White's interest. 29.'lMfg5 can
be met by 29 . . . 'lMfe7, so if White is feeling
a
b
c
d
e
f
g
h
In this position, the decision to invest two
tempos to provoke a pawn move on the
queenside looks too optimistic.
1 992
30.Eldfl !
After the less accurate 30.Elh3?! Wf8 3 1 .Elfl
Wig7 32.a4 b5 the position is balanced.
30 . . . Wd8
Preparing to meet the attacking plan of Elh3
and Wh6 with . . . cj;Jh8 and . . . Wg8.
If 30 . . . cj;Jh8 3 1 .fXg6 fXg6 32.Elf7 the rook on
the seventh rank is unpleasant for Black.
a
b
c
d
e
f
g
28 . . . b5!?
b
c
d
e
f
g
This is another interesting idea. Just as in
the game, it is not easy for White to break
through on the kingside.
29.Eldfl
This looks like the most logical attacking
move.
Black manages to defend himself after
29.Wh6 Wf8 30.Wd2 We7 or 29.Elh3 Wf8
30.Elfl Wg7.
29 . . . Wb4
29 . . . .ial ? 30.Elh3 is too dangerous.
30.Wh6 Wf8
h
3 1 .a4!?
With no direct attacking win, White takes a
moment to safeguard his a-pawn.
3 1 .fXg6 fXg6 32.Elf7 ltJ b8 defends.
3 l . . .a5
This is not forced, but it looks logical for
Black to secure the b4-square for his pieces.
And though it is not a serious threat in the
present position, a future b4 and Elb3 has the
potential to be unpleasant.
32.Elh3 cj;Jh8 33.Elff3 Wg8 34.ltJc3!
White surprisingly gets the upper hand on
the queenside.
a
199
Anatoly Karpov - Gata Kamsky
h
a
b
c
d
e
f
g
h
3 1 .Wg5!?
Inviting an interesting tactical skirmish,
but it turns out that Black has enough
resources.
The alternative is 3 1 .Wf4 Wg7 32.h4 to
continue the attack. However after 3 1 . . . We7!
it is not easy to increase the pressure, for
instance 32.Elh3 Wf8 defends.
3 1 . . . ltJxd4!
Despite the shaky appearance it turns
out that Black is holding his own in the
complications.
After 3 1 . . .Wd8 32.Wf4 Wf8 33.h4 White
keeps his play on the kingside alive,
nevertheless Black's position is quite
solid.
32 ..ixd4 .ixd4 33.fXe6 ltJxe5 34.exf7t ltJxf7
35 .Elxf7 Elxf7 36 . .ixd5 Elcc7 37.Wf4 .ib6
38.ltJc3 Elcd7
Black maintains the balance.
The Prime Years
200
3 1 . . .E1cl t is refuted by 32.�fl !. Interestingly
Karpov misses this winning move in his Chess
Informant analysis.
32.E1fl �xe5 33.fxg6
Black has little for the piece.
a
b
c
d
29".Wlffi 30.Wlg5 Wig?
30 . . . �d8 would not have achieved much
after 3 1 .�f4. Kamsky's move is sensible,
although it should be noted that his queen is
now cut off from the queenside, as retreating to
f8 would expose her to a lethal pin after fxe6.
e
f
g
h
29.Wlh6
Karpov naturally avoids a queen exchange
and takes the opportunity to shift his heavy
artillery to the vicinity of Black's king.
29.�f4!?
This move was also possible, as the following
tactical reply favours White.
29 . . . �xd4?
Black should prefer 29 . . . �f8! with similar
play to the game, although White has saved
a bit of time with his queen.
30.�xd4 tt:\xd4
3 1 .Wldl
Karpov takes away the b4-square from the
knight.
3 1 ".b6
Kamsky prepares to regain control over the
b4-square.
32.E1dfl
The rook no longer needs to defend the d4pawn, so Karpov recruits it for more active
duties.
32".a5 33.h4 � b4
The knight vacates its outpost for the rook.
34.a3
Capturing the bishop would have cost White
his queen.
8
7
a
b
c
d
e
f
g
h
3 1 .E1xd4!
The finesse 3 1 .a3!? results in a smaller
advantage for White: 3 1 . . .�c5 32.E1xd4
�cl t 33.E1fl �xf4 34.E1dxf4 White is better
but Black has some drawing chances.
3 1 . . .�e l t
6
5
4
3
2
1
a
b
c
d
e
f
g
h
20 1
Anatoly Karpov - Gata Kamsky
1 992
34 J�c2
34 . . . lLlc2!?
Having provoked the a-pawn into moving,
the knight can cheekily return to its previous
square. It leads to an interesting fight,
although White can keep the upper hand
with the help of some accurate moves.
35.h5! g5
35 . . . gxh5?! is risky: 36.gxh5 (There is also
36.fxe6! ? fxe6 37.gxh5 when Black's king
is more exposed, for instance 37 . . . lLl xd4?
3 8.EH7 wins.) 36 . . . lLl xd4!? 37.�xd4! White
gets enough material for the queen this
way. 37 . . . l':k2 3 8 .�xb2 )"1xd2 39.lLlxd2 Wg5
(39 . . . )"1c2? 40.�h3) 40.fxe6 fxe6 4 1 .�h3
Black is living dangerously.
3 5 . . . lLlxd4 36.�xd4 )"1c2 37.fxe6 fxe6 3 S .We3
gxh5 39.)"10 WxO Now it is Black's turn to
give up his queen for several pieces, but it is
not so favourable here. 40.)"1xO cj;JxO 4 1 .lLl d2
Black does not have full compensation as his
king is uncomfortable.
..
Premature is 36.fxe6 fxe6 37.a4 (37.Wd3
)"1fS) 37 . . . h6 3S .Wd3 lLl b4! when Black
chases the queen from her ideal home.
36 . . . lLl fS
36 . . . h6? 37.fxe6 fxe6 3S .)"10 wins.
36 . . . cj;JhS?! does not lose outright, but after
37.fxe6 fxe6 3S .)"10 WgS Black can hardly
feel happy.
a
b
c
d
e
f
g
h
37.)"13f2!!
Suddenly White switches to his attention to
the c2-knight and the queenside.
37 . . . �a 1
After 37 . . . b5 3S.)"1d2 b4 39.a4 )"1c6 40.�f3 �al
4 1 .Wb5 White invades on the queenside.
38.)"1d2 a4 39 .Wb5 axb3 40.Wxb3 )"1c4
4 1 .Wxb6
White has won a pawn for very little.
3S.Wf4 lLlc6 36 ..th3
a
b
c
d
Karpov keeps building his position and sends
a message that he may go after the e6-pawn.
e
f
g
h
36.�d3!
The queen stands superbly on d3, defending
the d4-pawn and setting up the threat of
fXe6 followed by )"10, as well as keeping an
eye on the c2-knight.
Less convincing is 36.h6 Wxh6t 37.)"1h3 Wg7,
when White has some compensation for the
pawn but nothing conclusive. Besides, it is
by no means clear than an attack on the h-file
will be more effective than one on the f-file.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
202
The Prime Years
36 tL1d8?
37 . . . fXe6
•••
Kamsky defends the e6-pawn in advance, but
the problem is that he relinquishes the pressure
on the d4-pawn. The way Karpov exploits it is
majestic.
After 36 . . . lLl dxe5?! 37.dxe5 �xe5 3S.Wle3
Black does not really have enough for the
piece.
Correct was:
36 . . . b 5 !
Th e idea behind this move i s t o free the c4square for a rook to put more pressure on
the d4-pawn. The immediate 36 . . . a4! ? might
also be considered.
37.fXe6!
White should keep his opponent busy with
defending.
37.h5 g5 ? 3S .Wle3 �xd4 Black wins.
If White plays like in the game a horrible
surprise is waiting for him: 37.�e3?
a
b
c
d
e
f
g
a
b
c
d
e
f
g
h
3 S .g5
3 S .Wlf7t Wlxf7 39.E!:xf7 lLl fS 40.�e3 �xd4
4 1 .�h6 �c5 42.E!:g7t <j;>hS Black seems to
be able to keep his position together.
3S . . . E!:eS 39.Wlg4
39.Wlf7t ?! Wlxf7 40 .E!:xf7 lLl fS does not
achieve much for White.
39 . . . Wle7 40.h5 a4
The position remains double-edged.
h
37 . . . g5!! 3S.hxg5 (3S.Wlxg5 Wlxg5 39.hxg5
lLlxd4-+) 3S . . . lLl dxe5! 39.E!:g3 lLl d3 40.Wlf3
lLl c 1 White is in big trouble.
37.E!:e l a4 (37 . . . b4 3 S . axb4 lLl xb4 is also
okay for Black) 3S .�f1lLl dxe5! 39.dxe5 �xe5
40.f6 (40.Wle3 ? �d4!) 40 . . . �xf4 4 1 . fXg7 �d6
Black has good play for the piece.
37.E!:d l is possible, although White would
ideally prefer not to resort to such purely
defensive moves. 37 . . . a4 3S.bxa4 bxa4
39 .�f1 lLl a5 The position is complicated.
a
b
c
d
e
f
g
h
37 ..ie3!
This is an especially sly move, which plays
for a remarkable idea which will be revealed
shortly. The point of the move is to take away
the c 1 -square from the b2-bishop. 37.E!:3f2 is
less good due to 37 . . . E!:xf2 3S .E!:xf2 �c 1 when
Black is in the game.
37... b5
Anatoly Karpov
1 99 2
Kamsky probably spotted Karpov's wicked
plan and tries to help the b2-bishop.
37 . . .�c3 38JBf2! (After 3 8 . ltJ xc3 2"18xc3
39.�d2 White is somewhat better thanks to
his two bishops and extra space, but he can
play for more.) 38 . . . �a l
-
203
Gata Kamsky
queenside with the plan of exchanging rooks
and targeting the misplaced bishop on b2.
Knights and pawns can easily get into trouble
if they stray too far from their camp, but this
rarely happens to a bishop due to its great
mobility. The present position is an exception,
and the closed centre makes it impossible for
the bishop to escape in a satisfactory way.
Kamsky was hoping that the bishop would be
able to target the d4- and eS-pawns, but he
misjudged the fine line and it turns out that
these pawns will continue to stand firm and
prevent the bishop from escaping.
38 b4
...
a
b
c
d
e
f
g
h
39.a4! Vacating the a3-square for the knight.
(39.2"1xc2 2"1xc2 40.2"1£1 is advantageous for
White, but not yet decisive.) 39 . . . 2"1xf2
(39 . . . <j;>h8 40.ltJa3 2"1xf2 4 1 .Wxf2 �c3 42.2"1c 1
!"le6 43 .ltJbS �b4 44.2"1xc6 ltJ xc6 4 S . ltJ c7+-)
40.'\Wxf2 ltJ c6 4 1 . ltJ a3 �c3 42.ltJ bS �b4
43.�h6! Wxh6 44.fxe6 fxe6 4S .Wf7t �h8
46.'\Wxd7 ltJ e7 47.Wxe6 White is winning.
38 . . . i.a l
This move also fails to solve the problem of
the lonely bishop.
39 .Wh2!
39.2"1xc2 2"1xc2 40.2"1£1 2"1x£1 4 1 .Wx£1 ltJ c6
(4 1 . . .hS ? 42. f6 Wh8 43.gS+- What a
position! So many black pieces are at the
edge of the board and his queen looks
especially sad.) 42.�f1 gxfS 43.�xbS ltJdb8
44.gS White keeps the advantage, but Black
is still in the game.
39 . . . 2"12c6
39 . . . 2"1xf2 40.Wxf2 ltJ c6 4 1 .hS gS 42.2"1c1
Wf8 43 .i.f1 b4 44.i.a6 2"1c7 4S .Wa2 Having
made the right preparations, White is ready
to harvest the crop.
8
7
,,, ". J//////•• :-;:-:-:-.//'
6
5
4
3
a
b
c
d
e
f
g
h
38JHfl!
Just as he seems to be poised for a kingside
breakthrough, Karpov switches to the
2
a
40.2"1a2
b
c
d
e
f
g
h
204
The Prime Years
40.hS!? can also be played, for instance:
40 . . . a4? (40 . . . gS 4U'1a2 .ic3 transposes to
the main line below) 4 1 .h6 Wf8 42.bxa4
bxa4 43.fxe6 tLlxe6 44.gS Black's position
falls apart.
40 . . . .ic3 4 1 .hS gS 42.f6 Wh6
42 . . . Wf8 43 . .ixgS .ixd4 44.Wd2 WcS
4S . .ih6 wins.
43 .We2 b4 44.tLlxc3 Ei:xc3 4S . .ig2
Black's situation is becoming desperate, as
the gS-pawn will soon fall.
39.axb4 axb4
8
7
6
5
8
4
7
3
6
2
5
4
a
3
b
c
d
f
e
g
h
43.\Wc2!
2
1
principles, but the downside is that the bishop
is more exposed to being taken: 43.f6 Wf8
44.Wc2 tLlxeS ! ? (44 . . . .ie 1 4S .Wc7+-; 44 . . . tLl b8
4 S . tLl xc3 bxc3 46.Wxc3 Black has no chance
to survive with the pawn deficit.) 4S.dxe5
.ixeS 46.gS Wd6 47.WcS WxcS 48 . .ixcS .ig3
49.tLld2 .ixh4 SO.tLlf3 The ending is easily
winning for White.
a
b
c
d
e
f
g
h
40.Ei:xc2!
40.WgS was not a bad alternative, but the
game continuation gives Black fewer chances
to resist. 40 . . . gxfS 4 1 .Ei:xc2 WxgS 42 . .ixgS
Ei:xc2 43.gxfS .ixd4 44 . .ixd8 tLl cS White only
has a few pawns remaining, which gives Black
some drawing chances.
It is remarkable how many times Karpov
switches between flanks in this game. Now
Black has no good defence against the invasion
on the c-file.
43 ... tLlxe5!?
Sacrificing the knight is a reasonable practical
try, but Black's pieces are too badly coordinated
for it to have much chance of succeeding.
43 . . . tLlb8
40 ... Ei:xc2 4UU2
Karpov removes the second rook from
around the bishop.
41 ..J:hfl 42.\Wxfl .ia3
The bishop is saved but it is out of play on a3
while the c-file becomes available for the white
queen.
42 . . ..ic3 would be more desirable on general
a
b
c
d
e
f
g
h
Anatoly Karpov
1 992
44.mh2!!
This lovely prophylactic move is directed
against the plan of . . . gxf5 and . . . Wlg3 . White
intends h5 combined with .tfl -d3 , which
seems to me to be the most convincing
winning plan.
44.WlcB ltJ bc6 4 5 . .tg5 WlfB 46 . .txdB
(46.ltJd2 exf5 47.gxf5 .td Black is still alive.)
46 ... ltJxdB 47.ltJd2 .tb2 4 B . ltJ f3 According
to Karpov White is winning, but I do not
see a definitive winning plan at this stage,
although Black is certainly on the ropes.
Ftacnik suggests 44.f6 WlfB 4 5 . ltJ d2 which
should be good enough to win eventually,
but the main line looks clearer to me.
44 . . . WlfB
44 . . . ltJ dc6 4 5 . ltJ d2 ltJ dB 46.h5 is winning.
44 . . . ltJ bc6 4 5 . f6 WlfB 46 ..tfl WleB (46 . . . ltJ bB
47.h5 ltJ dc6 4B . .td3+-) 47.h5 ltJ a5 4B.hxg6
hxg6
-
Gata Kamsky
205
46.f6
Exchanging is also good enough: 46.hxg6
hxg6 47.fxe6 fxe6 4 B . ltJ d2 ltJ e7 49 . .tfl
WleB 5 0 . .tg5 Black's resistance will soon be
broken.
46 . . . g5
46 . . . ltJ a 5 47 . .tfl ltJ db7 4B . .td3+47 . .tfl h6 4B . .tb5 WleB 49.ltJxa3 bxa3 50.Wld
a2 5 1 ..txg5 a l =WI 52.Wlxa l hxg5 53.Wld
Black can resign.
44 .dxe5 V«xe5
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
45.Wlc8!
a
b
c
d
e
f
g
h
49 .mg3! White's queen will switch to the
h-file, with mate to follow shortly.
45 .h5 ltJbc6
After switching between the kingside and the
queenside several times, Karpov commences
his final assault using another part of the
board, the eighth rank.
45 V«e4t
.••
Black has no defence, for instance 4 5 . . . m g7!?
(45 ... Wlxe3? 46.WlxdBt mg7 47.f6t <tt> h 6
4 B.WlfB#) 46.f6t! Wlxf6 47 . .th6t! mgB 4B.Wld7
Wlf3t 49 . .tg2 with mate to follow.
46.i.g2 V«xbl t 47.c;t,>h2 i.b2
The bishop defends the diagonal, but it is
not enough to save the king.
48.V«xd8t c;t,>g7 49.f6t!
Karpov finds the most accurate solution.
a
b
c
d
e
f
g
h
206
The Prime Years
49 i.xf6
•••
a
b
c
d
e
f
g
h
50.i.h6t!
The black king is drawn to the edge where it
is fatally trapped.
50 @xh6 5 1 .Y;Vxf6 Y;Vc2 52.g5t
•••
52.Y;Vxfl Wb2 5 3 .Wf8 t was even faster.
52 ... @h5 53.@g3
5 3 .ctfh3 ?? Wf5 t was no good, but 5 3 .Wf3t
ctfxh4 54.Wf4t ctfh5 5 5 .'it>g3 would also have
led to mate.
53 Y;Vc7t 54.@h3
1-0
•••
Finally Kamsky resigned. It was a grand
game, rich in original ideas.
In the last round Karpov agreed a quick
draw with Salov to finish on 3 Y217, which
was enough to share 4th-6th place. Anand
and Gelfand won the tournament, showing
that the new breed of super-grandmasters
were good enough to outshine former world
champions in some tournaments, although it
would take a long time for them to take over
on the rating list.
For his final tournament of 1 992 Karpov
travelled to Baden-Baden, for a tournament
featuring several prominent German players
and some other top grandmasters. Karpov
started by beating Bonsch with the white
pieces, then he scored an excellent win with
Black against Hubner, outplaying his rival in
an equal endgame. In Round 3 Kindermann
became Karpov's third victim from the host
nation - part of this game can be found on
page 325 in the notes to Game 4 1 .
In the next game Karpov drew quickly with
Psakhis, then in rounds 5 and 6 he continued
his harsh treatment of the local players, winning
two games that are referenced elsewhere in the
present work: his Round 5 victory over Lobron
was mentioned in the notes to Game 26 in the
present chapter, and his Round 6 win over
Hertneck can be found in the notes to Game
6 1 of the first volume.
In Round 7 Karpov faced Yusupov, who by
that time was living in Germany. Karpov was
pressing for a long time but eventually had to
settle for a draw. Then in Round 8 he defeated
Lautier after a long fight. In Round 9 Karpov
dropped his first draw of the tournament
against a native German, Christopher Lutz,
who got a slight advantage after a 4.Wc2
Nimzo-Indian, but was content to take a
draw. In the next round Karpov returned
to his winning ways against the five-time
East German Champion Knaak, eventually
grinding out a win from a rook ending with
an extra a-pawn.
Karpov's final opponent of the tournament
and the year was Matthias Wahls, a German
grandmaster with whom he had drawn twice
previously. This was the last time the two players
met over the board. Wahls played a total of five
games against the world champions, drawing
three and losing two.
1 992
I Game l91
Matthias Wahls
-
Anatoly Karpov
Baden-Baden
207
Matthias Wahls - Anatoly Karpov
their 1 996 World Championship match. The
players contested this variation twice, scoring
one win each. Karpov's victory is examined in
Game 43 of the present book.
1 992
1.e4 c6
On the one previous occasion when he faced
the German grandmaster with the black pieces,
Karpov opted for 1 . . .e5 but had to settle for a
draw after twenty four colourless moves. Wahls
always favoured the sharp Panov variation
against the Caro-Kann, so it is safe to assume
that Karpov was playing for a win from the
outset here. It was a brave decision, as despite
his superb performance thus far, he had not yet
won the tournament outright, as Lutz had a
chance to tie for first in the event that he could
win and Karpov lost. Karpov's decision to play
for a win may have partially been caused by
the fact that Wahls had not been having a good
tournament. Karpov may also have felt some
pressure to elevate his rating; he was rated
almost 200 points higher than Wahls, and by
the end of 1 992 Ivanchuk was closing in on
the number two spot in the world rankings.
2.d4 d5 3.exd5 ad5 4.c4 ttJf6 5.liJc3 e6
6.tLla .ib4
Prior to this game Karpov had only played
this move once back in 1 980 when he drew
with Timman in just seventeen moves. Over
the course of his career he played both this
and the alternative 6 . . . fie7 in roughly equal
measure.
7.ad5 ttJxd5 8 .id2
•
According to the database Wahls had played
8.Wc2 previously.
8 ... ttJc6 9 .id3 0-0 10.0-0 .ie7 1 l.a3
Preventing . . . tLl b4 is a typical idea, and one
of the main lines in the position. Kamsky
prepared I I .We2 to use against Karpov in
1 l ....if6 12JWe2
The main line is 1 2 .Wc2 which according
to the database was first played by Kavalek.
1 2.fie3 is also frequently played.
a
b
c
d
e
f
g
h
12 ... ttJxc3!?
Karpov introduces a novelty, which leads to
a complex situation involving hanging pawns.
Black has two safe routes to equality, but
Karpov was clearly looking to battle it out in
this game.
After the forcing sequence 12 . . . tLlxd4 1 3 .tLlxd4
fixd4 1 4.fixh7t �xh7 1 5 .We4t �g8 1 6.Wxd4
Black scores fine with all the knight moves
1 6 . . . tLl xc3, 1 6 . . . tLl b6 and 1 6 . . . tLl e7. Almost all
games have resulted in draws from here.
In 1 984 Korchnoi introduced another idea in
the form of 1 2 . . . fixd4!? 1 3 .tLlxd5 Wxd5. He
eventually lost to Pia Cramling, but Black's
position at this stage is quite reliable.
•
13.bxc3 g6 14 .ie4
•
White inhibits the development of the c8bishop to bY, while planning tLle5 and f4.
208
The Prime Years
1 4 .�h6 has been seen more frequently. Here is
an example involving a world-class player (the
game actually reached this position via a slightly
different move order) : 1 4 . . . 2"i:e8 1 5 .2"i:fd l �d7
1 6.ttJe5 ttJ xe5 1 7.dxe5 �h8 Black has a safe
position, Collinson - Speelman, Birmingham
1 999.
White can also try: 1 4.2"i:ab l b6 1 5 .ttJ e 5 (After
1 5 .�e4 �b7 Karpov prefers Black.) 1 5 . . . ttJ xe5
1 6.dxe5 �xe5 ( 1 6 ... �g7!?) 1 7."lMfxe5 (i7."lMfe4!?)
1 7 . . . "lMfxd3 1 8 .�h6 f6 1 9 ."lMfc7 2"i:O 20."lMfc6
�b7 2 1 ."lMfxe6 "lMfe4 Black is safe despite his
somewhat exposed king.
14....id7 15.�abl
Though this move takes place o n the
queenside, White's focus remains firmly on
the other wing - he only wants to provoke
. . . b6 or drag the queen to c7, while setting
up a possible rook transfer from b5 to the
kingside.
Playing in the centre is not dangerous: 1 5 .2"i:fd 1
2"i:c8 1 6.�h6 2"i:e8 1 7.c4 At this point Black can
break up White's centre with 1 7 . . . e5 1 8 .dxe5
ttJxe5 , or win some squares around the hanging
pawns with 1 7 . . . ttJ a5 1 8 .c5 �c6, with good
chances in both cases.
Karpov chooses the right way to safeguard
the b-pawn. Even though his bishop cannot get
to b7 in the immediate future, the availability
of this square might become important later
in the game.
The inaccurate 1 5 . . . "lMfc7?! walks into a big
tempo-gainer: 1 6.g3! (if 1 6.c4 2"i:ac8 Black
is fine) 1 6 . . . ttJa5 1 7.�f4 "lMfc8 1 8 .c4! White's
pieces and pawns control a lot of important
squares in the centre.
16.ltJe5
White proudly pOSItIons his knight on a
strong central outpost. Some other moves were
also possible.
1 6.c4 2"i:c8 ( 1 6 . . . 2"i:e8!?) 1 7.d5 exd5 1 8 .cxd5
ttJ d4 The d-pawn is securely blockaded and
according to Karpov the position is equal .
1 6.�f4
The bishop stands well on the h2-b8
diagonal.
1 6 . . . 2"i:c8 1 7."lMfa6! ?
Th e queen's lunge i s a n interesting way to
cause problems.
If 1 7.2"i:fd l ttJa5 1 8 .2"i:bc l �a4 Black generates
enough play against the hanging pawns.
8
7
6
5
4
3
2
1
a
a
15 ... b6!
b
c
d
e
f
g
h
b
c
d
e
f
g
h
1 7 . . . ttJa5!?
This positional sacrifice seems fairly
promising, although it is by no means
forced.
1 992
209
Matthias Wahls - Anatoly Karpov
The solid 17 . . .We7 is also fine: l S .:B:fe 1
:B:fdS Black plans to finish his development
by dropping his bishop back to eS.
1 9.d5 (I 9.:B:e3 ieS is solid enough)
1 9 . . . exd5 20.ixd5 ie6 Black has no
problems.
1 8:�xa7 ic6!
1 8 ... :B:xc3?! is worse: 1 9 .Wxb6 Wxb6 20.:B:xb6
:B:dS (20 . . . :B:fcS 2 1 .tLle5) 2 1 .:B:a6 tLl b3 Black
can probably hold this position, but it is still
unattractive for him.
1 9.tLld2 ixe4 20.tLlxe4 tLl c4
Black has good compensation. Karpov once
sacrificed a pawn under similar conditions
against Taimanov - see page 3 5 7 of the first
volume.
16 :B:c8 17.£4
...
In his Chess Informant analysis Karpov calls
this move dubious, without suggesting an
alternative. I doubt that it is any weaker than
other moves, although Karpov was probably
happy to see it at the board, as it keeps the
position complex in a way that would enable
his tremendous playing strength to shine
through.
Eight years later White deviated but got
nowhere: 1 7.tLlxd7 Wxd7 l S .g3 ( I S .:B:fd l
8:fdS) I s . . . ig7 1 9.h4 e5 20.dxe5 tLl xe5
2 1 .:B:fd l We6 22.ie 1 tLl c4 Black obtained a
slight plus, Zlatic - Bodiroga, Vrsac 2000.
White can also try 1 7.tLlg4 ig7 l s .id3
( l 8.if3 tLla5) I S . . . We7 when the position is
roughly equal.
17 ig7 18J�be1
...
Wahls concentrates on the kingside.
He could have changed tack with l S .Wa6,
but after l S . . . Wc7 1 9 .tLlxc6 ixc6 20.ixc6
Vf1xc6 2 1 .Wxa7 b5 22.Wa5 :B:bS Black has
fine compensation. 23 .Wb4 (Otherwise Black
might improve his position with . . . :B:b6 and
. . . :B:aS .) 23 . . . WaS 24.:B:fc 1 :B:fcS Black has a
lovely position.
a
b
c
d
e
f
g
h
18 :B:c7!
•.•
This subtle move may not be dazzling, but
it is strong and absolutely typical for Karpov.
Black's ideas include a possible re-routing
of the bishop from d7 to b7 via cS, as well
as potentially using the rook to defend the
kingside along the seventh rank.
19.:B:0
Wahls is not hiding anything, and shows a
clear desire ro attack on the kingside.
19 .!lJa5
•••
Karpov does not panic, but merely continues
to improve his pieces.
20.g4!?
Wahls continues to play aggressively. His last
move might help to prepare f5 , while in certain
positions he may prefer to advance with g5 in
order to fix the h7 -pawn.
White cannot make use of the fact that the
black knight no longer attacks the d4-pawn
with 20.c4?, on account of20 . . . tLlxc4! 2 1 .tLlxc4
(2 1 .ib4 :B:eS) 2 1 . . . ib5 22.id3 ixc4 23.ixc4
Wxd4t and Black wins material.
210
The Prime Years
23.g5
23 .Wh4 h6 defends.
23 . . . tt:l xd2 24.Wxd2 Wxa3 2 5 .:gfl
White is not helped by 2 5 .:gc l :gfc8.
2 5 ... :gfc8 26.:gff3 a5
Black has good chances to convert his extra
pawn.
8
7
a
b
c
d
f
e
g
h
20 ... :ae8!?
By putting his rook on the same file as
the enemy queen, Karpov discourages his
opponent from playing f5 . Nevertheless there
was an even stronger continuation:
20 . . . tt:l b3!
Hanging pawns are like isolated pawns;
when playing against them, it is useful to
exchange minor pieces in order to attack
them more effectively while also reducing
the opponent's attacking chances.
2 1 .:gh3
2 1 .tt:lxd7 Wxd7 22.Wg2 :gfc8 23.f5 e5 24.d5
tt:lc5 Black has strong play in the centre.
2 1 ..ie3 :gxc3 22.Wb2 (or 22 . .if2 :gxf3
23 . .ixf3 Wc7 24.We3 tt:l a5 2 5 .:gc l Wd6 and
White does not have much for the pawn)
22 . . . Wc7 23 . .if2 :gxf3 Black has safely won
a pawn.
2 1 . . . .ia4 22.Wf2
22 . .ig2 should also be met by 22 . . . Wd6.
22 . . . Wd6
6
5
4
3
2
1
a
b
c
d
e
f
g
h
21.i.bl ?
This unfortunate move is not only slow,
but also leads to unfortunate consequences
in a few moves' time. In this complicated
position White had several other ideas
available.
2 1 .g5 White plans to attack the h7-pawn,
but in the meantime Black can create
counter-threats. 2 1 . . . .ic6!? (This was Karpov's
suggestion, although 2 1 . . . tt:l b3 and 2 1 . . .We7
also deserve attention. 22.tt:lxc6 tt:l xc6 Black is
not worse.
2 1 .:gh3 tt:lb3!?
This sharp continuation is perfectly valid,
although Black has several other possibilities
including 2 1 . . .We7, 2 1 . . . .ic8, 2 1 . . . .ia4 and
2 1 . . .f5 ! ? 22.gxf5 exf5 23 . .id5 t .ie6 24.tt:ld7!
:gxd7 25 . .ixe6t <j;>f8 with a roughly equal
position.
22 . .ie3 :gxc3! ? 23 .Wb2
a
b
c
d
e
f
g
h
1992
Matthias Wahls - Anatoly Karpov
8
2 1 . f5 ! ? exfS
This was surely Karpov's intention, although
Black can also consider 2 1 . . .�xe 5 ! ? 22.dxe5
lLl c4 23 .�f4 We7 when the position is
complicated.
22.gxf5
7
6
5
4
3
2
21 1
V=U""C=
a
b
c
d
e
f
g
h
23 . . .'!Wc7!
Giving up the exchange would not offer
Black enough compensation.
24.tLlxd7 tLlxd4 2 5 . lLl e5 �xe5 26.fxe5 Wxe5
Black's three pawns give him enough
compensation for the sacrificed piece.
2 1 .�f2 The queen sensibly vacates the e-file.
Now Black has several roughly equivalent ways
to start exchanging pieces. 2 1 . . .lLl c4 (Also fine
is 2 l . . .lLl b3 22.Ei:h3 �a4 23.g5 [23 .�c2 b5]
23 . . . tLlxd2 24.Wxd2 We7 2 5 .Ei:al Ei:ecS and
Black has nothing to worry about.)
a
b
c
d
e
f
g
h
22.tLlxc4 (22 .�c l f5 23.gxf5 lLl xe 5 24. fxe5
gxf5 25 .�b 1 �c6 26.Ei:g3 Wd5 27.cj;>fl Black
has a repetition at his disposal with 27 . . . �b5t.)
22. . . Ei:xc4 23.f5 exf5 24.gxf5 Ei:xe4 2 5 . Ei:xe4
ic6 26.Ei:e2 �xf3 27.Wxf3 Black might be able
to claim a marginal edge, but nothing more.
Finally, the most aggressive and interesting
move at White's disposal was:
a
b
c
d
e
f
g
h
22 . . . �xe 5 ! ?
Black gives u p his important bishop t o take
advantage of White's dubious coordination.
22 . . . �cS 23 .Wg2 �xe5 24.dxe5 Ei:xe 5
2 5 .�f4 Black will not get quite enough
compensation for the exchange.
There was a safe alternative in 22 . . . �xf5
23.�xf5 gxf5 24. lLlxf7! Ei:xe2 2 5 . lLl xdS when
the endgame is equal.
23.dxe5 lLl c4 24.�f4
White must avoid 24.e6? �xe6!'
24 . . . lLl xe 5
It seems like White i s i n trouble, but h e can
keep himself in the game with the help of
some accurate moves:
2 5 .Ei:e3 lLl c4
2 5 . . . Ei:c5 26.fxg6 hxg6 27.�xg6 lLlxg6
2S.Ei:xeSt �xeS 29.WxeSt WxeS 30.Ei:xeSt
cj;>g7 Black is somewhat better, but White
has decent drawing chances in view of the
limited material on the board.
26.fxg6
26.Ei:g3? �xf5 is no good for White.
26 . . . hxg6 27.�xg6!
27.�xc7? Wxc7 wins.
27 . . . Ei:xe3
Th e Prime Years
212
21 . .i.c8!
.
Karpov gets ready to reposition his bishop
on the diagonal which its opposite number
j ust vacated. White's king will soon become
uncomfortable, as he can no longer block the
diagonal with his pawns.
a
b
c
d
e
f
g
h
2B .i.xf7t!!
This lovely tactical shot keeps White in the
game.
2B . . . \t>g7!?
2B ... <;t>xf7 leads to a draw: 29.�h 5 t \t>e6
30.Ei:xe3t tLlxe3 3 1 .�e5t \t>f7 32.�h5 t
<;t>e6 33.�e5t
29 ..be3!
The showy 29 .i.h6t?! does not have the
desired effect: 29 . . . <;t>xf7 30.�h5 t <;t>e6!
3 1 .�g4t <;t>d5! 32.�d4t <;t>c6 33.�xc4t
\t>b7 34.�d5 t i.c6 3 5 .�xdB Ei:xe 1 t 36.<;t>f2
Ei:e4 Black keeps some winning chances, as
White's king is exposed.
29 . . . tLlxe3 30.�xe3
Both sides have open kings, but neither is
able to get the upper hand.
30 . . . Ei:c5
After 30 . . . \t>xf7 3 1 .:8:n t <;t>gB 32.�g3t i.g4
33.h3 the game will end in a draw.
3 1 .i.h5 Ei:xh5 32.� e7t
The endgame should be drawn.
The plan of exchanging pieces with 2 1 . . .tLlb3
was also promising: 22.Ei:h3 (22.i.e3 ? Ei:xc3)
22 . . . �e7 (22 . . . tLl xd2 23 .�xd2 �e7 24.f5
exf5 2 5 . gxf5 i.cB 26.Ei:g3 reaches a balanced
position) 23 .�d l (23 .f5 ? exf5 24.gxf5 i.xe5
2 5 .dxe5 �c5 t-+) 23 . . . tLl xd2 (23 . . . �xa3 24.c4)
24.�xd2 Here Black can keep some advantage
with either 24 . . . �xa3 or 24 . . . i.cB .
22.g5
White fixes the h7-pawn, but he is unlikely
to do any damage there. His position would
have remained difficult after other moves as
well.
22.f5 exf5 23.gxf5 tLl c4 (23 ... i.b7!?) 24.i.f4
i.b7 2 5 .i.e4 i.xe4 26.�xe4 g5!? (There is also
26 . . . Ei:ce7 27.Ei:d3 f6 when White is awkwardly
tied up.) 27.i.xg5 f6 2B.Ei:ffl (2B .i.h4 tLl d2-+)
2B ... fxg5 29.f6 i.xf6 30.Ei:xf6 �xf6 3 1 .tLlg4
White is clearly worse although he retains
some chances to hold.
It was worth considering 22.i.a2!?, taking the
d5-square away from the black queen while
also preventing the knight from hopping to
b3 to force exchanges. Nevertheless Black
keeps the advantage here too: 22 . . . i.b7
23.Ei:e3 i.d5 24.i.xd5 �xd5 2 5 . f5 If White
does nothing then his hanging pawns will
soon come under fire. 2 5 . . . tLl b3 26.c4
�xd4 27.i.c3 �d6 2B.tLlxf7 Ei:xf7 29.fxg6
hxg6 30.i.xg7 <;t>xg7 3 1 .Ei:xb3 Ei:f4 White's
position is rather unpleasant as his kingside is
somewhat open.
22 . .i.b7 23J�g3 Y;Vd5
.
a
b
c
d
e
f
g
h
1992
213
Matthias Wahls - Anatoly Karpov
The queen takes up a dominating post in
the centre. Black's pieces are all working in
harmony, and White's king is under heavy
pressure on the diagonal.
2 8 .Elge3 Wa3 is also winning for Black.
28 . . . Wa3 29.Elee3 f6!
Black attacks from a new angle.
30.gxf6 �xf6
White's position is about to collapse.
25 ... tLlc4 26.�a2
The endgame resulting from 26.t2Jxc4 Wxc4
27.Wxc4 Elxc4 is virtually hopeless for White,
for instance 2 8 .Elge3 Ela4 29.�c l �f8 30 .�c2
Ela5 3 1 .a4 �d6 32. <;!;>g3 �c6 and Black wins
a pawn.
a
b
c
d
e
f
g
h
24... :aec8
Karpov no longer needs to pay attention
to f4-f5 , so he uses the rook to increase the
pressure on the backward c-pawn.
25.h4
Wahls must have known he had no chance
of hurting Karpov on the h-file, but he
probably went for it anyway as there were not
many other moves available.
25 .a4!?
It is hard to say, but perhaps this would have
been a bit more stubborn.
25 .. .tL'lc4
After 2 5 . . .'!Wb3 26.Wd l Wa3 27.�c2 t2Jc4
28.t2Jxc4 Elxc4 29.�b3 El4c7 White is
struggling but it seems he can avoid a direct
loss for the time being.
26.�c2
If 26. t2Jxc4 Wxc4 27.Wd l �c6 28 .�c2 �d5
White can hardly move.
26 . . . t2Jxd2 27.Wxd2 Wa2 2 8 .Elh3
28 .Elb I ? �e4-+
26.h5 is even worse: 26 . . . t2Jxd2 (There is
also 26 . . . t2Jxa3 27.�d3 b5 when Black is
a pawn up for nothing.) 27.Wxd2 �xe5!
This wins by force. 28.hxg6 (28 . fxe5 Elxc3)
28 . . . �xd4t 29.cxd4 hxg6 White is completely
lost.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
26 ...i.a6
Karpov plays pragmatically and maintains
control, but he could have won more or less
by force: 26 . . . �xe5! 27.fxe5 Wa5 28 .�xc4
(28 .Elcl Wxa3) 28 . . . Elxc4 29.Elal Ela4 Black
wins a pawn, and White has no real attacking
chances.
27.hc4?
Exchanging accelerates the end as it becomes
easier for Black to invade on the light squares.
Th e Prime Years
214
27.h5! would have given Black a chance to go
wrong:
a) 27 .. .tt'lxe 5 ? This simplification wastes a
great part of Black's advantage. 2 8 .�xd5 �xe2
29.'it>xe2!! White king gets close to the pawns,
with its help White suddenly has realistic
chances to hold. 29 . . . tLl c4 30.�xc4 1'!xc4
3 1 .hxg6 hxg6 32. 'it>d3 1'!a4 33.�c l White is
worse, but he can continue to resist for quite
some time. It is hard for Black to bring his
bishop to bear on the weak queenside pawns
in view of the open h-file.
b) 27 . . . Wa5 ! 28 .�xc4 �xc4 29.tLlxc4 1'!xc4
30.1'!e3 Wd5! (30 . . . Wxa3 3 1 .hxg6) 3 1 .Wf3
White avoids collapsing in the short term, but
Black remains in full control and is likely to
find a way through eventually.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
30 ... gxhS!
Karpov wins a pawn and reduces White's
chances of achieving something on the h-file.
27 ....L:c4 28.We3
8
3 1 .�h3 .tb7!
7
This bishop has been the star piece of the
game. With its latest seemingly modest retreat,
it cleverly nullifies the threat of 32.1'!xh5,
which can now by met by 32 ... 1'!xc3! .
6
5
4
32.We2 WdS
3
The queen has done her duty on the
queens ide, so now she returns to the centre
and resumes the mating threats.
2
1
30.hxg6 should be met by 30 . . . fxg6! 3 1 .'it>gl
(3 1 .1'!d1 1'!xc3) 3 1 . . .�xe5 32.fxe5 1'!0 White
is under pressure on both flanks, and has little
chance to survive.
a
b
c
d
e
f
g
h
28 . i.a6!
..
Karpov clears the c-file and opens the way for
his queen to invade the opponent's territory.
29.hS Wal!
Threatening . . . 1'!xc3 .
30.i>gl
If 30.1'!cl �b7 3 1 .h6 �f8 32.We 1 �d6 and
Black will invade one way or another.
32 . . . f5 ! was also strong, although it would have
led to a position with some practical chances
for Black to go wrong: 33.1'!h2 (33.gxf6 �xf6
34. i> h2 1'!g7-+) 33 . . . �e4! (33 . . . h4?! 34.Wh5
�d5 3 5 .Wxh4 �xe5 36.fxe5 Wxa3 37.g6! We7
[37 . . . 1'!g7 3 8 . i> f2! White is alive] 3 8 .�g5 Wg7
39.mf2 Wxg6 40.�d2 Black is two pawns up
but White has a dangerous initiative.) 34.Wxh5
�xe5 3 5 .fxe5 1'!g7 Black succeeds in defending
his kingside, meanwhile White has serious
problems on the light squares.
1992
21S
Matthias Wahls - Anatoly Karpov
33.�h2
Wahls hopes to get rid of the mating threat
on the g2-square.
If instead 33.Ei:xhS Ei:xc3! 34.�xc3 Ei:xc3
35 .�f2 �xeS 36.dxeS Ei:f3 the end is near for
White.
a
a
b
c
d
e
f
g
h
33 ... b5!
By threatening . . . as and . . . b4, Karpov
increases the pressure even more.
34.Ei:gl?
The German grandmaster commits a
serious mistake in time trouble. Perhaps he
was worried about . . . f6 earlier and therefore
decided to prevent it.
The best chance was:
34.Ei:xhS!
White should take the opportunity to
restore material equality. His position
remains difficult, but Black still faces a
challenge regarding how best to convert his
advantage.
34 . . . �xeS!
The opposite-coloured bishop position is the
most unpleasant one for White to face.
34 . . . a5? would be a mistake as after 3 S . lLl g4!
White suddenly becomes active, and 3 S . . . WTf3
can be adequately met by 36.Ei:h3!.
35 .fxe5
b
c
d
e
f
g
h
3S . . . aS!
Black gets ready to sacrifice a pawn in order
to open the way for his rooks to penetrate
to c2.
The exchange sacrifice 3S . . . Ei:xc3?! is playable,
but not decisive: 36.�xc3 Ei:xc3 37.Ei:h3!
WTxd4 (37 . . . Ei:c4 3 8 .Ei:d3) 38 . Ei:xc3 WTxc3
39.WTe3 White should be able to keep his
position together.
36.Ei:h4
36.Ei:h3 b4 37.cxb4 axb4 3 8 .WTf2 b3! 39.Ei:b l
b2 40.Ei:xb2 WTh l t 4 1 .<;t>g3 Ei:c3t Black
WIns.
36.WTf2 b4 (36 . . . WTa2 37.Ei:e3) 37.cxb4 axb4
38.axb4 Ei:a8! 39.Ei:fl Ei:a2 White is under
pressure.
36 . . . b4!
This seems best, although 36 . . . Ei:xc3!? is
more tempting here than on the previous
move, as Black's queenside pawns are more
mobile: 37.�xc3 Ei:xc3 3 8 .Ei:h3 Ei:c4 39.Ei:d3
Ei:xd4 40.Ei:xd4 WTxd4 4 1 .<j;lg3 b4 White faces
an unpleasant defence.
37.axb4 axb4 3 8 . cxb4 Ei:c2 39.g6
39.WTf2 Ei:a8 wins.
39 . . . fxg6 40.Ei:f4 Ei:a8!
Technically speaking this is the most accurate
move.
40 . . . WTd8 is also good: 4 1 .bS (4 1 .Ei:efl WTgS)
4 1 . . .Ei:b2 42.WTe3 Ei:cc2 43.Ei:d 1 It is possible
that White can hold this position with perfect
defence, but in an over-the-board game the
odds would be in favour of a black win.
216
The Prime Years
35.Ei:c1?!
4 1 .Ei:efl
In a difficult situation and with no time on
the clock, Wahls blunders.
3 5 .Ei:e l ! was the best chance, when Black still
has to choose the right path. There are two
ideas:
a
b
c
d
e
f
g
h
4 1 . . . g5!!
The surprising pawn move forces the rook to
give up either the f-file or the fourth rank.
After 4 1 . . .ia6 42.Ei:f8t Ei:xfS 43.Ei:xfSt <;t>xfS
44.ih6t <;t>gS 4 5 .'lWxc2 'lWxd4 46.'lWc6 Black
is certainly better, but White has decent
drawing chances.
42.Ei:4f2
42.Ei:g4 h6-+
42 . . . 'lWxd4 43 .'lWe3 'lWh4t 44.'lWh3 'lWxh3t
4 5 . <;t>xh3 h6
Black is winning.
a
b
c
d
e
f
g
a) 3 5 . . . ixe5 is possible, but does not work
quite so well here: 36.fxe5 (36.dxe5 Ei:d7)
36 . . . ia6 (36 . . . 'lWc2 37.Ei:xh5 Ei:xc3 3s .ixc3
'lWxe2t 39.Ei:xe2 if3 40.Ei:xh7! White is worse
but continues to resist) 37.'lWg2! (37.'lWf2 Ei:xc3)
37 . . .'lWxa3 3S .'lWe4 (3S .Ei:xh 5 ? ib7 39 .'lWf2
'lWb2 [39 . . . Ei:xc3? 40.Ei:xh7!] 40.Ei:e2 a5-+)
3S . . . b4 (3S . . . 'lWb2 39.Ei:xh5 'lWxd2t 40J�le2
'IW d 1 4 l .'lWxhlt <;t>fS 42.Ei:f2 White is still alive)
39.Ei:xh5 <;t>f8 40.'lWxh7 'lWa2 4 l .'lWhSt <;t>e7
42.'lWf6t <;t>d7 Black will walk to the queenside
with his king and maintain the better chances,
but his position is not winning.
b) The right way is: 35 . . . a5! 36.Ei:xh5 b4!
(36 . . . ixe5?! 37.fxe5 if3 38 .'lWxf3 'lWxd2t
39.Ei:e2 Ei:xc3 40.'lWxf7t! White is still breathing)
37.axb4 (37.tZJg4 bxc3-+) 37 . . . axb4 3S.cxb4
Ei:c2 Black is winning.
h
34 J1*Ia2!
a
•.
Karpov prevents White's intended 'lWxh 5 ,
and also prepares t o exploit the fact that the
white rook gave up its control over the e4square.
b
c
d
e
f
g
h
35 .ie4!
.•.
This final flourish from the brilliant bishop
seals White's fate.
1992
Matthias Wahls - Anatoly Karpov
217
36J�e3 1.f5
Black's kingside is now completely secure, while his position on the opposite flank is
overwhelming.
37.iel W1xa3
0-1
Wahls lost on time but his position by this stage is absolutely hopeless.
This final victory gave Karpov a superb final total of 9 Yz/ 1 1 , two points ahead of Lutz who
finished in second. Interestingly Karpov drew with the three players who finished in 2nd-4th and
defeated the remaining eight. Like the Madrid tournament earlier in the year, the Baden-Baden
event was slightly below Karpov's usual level of competition, nevertheless he won them both with
remarkable authority. Compared with the previous year, Karpov slightly reduced his number of
losses from eleven to ten, and increased his number of wins from twenty five to thirty one.
218
1992 Summary
Linares (4th place) : 7Y2/ 1 3 (+5 = 5 -3)
Candidates semi-final match versus Short: Lost 4-6 (+2 =4 -4)
Madrid ( l st place) : 7Y2/9 (+6 =3 -0)
Biel ( l st place) : 1 OY2/ 1 4 (+8 = 5 - 1 )
Tilburg Knockout, Round 1 versus Chernin: 1 12 ( 0 =2 -0) (lost rapid tie-break)
Alekhine Memorial, Moscow (4th-6th place) : 3Y2/7 (+2 =3 -2)
Baden-Baden ( l st place) : 9Y21 1 1 (+8 =3 -0)
Total 65.9% (+3 1 =25 - 1 0)
II Wins • Draws • Losses
1993
Rating 2725 (2 in the world)
Karpov's first event of the new year was Wijk aan Zee, which followed the knockout format at
that time, with two-game matches followed by a rapid tie-breaker if needed. Karpov began in the
worst possible way, blundering a piece after j ust twelve moves against Larry Christiansen. (You
can see what happened in the notes to Game 3 1 later in the chapter.)
I was in Wijk aan Zee on that day to visit Peter Leko, who was my pupil at the time. During
the evening I saw Karpov having supper with Lembit 011. Later I asked the Estonian grandmaster
whether Karpov had said anything about the game. He told me that Karpov did not mention it
at all.
The one good thing for Karpov was that his blunder occurred when he had the black pieces, so
he still had a fair chance to win with White the next day to tie the mini-match. He managed to
do it, and then eventually overcame Christiansen in a series of hard-fought rapid games.
In the second round Karpov faced John Nunn, who employed a Modern Benoni set-up via a
Samisch King's Indian move order. For a while the position was sharp and dynamically balanced,
but shortly before the time control the English grandmaster blundered a piece and resigned soon
after. In the second game Karpov surprised his opponent with a Fort Knox French. For a while his
position looked uncomfortable, but he gradually managed to ease the pressure and make a draw.
Karpov's next opponent was Valery Salov. The opening phase of their first game was very unusual
indeed - we will look at it briefly.
l.d4 d5 2.c4 c6 3 . lLl c3 dxc4
Karpov never before and never again encountered this move.
4.e4 e5 5 . dxe5
This move is rare, and may have been the product of Karpov's over-the-board improvisation.
5.lLlf3 is more common, and was once used by Portisch who defeated Saidy in an endgame
masterpiece.
5.. .�xd l t 6.<;t>xd l
The mutual double-pawn majorities promise a rich and sharp queenless middlegame. Despite
Karpov's success in the present game, this line did not attract many followers.
6 . . . h5?!
Later Stefanova improved with 6 ... �e6, which looks more solid.
7.Ad lLld7 8 . f4
Karpov bases his play on his strong central pawns.
8. ..Ac5 9.<;t>e2 a5
9 ... g5 should be met by 1 O.g3 .
220
The Prime Years
1 O. tLl f3 tLl e7
1 6. tLl d l h5?!
1 6 ... h6 was better.
8
8
7
7
6
5
6
3
4
5
4
3
2
a
b
c
d
2
e
f
g
h
1 l . f5 !
Karpov uses his pawns to choke Black's
position.
1 1 . . ..ib6 12 . .ixb6 tLlxb6
a
b
c
d
e
f
g
h
1 3 . 'it>e3!
Karpov puts his king to remarkable use. It is
strong in the centre, and may help to defend
the e5-pawn.
1 3 . . . f6
Perhaps Black should have focused on the
queenside with 1 3 . . . .ia6.
1 4 .exf6 gxf6 1 5 .g4 b4
Around here Black might have considered
1 5 . . . h6!? with the idea to exchange as many
kingside pawns as possible, and maybe even
sacrificing a piece for White's two remaining
kingside pawn, followed by swapping off
White's last few queenside pawns to make a
draw.
a
b
c
d
e
f
g
h
1 7.E!:g l ?
It is hard to guess why Karpov rejected the
natural and strong 1 7.g5! (even 1 7.lD£2
is stronger than the game continuation) .
Perhaps he was concerned abour 1 7 . . . fxg5
1 8 .tLlxg5 .ixf5 ! ? , but after 1 9.exf5 tLlxf5t
20.'it>f2 0-0 2 1 .E!:gl White has excellent
winning chances.
1 7 . . . hxg4 1 8 .E!:xg4
The position is unclear, Karpov - Salov, Wijk
aan Zee 1 993.
Later in the game Salov sacrificed a pawn in
order to obtain a protected passed pawn and
decent counterchances, but then he gave up
his strong pawn unnecessarily and Karpov
punished him. In the return game Karpov was
under pressure in a semi-closed double rook
endgame, but he managed to draw it and thus
secure his ticket to the final.
In the final Karpov faced Illescas in a four­
game match. In the first game he equalized
on the black side of an English Opening, and
the game remained balanced until j ust after
the time control when a draw was agreed.
In the second game Karpov got an edge in a
'iff1c2 Semi-Slav (see the notes to Game 27) ,
and subsequently won material. He made a
few shaky moves before the time control, but
nevertheless was able to win eventually. In the
22 1
1 993
third game Karpov again drew solidly against
the English.
The final game was a dramatic one. Requiring
only a draw with the white pieces, Karpov
played too timidly against the Semi-Slav and
the Spaniard skilfully outplayed him. At one
point he was a pawn up with a clearly winning
position, but allowed Karpov to escape to a
still difficult endgame. Illescas was able to win
a second pawn but Karpov narrowly managed
to hold the draw and thus secure the winner's
trophy.
The next big tournament was Linares, where
Karpov started with a beautiful attacking win
over Salov, which can be found on page 87, in
the note to White's 1 3th move in Game 1 3 . In
Round 2 Karpov repeated the defensive set-up
he used against Illescas in the English Opening,
but this time he failed to equalize, and later
forfeited on time in a losing position.
Karpov bounced back with a win over
Beliavsky, referenced on page 400 of the first
volume. In Round 4 he held Anand using a
Fort Knox French, then drew with White
against Ivanchuk. In Round 6 Karpov had the
white pieces for the second consecutive time,
and was able to defeat Gelfand with the help
of a beautiful positional exchange sacrifice.
He followed it with a solid draw against
Shirov.
In Round 8 Karpov faced Kamsky, who
played a most imaginative piece sacrifice on the
black side of a Samisch King's Indian. Karpov
was under some pressure, but survived and
eventually won the game after Kamsky erred.
In the next game he outplayed Ljubojevic on
the black side of an English Opening.
In Round 1 0 Karpov's aspirations of winning
the tournament took a hammer-blow as he
lost in spectacular style against Kasparov, who
played brilliantly. Karpov chose the Samisch
Variation against the King's Indian and made
an overoptimistic pawn grab. He then tried to
drive his opponent's knight from the centre
with the move f3-f4, to which Kasparov replied
with the powerful zwischenzug . . . b5-b4! ,
leaving his own knight temporarily e n prise
and targeting Karpov's c3-knight. According
to Anand, it was rumoured that Karpov had
overlooked this key resource. When this was
put to Kasparov, the then World Champion
retorted in his own inimitable style, "Did he
expect me to resign?"!
From there on, Kasparov's initiative
snowballed and by move 22 all seven of
Karpov's remaining pieces had been driven
back to the first rank. The game contained one
more twist: a couple of moves later Kasparov
used his c-pawn to capture a white rook on
d 1 , but he did not have a spare queen to hand.
The arbiter mistakenly brought a white queen,
after which Kasparov lost patience and pressed
his clock with the pawn still on d l , expecting it
to be captured on the next move. Upon seeing
this, Karpov cheekily left the pawn where it
stood and captured a different piece, claiming
that Kasparov could have promoted to a knight
or a bishop! Unfortunately for him this gambit
was declined, and three moves later he lost on
time in a hopeless position.
Karpov managed to put this disappointing
result behind him and won a nice game from
the black side of a Samisch Nimzo-Indian
against Yusupov. In the final two rounds he
found himself on the brink of defeat against
Kramnik and clearly worse against Bareev,
but thanks to some inaccuracies from his
opponents, combined with his own fine
defensive technique, he was able to draw both
games. His final score of 8 Y2/ 1 3 was enough to
tie for second place with Anand, a point and a
half behind Kasparov.
***
After the break-up of the Soviet Union,
Karpov began to play in some team events in
222
The Prime Years
different European countries. According to
the database Karpov played two games in the
Austrian Team Championship, drawing with
Blatny and beating Duer in a nice game which
is referenced in the notes to Game 32 later in
this chapter. In the European Cup final Karpov
drew against Mikhail Gurevich and defeated
Ekstroem.
Karpov was especially active in the French
Team Championship. His results included a
black draw against Pinter, a white draw against
Lautier, and three white wins against Belotti,
Chevallier and Kallai. We will examine the last
on the list. The Hungarian grandmaster played
three games against the world champions in the other two he drew with Spassky and
Kramnik.
I Game 30 I
1 993
l .e4 c5
Kallai also plays the 4 . . . ltJ d7 variation of the
Caro-Kann. I worked on that line a lot with
Gabor, before the Karpov - A. Sokolov match
which brought it into fashion. In a way it was
bad luck for us, as from then on many other
players started investigating that variation.
It is understandable that Gabor did not want
to find out how Karpov, the most prominent
expert on the 4 . . . ltJ d7 line, was planning to
meet his pet variation.
2.lLlf3 lLlc6 3 ..ib5
Karpov played the Rossolimo three times
back in 1 969 in Soviet team events, and
never repeated it after this game. Was it is
coincidence, or was there some reason why he
felt inclined to play it at team events?
3 ... g6
4.0-0 .ig7 5 .ixc6
•
The first deep investigation of the capture
on c6 was carried out by the great junior
trainer and opening theoretician Chebanenko.
Interestingly Fischer and Kasparov took
the knight one move earlier, but perhaps on
principle Karpov preferred not to follow their
play exactly.
5 ...bxc6
5 . . . dxc6 has
nowadays.
a
more
solid
reputation
6.d3
Anatoly Karpov - Gabor KaUai
French Team Championship
Kallai had a game against Benjamin where
he undertook doubled isolated pawns on the
c-file, so Karpov may have been hoping for the
same.
6.c3 is more popular than the text move,
but 6.l"i:e l ! is the main line, maintaining the
flexibility of White's pawns and in particular
keeping the option of playing c3 and d4
without losing a tempo.
Karpov's move is hardly critical, although
it should not be forgotten that he enjoyed
great success on the black side of the English
Opening with l .c4 e 5 , so he must have felt
happy in the reversed position with an extra
tempo.
6 ... f6?!
Black intends to put his knight on h6. There
is nothing wrong with this plan, although it
would have been more flexible to play 6 . . . tDh6
immediately, as in some lines Black may be
able to do without . . . f6.
The main moves are 6 . . . d6 and 6 . . . e 5 .
7.c3
Preparing d4 is the most promising plan in
the position, even though it involves a tempo
loss.
7 ... lLlh6 8.d4 cxd4 9.cxd4 lLlf7 1O.lLlc3 e5?!
1 993
223
Anatoly Karpov - Gabor Kallai
Black establishes some space in the centre,
but his pawn structure becomes a bit rigid.
i.f6 1 7.i.h6 White is a bit better, but Black
should be able to live with it.
1O . . . d6 leads to a standard type of position, in
which White has lost some time by playing d4
in rwo moves rather than one. Normally his
rook would be on e1 here; see for example the
thirteenth game of the 1 992 Fischer - Spassky
match, which was eventually drawn.
l l .dxeS
Karpov decides to fix Black's pawn centre
without delay.
11. .. fxeS 12.i.e3 0-0 13J!Bdl
a
b
c
d
f
g
h
Karpov gets ready to exert pressure on the
d-file. 1 3 .�a4!? was also interesting.
14.h4!?
13 ... h6
Karpov hopes to soften Black's pawns in the
centre and gain control over the d5-square.
Kallai probably wanted to prevent Karpov
from exchanging the f7 -knight, and in the
longer term he may have wanted to push his
kingside pawns. It is worth checking some
other plans, as Black had a few different ways
to handle his central pawns.
1 3 . . . ib7?! intends a pawn sacrifice, but it fails
to solve Black's problems: 1 4.l"i:ad 1 ! Thanks to
this accurate move Black will not be able to
take the b2-pawn with a tempo. (After 1 4.l"i:fd 1
d5 1 5 .exd5 cxd5 1 6.lLlxd5 e4 1 7. lLl e 1 �xd5
18.'&xd5 ixd5 1 9.1"i:xd5 i.xb2 20.l"i:ad l White
is a bit better, but Black has good chances to
survive.) 14 . . . d5 1 5 .exd5 cxd5 1 6.lLlxd5 e4
1 7.lI:l g5 Black has insufficient compensation
for the pawn.
Black could have considered the modest
13 . . . d6!? intending to develop normally.
14.b4!? ( l 4.h3 ie6 1 5 .l"i:fd l �e7100ks okay for
Black.) 1 4 . . . i.e6 (After 1 4 . . . i.g4 1 5 .lLlg5 lLl xg5
16.ixg5 White will soon start playing against
Black's central pawns.) 1 5 . lLlg5 ( l 5 .l"i:fd 1 and
15 .b5 can also be tried.) 1 5 . . . lLl xg5 1 6.i.xg5
e
14...i.h7 lSJ:Ud l
This seems natural, although 1 5 .l"i:ad 1
was also possible, and would have had the
advantage of removing the rook from the long
diagonal: 1 5 . . . d5 ( l 5 . . . a5 1 6. lLla4! [ 1 6. b5
cxb5 i s nothing special] 1 6 . . . d 5 1 7.ib6 �f6
1 s .lLlc5 White's pieces stand well.) 1 6.i.c5
l"i:eS 1 7.exd5 cxd5 1 s .lLlxd5 e4
a
b
c
d
e
f
g
h
1 9.1Lle7t! ( l 9 . lLl e l lLl e 5 20.lLle7t cj;Jh7 Black
has enough activity.) 1 9 . . . l"i:xe7 20.�xdSt
l"i:xdS 2 1 .l"i:xdSt lLlxdS 22.i.xe7 lLl c6 23 .i.c5
exf3 24.l"i:d 1 We have reached a rich endgame
224
The Prime Years
in which White's chances should be somewhat
higher.
If 1 9 .:B:ab l axb4 20.�xb4 �a6 Black's light­
squared bishop becomes active.
1 9 . . . axb4
15 ... d5
Black decides to play actively and advances
in the centre.
The attempt to exchange pawns on the
queens ide loses a pawn: 1 5 . . . a5? 1 6.bxa5 �xa5
1 7.�xd7 White has a serious advantage. This
line reveals one advantage of bringing the
f-rook to dl as . . . �a6 will not win a tempo.
The modest 1 5 . . . d6 does not fit well with the
previously played . . . h6. 1 6.�d3! The queen
defends e4 and prepares b5 and might attack
g6 in some lines. 1 6 . . . a6 1 7.a4 White is doing
well. His plans may include b 5 , or the knight
transfer lLl d2-b3-a5.
16 ..ic5 �e8 17JWc2
Karpov
advancing.
provokes
the
d-pawn
into
17 ... d4
And Kallai is happy to accept the invitation.
a
b
c
d
e
f
g
h
20.:B:ab l ! ?
20.�xb4 gives Black the opportunity to play
a promising pawn sacrifice: 20 . . . c5! (20 . .Wf6
2 1 .lLl c l �a6 [it is not to late to consider
2 1 . . .c 5 !? 22.�xc5 :B:ec8!?] 22.lLld3 �xd3
23 .�xd3 �f8 Black is close to equalizing, as
he can double his rooks on the a-file which
should restrain the a-pawn.) 2 1 .�xc5 Wd7!
Black threatens . . . :B:ec8 and . . . �c6. He has
excellent compensation, and it is White who
has to be more careful.
20 . . . �a5
Black can also look towards the kingside
with 20 . . . �f6.
2 1 .:B:xb4 �a6 22. lLl c l :B:eb8 23.lLlb3 �c7
Black is holding his own in this complex
positional battle.
.
The provocative 17 . . . a5 ! ? was also playable:
1 8 .b5 (White can win a pawn, but in doing
so he opens the position for the black pieces:
1 8 .exd5 cxd5 1 9.�xg6 axb4 20.�xb4 Now
both 20 . . . e4 and 20 . . . d4 offer Black reasonable
compensation.) 1 8 . . . d4 1 9 .b6 �a6 20.lLla4 :B:b8
2 1 .:B:ab 1 The position is rather unorthodox,
but White's strong passed pawn and queenside
squares should give him the upper hand.
Karpov discourages his opponent from
playing . . . a5.
18.tLle2 �h7
19 ....ic8
1 8 . . . a5!?
Once again this was an interesting
possibility.
1 9 .a4!?
White hopes that the a-pawn will become a
powerful force.
19.�ab l
Kallai has decided to stake his hopes on a
kingside attack.
It was not too late to fight back on the queen­
side with 1 9 . . . lLl d6!?, for instance: 20.ttJc1
(20.a4 �a6) 20 ... lLlb5 Heading for the c3-
1993
225
Anatoly Karpov - Gabor Kallai
square. 2 1 .E1b3 VfJc7 It is not easy for White to
proceed, as both 22.a4 tLl c3 and 22.tLld3 .ia6
give Black reasonable play.
20.tLld!
Karpov brings the knight towards its ideal
home on d3 .
20 ... h5
White could also have played for a knight
transfer to f5 : 23.tLle2!? g4 (23 . . . h4 24.h3
Black's kingside play comes to a dead end.)
24. tLl h4 White's position is preferable.
23 . . . g4 24. tLl d2 .ie6!?
24 ... .ih6 2 5 . tLl c4 h4 26.E1a5 .ie6 (Also
after 26 . . . g3 27.hxg3 hxg3 28 .E1b3 White's
chances are higher.) 27.E1b3! White can
utilize his rooks in a most instructive way.
27 . . . E1eb8 28.E1ba3 E1b7 29.E1a6! .ixc4
30.VfJxc4 tLl g5 3 1 .VfJc2 White's queenside
attack is more effective than Black's kingside
counterplay.
2 5 .E1a6 .ic8 26.E1a5 h4 27.E1b3 a6 28.tLlc4
E1b8
It is not easy for White to make progress on
the queenside.
23.lLld2
a
b
c
d
e
f
g
h
2 1 .E1d3!
Before putting his knight on this square,
Karpov uses it to activate his rook in a most
original way, while conveniently avoiding a
future pin with . . . .ig4. It has been suggested
that, out of all the other world champions, it
was Petrosian whose playing style most closely
resembled that of Karpov. Indeed, the ninth
World Champion made quite a few remarkable
rook manoeuvres over the course of his career.
Karpov anticipates the advance of Black's
g-pawn, and brings his knight to the
queenside.
23 ... g4 24.tlJc4 �b8
Black decides to keep the bishop, but this
may not have been necessary.
24 . . . h4! ?
This looks more challenging. Black accelerates
his attack and intends to weaken the f3square in order to put his knight there.
2 1 ...g5 22.E1a3! a6
22 . . . a5? was no longer a viable option, as
after 23.tLlb3 a4 24.tLla5 the pawn will soon
perish.
22 ... 'I1Nc7!?
Kallai probably wanted to keep his queen
free for attacking duties; nevertheless this
move was playable.
23.liJd3
a
2 5 . liJ b3 !
b
c
d
e
f
g
h
226
The Prime Years
White continues to improve his pieces.
Exchanging Black's light-squared bishop
does not take the sting out of his kingside
initiative: 2S .lbb6 Elb8 26.lLlxc8 Wxc8
27.We2 (27.f4 exf4 28.lLle2 d3! 29.Elxd3
lLleS is also promising for Black.) 27 . . . lLl gS
28.Elxa6 h3 29.Ela7 hxg2 Black has a
dangerous attack.
White can try to fight back on the kingside
with 2 S . f3!?, but this is also not fully
convincing: 2S . . . lLl gS (or 2S . . . WgS 26.lLlb6
Elb8 27.lLl xc8 Elbxc8 28 .Elbb3 Ela8 with
mutual chances) 26.fxg4 (After 26.lLld6
gxf3 27.lLlxe8 Wxe8 28.gxf3 Wf7 29.We2
�e6 Black has decent compensation for the
exchange.) 26 . . . �xg4 27.lLld3 �hS Black
has a full share of the chances.
2 S . . . lLl gS
2S . . . Elb8 26.We2 is also a bit better for
White.
a
b
c
d
e
f
g
h
26.lLl d6!
The knight heads for the outpost on fS .
26.lLl baS leads to complex play: 26 . . . Wf6
27.lLlb6 �b8 28.lLlxc8 (28 . lLl ac4 �e6)
28 . . . Elbxc8 29. lLl c4 Elcd8 30.We2 (30.lLld2
�h6) 30 . . . Wf4 3 1 .lLld2 Ela8 The position is
roughly balanced.
26 . . . Ele6 27. lLl fS h3 28 .We2! Elg6
28 . . . hxg2 29.Wxg4 �h6 30.h4 lLlf7 3 1 . lLl aS
Elg6 32.WhS White succeeds in gaining the
upper hand in the complications.
29.Wxg4 �h6 30.g3 Wf6 3 1 .Ele l
White is better, although making his extra
pawn count will not be easy.
8
7
6
5
4
3
2
1
a
b
c
d
f
e
g
h
25.ltJd3 Ele6
2S . . . h4
With the knight already on d3, White is in
a much better position to react to the direct
attack.
26.Elf1 !
It turns out that White can not j ust defend,
but even fight for the initiative on the
kingside.
26 . . . lLlgS
26 ... h3?! is too slow, and after 27.f4! gxf3
28.Elxf3 Black's pieces are not working in
harmony.
26 . . . WgS should be met by 27.f4! gxf3
28.lLl e l ! WhS 29.lLlxf3 when White keeps a
slight plus.
27. lLl d6 Ele6 28.lLlxc8 Wxc8 29.We2 Elg6
a
b
c
d
e
f
g
h
1993
30.h3! CLl e6
30 . . . gxh3? can be refuted by a few accurate
moves: 3 1 .'lWh5t 'kt>gB!? (3 1 . . . ih6 32.CLlxe5+-;
3 1 . . . Ei:h6 32.'lWxg5 hxg2 33.Ei:e l ! White should
be winning) 32.CLlxe5! (but not 32.'lWxg6??
lWg4-+) 32 ... Ei:e6 33.'lWxg5 Ei:xe5 34.'lWxh4
hxg2 35 .Ei:dl White is winning easily.
30 . . . CLl f7?! 3 1 .hxg4 Ei:xg4 White obtains a
clear advantage with 32.f4 (or first 32.f3 ! ?
Ei:g3 followed b y 33.f4)
3 1 .hxg4 CLlxc5 32.bxc5 Ei:xg4 33 .'lWf3
Both kings are somewhat vulnerable, but
White has the advantage as his knight is clearly
stronger than Black's bishop.
a
b
c
d
e
227
Anatoly Karpov - Gabor Kallai
f
g
26 . . . Ei:b7? was even worse, and after 27.Ei:xa6
Black loses a pawn for nothing.
26 . . . Ei:b5!
This was the correct option, after which the
position would have remained complex.
27.Ei:ab3
27.Ei:fl ! ? h4 2B.f4 gxf3 29.Ei:xf3 CLlg5 30.Ei:f5
Ei:eB reaches an unclear position with mutual
chances.
27.Ei:a5 ! ? leads to interesting play: 27 . . . CLl d6
(There is also 27 . . . h4 2B.a4 Ei:xa5 29.CLlxa5 [or
29.ib6 'lWeB 30.ixa5 h3 3 1 . CLl b6 CLlg5 and
Black's counterplay arrives in time) 29 . . . 'lWc7
30.'lWc5 CLlg5 and Black is active enough on
the kingside.) 2B.CLlxd6 'lWxd6 29.a4 Ei:xa5
30.bxa5 h4 3 1 .ic5 (3 1 .i.bB 'lWe7) 3 1 . . .'lWc7
32.Ei:b6 White has the better structure but
Black's position looks playable enough.
27 . . . Ei:g6 2 B . CLl c5
2 B . CLl a5 'lWc7 29.i.c5 h4 gives Black
reasonable counterplay.
h
26.ia7!
Karpov produces a curious echo of his
famous ia7 against Unzicker from 1 974 (see
Game 37 of the first volume) . On that occasion
the motive was purely positional, but here it
is connected with more immediately forcing
variations.
26.Ei:fl !? was an interesting alternative, which
can be compared with the previous note.
26 .. J::& a8? !
Kallai stays true to his plan and wins a tempo
that can be used to advance his kingside play,
but it turns out that the attack is not strong
enough.
a
b
c
d
e
f
g
h
2B . . . a5!
Black has to allow his opponent to create a
passed a-pawn in order to free his rook.
29.a4
29.CLlxa5? 'lWc7 traps the bishop, and after
30.a4 Ei:xa5 3 1 .bxa5 'lWxa7 Black is clearly
better.
29.bxa5 'lWe7 30.Ei:xb5 cxb5 3 1 . CLl b6 'lWxa7
32.CLlxcB 'lWxa5 33.CLle7 Ei:f6 White's knights
are strong, but Black's protected passed
d-pawn is also an important asset.
228
The Prime Years
29 . . J"lxb4 30.�xb4
30.j,b6 1'f!je7 Ieads to the same thing.
30 . . . axb4 3 1 .j,b6
Or 3 1 .�xb4 1'f!je7.
3 1 . . .1'f!je7 32.a5 lt:l g5
The position is dynamically balanced.
The a-pawn is undoubtedly strong, but
Black's kingside counterplay should not be
underestimated.
27..ib6
a
b
c
d
e
f
g
h
27...1'f!jh4!?
Kallai continues his policy of aggression. The
queen had several other options.
27 . . . 1'f!jf6?! From this square the queen
blocks the bishop from the defence of the
e5-pawn. 2B.�a5! h4 29.�f1 1'f!jg6 30.j,c7
1'f!jxe4 3 1 .�e l 1'f!jf5 32.j,xe5 It:l xe5 3 3 . lt:l cxe5
j,b7 34.f3 gxf3 3 5 . lt:l xf3 White wins a
pawn, which should give him good winning
chances.
From this square the queen defends the
weaknesses on c6 and e5, and later she might
also join in the attack from g6.
2B.�a5
White has a few other ideas:
2 B . lt:l c5 �g6 29.j,a5 j,e6 30.lt:lb6 �a7
3 1 .lt:lxe6 1'f!jxe6 32.lt:la4 lt:l d6 Black is alive.
2B .j,c5 It:l g5 29.lt:lb6 (29.j,d6 It:l f7 30.lt:lb6
It:l xd6 3 1 .lt:lxaB 1'f!jdB 32 . lt:l c5 �g6 33.�a5
h4 It is hard to j udge the position as the
aB-knight is trapped but hard to collect.)
29 . . . �bB 30.lt:lxcB 1'f!jxcB White keeps some
pressure after 3 1 .�c l or 3 1 .1'f!j e2, but Black's
position should be tenable.
2B . . . h4 29.1'f!je2 h3
29 . . . �g6 is met by 30.j,c7.
30.g3 �g6 3 1 .j,c7 �g5 32.lt:lb6 �a7 33.lt:lxc8
1'f!jxcB 34.j,b6
White keeps some initiative on the queenside,
but it is hard to say how serious his advantage
is.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
28J�a5!
Even in a hot situation Karpov is able to
find a cool-headed move. To send the rook on
a journey from f1 to a5 on a crowded board
takes great vision and creativity.
27 . . . 1'f!je7 28.�a5! h4 (if 2B . . . j,h6 29.�fl !
White opens the kingside) 29.1'f!je2 (29.�f1 ! ?)
29 . . . �g6 30.j,c5 White maintains some
pressure, but Black has reasonable chances to
defend.
28 .. J'�b8!?
27 . . . 1'f!jeB
A witty reply. Black would like to continue
the attack with . . . �g6, but at the moment this
1 993
229
Anatoly Karpov - Gabor Kallai
would drop the e5-pawn. Therefore Kallai
attacks the bishop in order to prevent the c4knight from taking on e 5 .
can be met by the calm 30.'kt>h l ! when Black
lacks a decent follow-up.
29 �e7?
...
I t i s possible that Gabor had previously been
planning 2B . . . g3 ?, but then wisely thought
better of it. 29.hxg3 Wxe4 If Black had time
to organize his position properly then the
removal of the e4-pawn would be a great
help to him, but in the present position it
merely opens additional lines of attack for
the opponent's pieces. 30.2"1e l Wg6 3 1 .Wb3
!l:bB 32.a4! White stops . . . 2"1b5 before going
after the e5-pawn. (White can consider both
32.'t'lc5 2"1eB 33.'t'lxa6 .ixa6 34.2"1xa6, and
32.1c7 2"1b5 33.'t'lcxe5 , but both are less
convincing than the main line.) 32 . . .Wg4
33.'t'lcxe5 't'lxe5 34.lLlxe5 White is winning;
he has an extra pawn, and Black's king is far
too exposed.
29.�e2?
Karpov tries to strengthen his kingside, but
paradoxically the position of the queen on
e2 adds strength to Black's attack, as a future
. . . 't'lf3 sacrifice might come with the gain of an
additional tempo after both sides capture on f3
with their g-pawns.
White could have shut down his opponent's
counterplay with: 29.lLlc5! 2"1g6 30.2"1b3! White
now has two rooks performing great work
laterally across the board. (30.lLlxa6?? is much
too greedy: 30 . . ..ixa6 3 1 .2"1xa6 lLlg5 32.2"1a5
[32.i>h l 2"1fB-+] 32 . . . lLl f3t Black's attack
breaks through.) 30 . . . 2"1aB (3o . . . lLl g5 3 1 .2"1g3!;
30 . . . \&f6 3 1 .lLlxa6) 3 1 . lLl d3 White keeps the
advantage while keeping his opponent's attack
at bay.
There was a second strong continuation in
29.a4!, improving White's position on the
queenside and challenging Black to come up
with a good plan. The key point is that 29 . . . g5
Gabor commits a sin that many of us do
from time to time: having embarked on a
plan, he changes his mind and reverts back
to a plan he rejected earlier in the game. fu
I have mentioned several times throughout
the two volumes, Karpov almost never
did this.
29 . . . 2"1g6? was no good as the e-pawn falls:
30 . .ic7 2"1b5 (30 . . . lLlg5 3 1 .lLl cxe5 wins.)
3 1 .2"1xb5 cxb5 32.lLlcxe5 Black is unable
to generate any meaningful threats on the
kingside.
The best move was 29 . . . lLlg5!. This leads
to serious complications and some magical
tactics. Despite his almost unparalleled feeling
for chess, it is possible that even Karpov would
have lost his way in the j ungle. White has
several plausible moves, but only one route to
an advantage.
a) 30 . .ic7? lLl f3t! 3 1 .gxf3 gxf3 32.Wxf3 2"1g6t
wins for Black.
b) 30.2"1e l ? lLl f3t! 3 1 .gxf3 gxf3 32.Wxf3 2"1g6t
33.'kt>h l .ig4 34.Wg3
a
b
c
d
34 . . . .ie2!! Black wins.
e
f
g
h
230
The Prime Years
c) 30.Wd 1 ? lLl f3t! 3 1 .gxf3 gxf3 32.�h 1 E:g6
33.lLlcxe5
a
b
c
d
e
f
g
e) 30.�h 1 lLl f3 3 1 .h3 (3 1 .gxf3! ? looks risky
but j ust seems to be playable: 3 1 . . .gxf3
32.Wfl E:f6 33.E:xe5 [33. lLl dxe5 ? i.h3 wins.]
33 . . . i.h3 34.Wc 1 i.g2t 3 5 .�gl E:g6 36.�f4
Black does not seem to have more than a
draw.)
h
33 . . . i.xe5! (33 . . . i.h3 leads to a draw after
34.lLlxg6 Wg4 3 5 .E:xh5t i.h6 with a perpetual
to follow.) 34.lLlxe5 i.h3 3 5 .Wb3 (3 5 . E:xa6
E:b7-+) 35 . . . E:g7! (35 . . . E:b7? 36.lLlxg6 Wxf2
37.E:xh5t �xg6 3S .WgSt White escapes with
a perpetual.) 36.lLlf7 E:bgS 37.i.c7 i.g2t
3S. �gl i.fl t Black wins.
d) 30.We I ! ? Unlike the retreat to d 1 , this has
the advantage of defending the f2-pawn. Even
so, White has no more than equality here.
30 . . . lLl f3t 3 1 .gxf3 gxf3 32.�h 1
a
b
c
d
e
f
g
h
32 . . . E:xb6!? (Black can also force a draw
with 32 . . . E:g6 33.lLldxe5 i.h3 34.lLlxg6 Wg4
3 5 . lLl f4 i.g2t 36.�gl i.h3t with a perpetual.)
33.lLlxb6 E:g6 34.lLlxcs Wh3 3 5 .Wg 1 E:xg 1 t
36.E:xg 1 WxcS The position is approximately
balanced.
c
d
e
f
g
h
3 1 . . . lLl g5 ! The knigh t returns, having succeeded
in provoking a weakness. 32.E:e 1 gxh3 33.g3
Wg4 34.f4 lLl f7 The position is roughly
equal.
f) 30.lLl e l !
a
a
b
b
c
d
e
f
g
h
This prophylactic move is the only way
for White to keep the advantage. 30 . . . E1b7
(30 . . . g3 3 1 .hxg3 Wxe4 might be playable, but
Black is definitely worse.) 3 1 .E:b3! The rook
does a great job of defending laterally along the
third rank. 32 . . . E:g6 32.E:g3 Black's attack is
going nowhere, while White keeps a promising
initiative on the queenside.
1993
23 1
Anatoly Karpov - Gabor Kallai
Also after 32 . . . Wc7 the position is roughly
balanced.
33.�a7 1"i:a8 34.llJc5 �e8
This is not a serious mistake, but I prefer
34 . . . We7! in order to take away the c7- and
d6-squares from the white bishop, and after
3 5 .j,b6 h3 36.g3 ttJ d6 the position is about
equal.
35.�b6 g3
a
b
c
d
e
f
g
h
30.a4?!
In a complex situation Karpov allows his
opponent back into the game. The queen's
retreat to e7 took a lot of the power out of the
attack, so this would have been a good time
to cash in on the queens ide with 30.ttJc5! 1"i:g6
3 1 .ctJxa6 j,xa6 32.1"i:xa6 when Black does not
have much compensation for the pawn.
3 5 . . . We7 36.b5!? maintains some pressure.
It was worth considering 35 . . . h3 36.g3 ttJg5
37.ttJd3 ttJ f3 with roughly equal chances in a
complex position.
36.llJd3
After 36.h3 gxf2 37.1"i:fI 1"i:g3! Black has good
counterplay.
30 ... 1"i:g6!
Black correctly switches his attention back
to the kingside.
a
b
c
d
e
f
g
h
36 gxh2?
..•
With little time remaining, Karpov finds a
good prophylactic move.
The huge strategic fight combined with a lot
of calculation finally took its toll, and Kallai
makes a fatal error. In a funny way Karpov's
time trouble may even have worked against
Gabor, by adding even more pressure to the
already tense situation. There were several
interesting alternatives.
31...h4 32.�c5 �e6
36 . . . 1"i:g5
a
b
c
d
e
f
g
h
31.@hl!
defends
the
e5-pawn.
37.j,c7
232
The Prime Years
(37.fxg3 hxg3 3B .h3 i.xh3 39.gxh3 'iffi e6
White's advantage has evaporated.) 37 . . . gxh2
(37 . . . Ei:a7!?) 3 B . tLl b6 i.g4 39.f3 Ei:a7 with huge
complications.
36 . . . cj;gB !? 37.i.c7 Black can defend his
e-pawn indirectly, by pawn by putting his king
on a safer square: 37 . . . gxf2 3 B.'iffixf2 h3 39.Ei:gl
c5 40.tLlxc5 'iffi c6 The position is messy.
37.c!LJdxe5!
After conducting the strategic aspect of
the game with great skill and imagination,
Karpov seizes his opportunity to win the game
tactically.
37... gh6
37 . . . tLlxe5 3B.'iffih 5t <;t>gB 39.tLlxe5 wins.
38.,hd4
36 . . . gxf2!? Taking the other pawn also drops
the e5-pawn, but the conditions are much
more favourable for Black than in the game.
37.tLldxe5 tLlxe5 3B.tLlxe5
White wins a second pawn and the rest is
simple.
38 ... c!LJg5 39.£4 c!LJe6 40.i.e3 h3 41 .g4
42.bxc5 'l'Nd8 43.'l'Nd2
1-0
c5
The whole game was rich in both strategic
and tactical ideas. Kallai held his own for a
long time, but after a terrific struggle Karpov
eventually caught him.
***
a
b
c
d
e
f
g
h
3B . . . h3! Thanks to this lovely move Black can
stay in the game. (3B . . . i.xe5 39.'iffi h 5t <;t>g7
40.Ei:xe5 h3 4 1 .Ei:g5 +-) 39.'iffih 5t Ei:h6! 40.'iffixeB
fl ='iffi t 4 l .Ei:xfl hxg2t 42.<;t>xg2 i.h3t 43.<;t>g3
Ei:xeB 44.Ei:£7 i.e6 Black is not worse.
8
l .e4 c6 2.d4 d5 3 . tLl d2 dxe4 4.tLlxe4 tLld7
5 . tLl g5 tLl gf6 6.i.d3 e6 7.tLl I f3 i.d6 B.'iffie2 h6
9 . tLl e4 tLlxe4 1 O.'iffixe4 tLl f6 1 l .'iffi h4
Nowadays most experts prefer 1 1 .'iffi e 2.
8
7
7
6
6
5
5
4
4
3
3
2
2
1
Karpov's next tournament was Dortmund. In
the first round he unleashed a fantastic novelty
against Kamsky in the 4 . . . tLl d7 Caro-Kann.
We will look at it briefly:
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
1 99 3
233
Christopher Lutz - Anatoly Karpov
1 1 . . .�e7!!
The point behind this astonishing move
is that Black's queen is now defending his
rook, thus making . . . g5 into a serious threat.
Karpov's idea has stood the test of time; it
is still regarded as the strongest move in the
position, and has been repeated in over a
hundred games to date.
1 2.tLle5
Kamsky chooses the most principled answer
and sacrifices a pawn.
12 . . . �xe5 1 3 .dxe5 �a5t 1 4.c3 �xe5t 1 5 .�e3
b6 1 6.0-0-0 g5 1 7.�a4
17.�h3 scores better.
1 7 . . . c5
An interesting middlegame ensued in
Kamsky - Karpov, Dortmund 1 993. For a
long time Black was better, but later he went
wrong and allowed his opponent to get back
into the game. But Kamsky failed to find the
right path, and after a tough battle Karpov
eventually secured the full point.
In Round 2 Karpov had the black pieces for the
second time. His opponent was Christopher
Lutz. The German grandmaster played sixteen
games against world champions, scoring one
victory, five draws and ten defeats. He made
two draws and two losses against Karpov,
including a draw in their one previous meeting
before the present game.
Black's idea is to lose a tempo with the bishop
in order to divert the white queen from the
d-file, thus preparing . . . c5 without allowing
the reply d5.
6.tDc3 c5 7.e4 cxd4 8.tDxd4 tDc6 9.tDxc6
.ixc6 IO .if4 1L1h5 1 l ..ie3
.
a
b
c
d
f
e
g
h
1 1 ...�b8
1 1 . . . �c5 is a popular alternative, but
Karpov's choice is the main line. Black
stops f4 and fights for control over the dark
squares.
Earlier in the year Karpov opted for an
alternative with similar motives, but which led
to disastrous consequences: 1 1 . . . �d6??
I Game JI I
Christopher Lutz
-
Anatoly Karpov
Dortmund
1 993
l .d4 tLlf6 2.c4 e6 3.tiJf3 b6 4.a3 .ia6
Karpov's usual preference was 4 . . . �b7.
He played the text move only four times in
his career, the present game being the last of
them.
a
b
c
d
e
f
g
h
1 2.�d l ! 1-0 Christiansen - Karpov, Wijk aan
Zee ( 1 ) 1 993.
The Prime Years
234
12.g3
16 . .!iJxe4 .!iJxe4
A few rounds later Lutz played the same
variation from the opposite side of the board
against Kramnik, who deviated at this point.
The game continued 1 2.0-0-0 �d6 1 3 .g3
�e5 and was eventually drawn.
In view of the improvement mentioned in
the note to White's 1 9th move below, it seems
likely that 1 6 . . . �xe4! is an objectively sounder
move. After 1 7. fxe4 �d6 1 8 .l"i:hf1 ( l 8 .�f4!?)
1 8 .. .'IWc7 Black has good control over the
dark squares and will be able to castle without
difficulty.
8
KS
�.i
/�.m
ti""'%� r_'l�_'l
" ' f�
".
�
�
"�
�
�
� � �
�
�I
�
�
: ';: �
!
!
�% "/,, ��,;: ��,�
3 � � � r[j
%�'if�� �� %W(f
� � =:t.:
7
6
2
1
' ' ' ' '
a
'''''
b
c
d
e
' ' ' '
f
g
h
12 ... f5!?
This was Karpov's novelty, which has since
become the main move in the position.
Black wants to fight for the initiative instead
of merely settling for a solid Hedgehog-type
position.
13.0-0-0
1 3 .l"i:g 1 has since become more popular,
although Black scores well there too.
17.fxe4
1 7.�xe4!? would have stopped Black from
putting his bishop on d6, but on the other
hand the c4-pawn might become vulnerable.
1 7 . . .�xe4 1 8 . fxe4 ( l 8 .1Wxe4? 1Wxe4 1 9 .fxe4
l"i:c8 The c-pawn falls.) 1 8 . . . l"k8 1 9. � b 1 �e7
( l 9 . . . b5!? 20.c5 �e7) 20.l"i:hfl l"i:f8 Since
the white pieces are tied up with defending
the c-pawn, Black will probably be able
to castle by hand, thus keeping a good
position.
17...�d6 18.ghfl �e5!
Karpov blocks the e4-pawn and renders his
opponent's light-squared bishop very passive.
The c6-bishop is also restricted, but it enjoys
better prospects than its opposite number.
Positionally Black is doing extremely well - the
only drawback to his position is his difficulty
castling.
13 ... .!iJf6 14.�d3 'lWb7!
By forcing White's next move, Karpov
increases his influence over the central dark
squares.
15.£3
Karpov mentions the extravagant 1 5 . ttJd5?,
but fails to point out the refutation of
15 ... exd5! 1 6.exd5 �xd5 1 7.cxd5 l"i:c8-+ .
15 ... fxe4
Karpov makes sure he will get control over
the e5-square.
a
19.�f4?!
b
c
d
e
f
g
h
1 993
235
Christopher Lutz - Anatoly Karpov
Lutz's plan is to entice Black into short
castling with the intention of starting an attack
there, but it never materializes.
White can try to prevent long castling with
1 9 .ig5 , but after 1 9 . . . h6! 20.,if4 Wb8
(20 . . . Wc7?! 2 1 .Wc3!) 2 1 .We2 (2 1 .,ie3 ,if6)
2 1 . . .0-0 Black has a pleasant game, and the
slight weakening of his kingside caused by
. . . h7-h6 is of little consequence.
19.'lWe2! was the best way to cause problems:
19 . . . 0-0-0 (If Black stubbornly tries to
arrange short castling then he could easily fall
into serious trouble: 1 9 . . . ,if6?? 20.e5 wins,
and 19 . . . Wc7?! 20.,ig5 ! is strong.) 20.cS <tt> b 8
2 1 .cxb6 axb6 22.<tt> b 1 White is slightly better.
Note that Black cannot take the e-pawn under
favourable conditions as 22 . . . ,ixe4? (Black
should instead prefer 22 . . . d6 23 .,if4;!;) 23.,ig5!
is strong.
19 Wb8!
Karpov keeps his control over the e5-square
with the more precise move.
a
b
c
d
e
f
g
21 ..Jhf4!
Karpov finds an excellent way to extinguish
his opponent's initiative. In return for the
exchange, Black gets a pawn and fortifies his
control over the central dark squares.
2 1 . . .,ixf4 t? 22.gxf4 Ei:f7 (22 . . . Ei:xf4 23.eS)
23.e5 g6 24.Ei:g 1 White has a powerful attack
on the g-file.
..•
Lutz was probably hoping for 1 9 . . . Wc7
20.'lWc3! ,ixc3 (20 . . . d6 2 1 .,ixe5 dxe5 is
possible, but Black is obviously worse due to
his pawn structure.) 2 1 .,ixc7 Ei:c8 22.,id6 ,if6
23.eS ie7 24.<tt> d2 when White has a small
advantage.
20 We2
White is somewhat ahead in development
and tries to make something of it.
.
Another idea was 20.,ig5 to prevent Black
from castling, but after 20 . . . h6 2 1 .,ie3 Wc7
22.Wb 1 (22.,ie2 Ei:f8) 22 . . . Ei:f8 23 .,ie2 c.j;; e7
Black's king stands safely in the centre.
20
.
..
0-0 21 .�h5
22.gxf4 hf4t 23.';t> b l �e5!
Black must keep a piece on this square in
order to prevent White from liberating his
bishop with e4-eS. Despite his slight material
disadvantage, Black runs virtually no risk of
losing this position as long as he is careful to
maintain control over the dark squares.
24.�xe5?!
Exchanging queens leads to a passive
position for White, who will have a hard time
improving his pieces.
24.We2!?
With queens on the board Black will not
be able to activate his king like he does in
the game. This move allows Black to take
a second pawn; on the other hand, fewer
white pawns mean more open files for his
rooks!
24 . . . ,ixh2
236
The Prime Years
Black is not forced to take the second pawn,
and can also consider 24 . . . i.g5 ! ?
2 5 .Wf3 h6
After 25 . . . Wb8?! 26.:§:f2 i.e5 27.:§:h l White's
attack is dangerous.
26.:§:d2 i.g3
26 . . . �h8 27.:§:g2 Wd6 28.:§:xg7 �xg7
29.Wf6t leads to a perpetual.
27.:§:g2 i.h4 28 .Wf7t �h8 29.Wg6
Black has enough resources to defend his
king, but he is in no position to advance his
passed pawns on the kingside, therefore the
position is balanced.
24 ...he5 25.h3
It is hard for White to justifY the sacrifice
2 5.b4 i.xh2 26.b5 i.b7 27.i.c2 d6 when
Black's passed g- and h-pawns give him decent
winning chances.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
25 ... a5
Karpov stops his opponent from gaining space
on the queenside - a sensible idea, although
in view of improvement mentioned in the
following note, 25 . . . d6 was worth considering.
26.b3?!
Lutz does not want to allow his queenside
pawns to be fixed, but this should not have
been his top priority. Instead 26.:§:f3! followed
by :§:dfl would have prevented the following
bishop manoeuvre, after which Black will have
a hard time improving his position.
26 ... d6!
Preparing to improve the light-squared
bishop. Now Black gets everything he could
have hoped for.
27.�d2 i.e8 28.@c2
28.:§:df2 is met by 28 . . . i.g6 when Black
covers all the entry squares on the f-file.
28 ...i.g6
The bishop arrives on its perfect square,
from which it menaces the e4-pawn while
conveniently guarding the entry square on f7.
29.�df2 �c8 30.�dl?!
White's defensive task is not easy - he cannot
anticipate exactly how his opponent will try
to exert pressure. It is never easy to choose
between numerous defensive plans, when all
of them result in a slightly worse position.
With the benefit of hindsight, it is apparent
that queens ide play does not achieve much
for White. Therefore it looks better for him to
wait patiently with 30.�d2, keeping his rooks
on the f-file in order to keep the enemy rook
tied to the eighth rank. Play might continue:
30 . . . i.d4 3 1 .:§:f3 i.c5 32.e5!? Sacrificing the
e-pawn to open files for the rooks. (Alternatively
32.a4!? is passive but Black will have a hard
time breaking through.) 32 . . . i.h5 33.:§:f4
(33.:§:g3!? i.xa3 34.exd6 is also possible.)
33 . . . dxe5 34.:§:h4 g6 3 5 .:§:e4 i.d4 36.ie2
White should be able to hold this ending.
30 ...i.d4!
Karpov forces the rook away from the f-file
and opens the fifth rank for the rook.
3 1 .�a2
3 1 .:§:f4? i.e3 costs White his extra exchange.
237
Christopher Lutz - Anatoly Karpov
1 993
8
much danger.) 36.�b3 .ia5 (36 . . . .ic5 ??
37.:9:a2!+-) 37.:9:hf2 h6 38.:9:f8t �h7
39.:9: 1 f3 :9:e5 40.gd8 White has nothing to
worry about as his pieces are active enough.
33 . . . :9:xh3
33 .. J!:h4 34.:9:d2 :9:xe4 35 . .ig4 should be
fine for White.
7
6
5
4
3
2
a
b
c
d
e
f
g
h
31..J�c5!
Karpov transfers the rook to the kingside
where it helps to keep up the initiative.
a
32.g g2
The German grandmaster changes his mind,
and decides not to take any action on the
queenside. More challenging looks:
32.b4! ? gh5!
I believe this to be the strongest answer.
Karpov only mentions 32 . . . gg5 , and Lutz's
move in the game suggests that he was also
concerned about the rook occupying the
g-file. However, White has a surprising
way to activate his pieces: 33 . .ie2! .ixe4
(33 . . . h5 34 . .if3 :9:g3 3 5 . :9:g2! Black's
rook will be eliminated, along with all his
winning chances.) 34.:9:d2 .ie5 (If 34 . . . e5
35.c5! White becomes active.) 3 5 . .ig4 .if5
36.ixf5 exf5 37.:9:d5 c;t>f7 38 .:9:b5 White has
succeeded in activating his rooks, and the
position is unclear.
33.ie2!
White's counterplay is based on the eighth
rank.
33.Elh2 looks passive, but is not a bad move.
33 . . . axb4 (33 .. J''1 h4 34.�c2) 34.axb4 .ic3?!
This ambitious approach gets nowhere.
3 5. c;t>c2 .ixb4?! (Black should probably
admit his mistake with 35 . . . .if6, although
after 36.c;t>b3 White should not be in too
b
c
d
e
f
g
h
34.:9:d2!
White plays to exchange the rooks rather than
the bishops. If 34 . .ig4 .ih5 35 . .ixh5 :9:xh5
36.gd2 e5 Black keeps some advantage.
34 . . . .if6
34 . . . .ie5 ?? 3 5 . .ig4 .ih5 36 . .ixh5 Elxh5
37.c5! wins for White.
35 . .if3 gxf3
35 . . . .ie7 is met by 36.e5 ! .
36.:9:xf3 .ih5 3 7 . c;t>e2
White should be safe as Black has no rook.
32 J�h5 33J�f3
..
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
238
The Prime Years
33 ...i.e8!
The bishop has done its duty on g6. With
this subtle retreat, Karpov clears the path of
the g-pawn in order to improve his king.
38 .h4? is a mistake due to 38 ... j,h5 39.l"1fg3
j,xg4 40.l"1xg4 j,e3 (40 . . . h5 4 1 .l"1xg5 t should
be a draw) 4 1 .hxg5 j,xg5 Black has excellent
winning chances as his extra pawn is both
distant and passed.
34.�c2
7
White could have obtained excellent drawing
chances with the resolute move: 38.l"1fg3!
h6 (The alternative is 38 . . . c;t>f6 39.h4 h6
[39 . . .gxh4 40.l"1f3t!] 40.hxg5 t hxg5 4 1 .l"1f3t
when White's activity should enable him to
hold.) 39.h4 j,h5 (39 . . . \t>g6 40.hxg5 hxg5
4 1 .b4 White is safe as he can play l"1h3 on the
next move, and 4 1 . . .\t>h5? achieves nothing
after 42.j,e2.) 40.hxg5 ! White solves his
problems by returning the exchange. 40 . . . j,xg4
4 1 .gxh6t \t>xh6 42.l"1xg4 Black has a slight
positional advantage, but virtually no winning
chances.
5
38 ...i.c5 39.�b2
3
39 .a4?! completely immobilizes White's
queenside, and might allow a king invasion to
b4 later in the endgame.
34.b4!? deserved consideration.
34 ...g6 35.i.e2 �e5!
36.l'!d3 was a serious threat, so the rook
had to leave the h-file. Karpov chooses a good
square, where the rook defends e6 and attacks
e4.
36.i.d3 �g7 37.�g4
8
6
4
2
39 ...i.g6
a
b
c
d
e
f
g
h
37 ... g5!
Having improved his king slightly, Karpov
starts pushing his kingside pawns, which he has
to do at some point if he is to generate serious
winning chances. It is quite possible that
Karpov deliberately timed it to occur shortly
before the time control, when his opponent
would have the most difficulty reacting to the
change in the position.
38.�fl
This is playable, but rather passive. It is
possible that Lutz was short of time and simply
wanted to get closer to move 40.
39 . . . j,h5 40.l"1g2 j,g6 4 1 .l"1e2 defends.
40.h4!?
Finally Lutz decides to change the pawn
structure of the position. It clears some space
for the rooks, but the black rook turns out to
be the one which is best placed to utilize the
g-file.
Waiting passively would have led to a joyless,
bu t still tenable position for White, for
instance: 40.j,c2 d5!? (After 40 . . . h5 4 1 .l"1g2 I
do not see how Black can effectively improve
his position. If his king goes to h6 then l"1f6
will be annoying, and if he puts his bishop on
f4 then white will play l"1d 1 .) 4 1 . \t>c3 (4 1 .cxd5
exd5 42.exd5 [42.b4 j,d4t] 42 . . . l"1e2 43 .l"1c4
1 993
239
Christopher Lutz - Anatoly Karpov
id3 44.�d l j,xc4 4 5 .bxc4 Black is j ust a
fraction better.) 4 1 . . .dxe4 (4 1 . . .j,xa3 42.exd5)
42.h4 axb4t 43.axb4 j,e7 44.�e l White
holds.
It is understandable that White does not
wish to calculate the consequences of a bishop
check on c3 .
40 ... gxh4 4Uhh4
a
b
c
d
e
f
g
h
47 ... gg5 48.!!fhl i.e5 49.gh3 i.d4
41...!!g5!
Going after the a-pawn with 49 . . . �g2t
50.'it>e3 �a2 does not win a pawn due to
5 1 .Elb l ! .
Karpov occupies the file that was j ust opened.
His rook gets stronger and stronger.
50.!!3h2 !!g3
a
b
c
d
e
f
g
h
Karpov improves his rook a bit.
Black had another strong continuation in
41 . . . d5!. Karpov evaluates this move as winning
for Black, but this seems too optimistic. Black
indeed wins a pawn, but White can try to
sacrifice an exchange for one of the pawns
and hope to survive in the opposite-coloured
bishop endgame a pawn down. 42.�d 1 (42.j,c2
dxe4 43.�h2 �f5 44.�e l �f4 45.�he2 j,f2
46.�xe4 White is struggling but has chances
to hold.) 42 . . . dxe4 43 .j,c2 h5 44.�d2 (44.�e l
if2 45.�hxe4 j,xe4 46.�xe4 �xe4 47.j,xe4
also gives some drawing chances.) 44 . . . 'it>h6
45.Ele2 White is clearly worse but he certainly
has chances to survive.
42.!!h2 !!g3 43.i.c2 i.d4t 44.<j;>c1 a4!
Karpov creates
queenside.
a weakness
in White's
45.bxa4 !!xa3 46.<j;>d2 !!g3 47.i.d3
5 1 .gfl i.f6
5 1 . . . d5 can be met by 52.�e2, so Karpov
continues to manoeuvre in an effort to wear
down his opponent.
52.!!bl i.g5t 53.<j;>c2 i.e3 54.gal
Knowing that he may have to suffer for a
long time to come, Lutz sensibly decides to
exchange his weak a-pawn.
Stopping the king from marching to the
centre with 54.�f1 ? costs White a pawn after
54 . . . d5.
More resilient was: 54.Elbh l ! \!If6 Obviously
Black could shuffle around for a few mores
in order to tire his opponent, but at some
point he will have to take action. (It it worth
pointing out that 54 . . . d5? does not work, and
240
The Prime Years
after 5 5 .exd5 .ixd3t 56.\t>xd3 .ig l t 57.\t>e4
White may even start pressing.) 5 5 .�xh7!?
Perhaps White can hold by waiting passively,
but attempting to break free is much more
natural. (Another idea is 5 5 .e5t \t>xe5 56 . .ixg6
hxg6 when it is hard to say how great Black's
winning chances are with two pawns for the
exchange.) 5 5 . . . .ixh7 56.�xh7 me5 57.mc3
.id4t 58 .\t>c2 �g2t 5 9.\t>b3 �b2t 60.ma3
White is extremely passive, but in view of the
limited number of pawns remaining, he has
reasonable chances to hold.
56.a5
Lutz gets rid of a weakness, yet his suffering
is hardly eased.
Thanks to Karpov's last move, White had an
opportunity to swap off his bad bishop with
56.e5t! dxe5 57 ..ixg6 mxg6 5 8 .a5 when he
has decent drawing chances .
56 ... bxa5 57.�xa5
Sacrificing the e-pawn was still possible:
57.e5t!? dxe5 (57 . . . \t>xe5 58 ..ixg6 hxg6
59 .�e l t! .ie3 60.�d3) 5 8 . .ixg6 mxg6 59.�xa5
.id4 60.�d3 White should be able to hold this
endgame.
57...i.c5 58.�al
8
7
6
5
a
b
c
d
e
f
g
h
54 i.gl !
•.•
4
3
2
Karpov shows fine j udgment i n rejecting
the tempting 54 . . . d5, which wins a pawn
but allows White to reduce the number of
remaining pawns significantly: 5 5 . cxd5 exd5
56.a5! The following line looks frightening
for White, but he seems to get away with it:
56 . . . dxe4 57 . .ie2 .id4 5 8 .�a4! .ie5 59.�h l !
�c3t 60.\t>d l e3 6 1 .�gl bxa5 62.�xa5 .ic7
63.�a6 �b3 64.�c6 After walking a narrow
path White manages to survive.
59.�f1 i.gl
55.�d2 \t>f6
Karpov hints at the plan of pushing the
h-pawn to h2.
Karpov centralizes his king, but there was an
even stronger continuation in 5 5 . . . e5! 56.a5
bxa5 57.�xa5 h5 when Black has excellent
winning chances.
1
a
b
c
d
e
f
g
h
58 \t>e5!
.•.
Karpov centralizes his king and finally
prevents his opponent from exchanging his
passive bishop by sacrificing his e-pawn.
60.�ddl i.e3 61 .�f8 �g2t 62.\t>b3 �h2!
Taking away the h-file from the opponent.
63.i.bl?
1993
63.ic2 should have been preferred, as the
dl -square is important in some variations, for
instance: 63 . . . ih5 64J'1d3 ic5 (64 . . . id4 can
be met by 65.!'i:g3! as with the bishop on c2,
Black does not have a check on the d l -square.)
65.mc3 Black can continue to press, but White
has reasonable chances to survive.
White should have preferred 65.!'i:cl , defending
the c-pawn and keeping the rook closer to the
king. 65 . . . ig4 (65 . . . ic5 66.!'i:ffl ) 66.!'i:gB (or
66.ma4 h5 67.!'i:gB) 66 . . . h5 67.\t>a4 White's
position remains unpleasant, but at least he
has avoided a quick collapse.
6S ...J.cS 66J�el?
8
Lutz wants to defend the c4-pawn with
his very passive bishop, but his king ends up
becoming vulnerable. White's position was
highly unpleasant, but it was still possible to
offer tough resistance.
7
6
5
4
3
24 1
Christopher Lutz - Anatoly Karpov
66.!'i:gB ie2 67.!'i:g5t \t>d4 6B.!'i:cl (6B.!'i:fgl
ixc4t) 6B . . . h5 69 .!'i:c3 h 4 7o .ic2 White is
dangerously tied up.
V////O/// \.....L..J
2
a
b
c
d
e
f
g
h
66.!'i:c l ! was the best chance, and after 66 . . . ie2
67.!'i:hB (67.!'i:gB h5) 67 . . . h5 6B.\t>a4 if3
69.!'i:fl White still has chances to survive.
63 ...J.hS!
Karpov finds a more effective role for the
bishop than attacking the e4-pawn.
64J:�el
The main alternative was 64.!'i:d3 id4
65 .ic2 (65 .\t>b4 !'i:b2t! 66.!'i:b3 ic5 t! 67.\t>c3
Elh2 Black has nicely cut the b3-rook out of
play and is now ready to push his h-pawn.)
65 ... ie2 66.!'i:g3 ixc4t 67.\t>xc4 !'i:xc2t
68.mb3 !'i:b2t 69.\t>a3 !'i:h2 70.!'i:g4 (70.!'i:ff3
�xe4 7 1 .!'i:h3 !'i:d2) 70 . . . h5 7 1 .!'i:g6 and it is
hard to say whether Black can win.
64...J.f2
64 . . . ic5 was a good alternative, but perhaps
Karpov was looking to provoke White's next
move.
65.!'i:f1?
Lutz attacks the f2-bishop, but helps the
more dangerous bishop to join the attack with
a huge tempo.
a
b
c
d
e
f
g
h
66 ...J.e2!
Throughout these two volumes, I have often
commented on Karpov's remarkable ability
to catch the opposing king in endgames. The
end of the present game provides yet another
compelling example.
67.J.a2
67.!'i:c l would now be too late, as after
242
The Prime Years
67 . . . .te3 6S.Ek3 .td2
exchange and loses.
White
drops
the
67 ... �h3t 68.Wb2 �a3t 69.Wa1
The king is forced into the corner. The a2bishop is contributing nothing to the defence,
and even gets in the way by blocking a possible
escape square from the king.
69.�c2 loses to 69 . . . .td3t 70.�b3 .tc5 .
***
69 ...�d3 70.�bl
70.l"i:bS .tc5 7 1 .�b2 .td4t wins.
a
b
c
Samisch King's Indian. In the next game he
suffered a setback against Lautier, who played
an enterprising pawn sacrifice leading to a
dangerous attack which decided the game.
Karpov recovered well to register good wins
against Serper and Lobron, before agreeing a
quick draw with Kramnik in the last round.
Karpov won the tournament with the superb
score of 5 Y2/7, a point and a half clear of
Kramnik and Lutz.
d
e
f
g
h
70 ...�b4!
Now White must lose too much material to
avoid checkmate.
71 .�cl
After 7 1 ..txd3 .txe l the rook and opposite­
coloured bishop endgame offers White little
chance of survival, as his king is weak and his
bishop is too passive.
71 ...�d2 72.�dl �c3t 73.Wa2 hc4t
74.Wa3 �e2
0-1
Karpov did not slow down, and in the next
round he won a surprisingly one-sided game
against Dolmatov on the white side of a
Karpov's next tournament was in Dos
Hermanas. In the first round he played
the English Opening against Epishin, but
achieved no advantage and an early draw
ensued. In Round 2 he equalized against Izeta
Txabarri with the Caro-Kann, then outplayed
his Spanish opponent and won. Karpov's next
game against Adams featured an unusual
opening which looked like a hybrid between
a Sicilian, Hedgehog and Queen's Indian.
Karpov castled long and developed a promising
initiative, which allowed him to simplifY to an
advantageous four-rook endgame, which he
converted smoothly.
In Round 4 Fernandez Garcia played the
Griinfeld, but posted three of his pieces on
the a-file and Karpov punished him on the
kingside. In the next round Magem Badals
refused the famous ltJ g5 repetition in the
Zaitsev, but went wrong in the Breyer-type
middlegame and Karpov punished him.
Karpov then scored his fifth consecutive win
- his victim this time was Yudasin, who played
a Queen's Indian with hanging pawns, but
failed to equalize and despite stiff resistance he
eventually went down.
In Round 7 Karpov had to defend a dangerous
attack against Khalifman, but found a good
solution, giving up his queen for a rook and
knight under relatively favourable conditions.
The game was subsequently drawn. In the next
round Judit Polgar played the King's Indian and
1993
24 3
Anatoly Karpov - Miguel Illescas Cordoba
Karpov reacted with the Fianchetto Variation.
After some pawn exchanges Karpov emerged
with the better structure, and he was able to
press home his advantage. In the ninth and
final round Karpov made a solid draw with
Rivas Pastor to finish on the excellent score of
7Y2/9, a point clear of Polgar who was second.
***
Karpov's next tournament was in Leon in Spain,
where his first opponent was the thirteen-year­
old Peter Leko. Your author was Peter's trainer
at that time, having started to work with
him three years before. Competing in such a
tournament was a big achievement for Peter,
nevertheless after playing a quiet line against
the Caro-Kann, he reached an equal ending
and was able to survive despite Karpov's best
efforts to squeeze a win from the position.
In Round 2 Karpov was close to grinding
out a win against Morovic Fernandez, but the
Chilean grandmaster defended extremely well
and narrowly managed to survive for a draw.
In Round 3 against Garcia Illundai Karpov
played a Queen's Indian with hanging pawns,
but failed to equalize. For a long while he
looked to be in trouble, but his opponent did
not make the most of his advantage and a draw
ensued.
In Round 4 Karpov broke his run of draws
and overcame Sion Castro with the help of a
strong positional exchange sacrifice. But then he
only managed to draw a back-and-forth game
with White against Romero Holmes. There
followed a quick draw against Vyzmanavin, a
longer one against Topalov, and a short draw
with Black against Yudasin.
In the ninth round Karpov faced Miguel
Illescas, who played sixty nine games against
the world champions. He won two of those
games, drew thirty five and lost the other
thirty two. Karpov played him fourteen times,
winning seven, drawing six and losing one.
I Game 3l 1
Anatoly Karpov - Miguel Illescas Cordoba
Leon
1 993
l .c4 c5 v!tjf3 t£Jf6 3.�c3 e6 4.g3 �c6 5.i.g2
d5 6.ad5 exd5
Through a slightly unusual move order we
have arrived at a Tarrasch Defence. It is a
somewhat risky choice against Karpov, who
proved throughout his career that he was
tremendously adept at playing against a fixed
weakness, such as the isolated pawn which
often arises in this opening.
7.d4 i.e7 8.0-0 0-0 9 ..ig5
Karpov plays the main line, with which he
twice defeated Kasparov.
Nine years before, Karpov won another
interesting game using the second most popular
move: 9.dxc5 �xc5 1 O.�g5 d4 1 1 .�xf6 Wfxf6
1 2 .ttJd5 Wfd8 1 3 .ttJ d2 2"i:e8 1 4.2"i:cl �b6
1 5 .2"i:e 1 �e6 1 6. ttJ f4 �xa2 1 7.b3 �a5 1 8 .2"i:c2
�xb3 1 9 .ttJxb3 d3 20.2"i:xc6 �xe l 2 1 .2"i:cl d2
22.2"i:b l A highly irregular position has been
reached. Karpov skilfully manages to turn the
seemingly powerful d2-pawn against its owner,
by exploiting the fact that the unusually placed
bishop on e 1 is immobile.
a
b
c
d
e
f
g
h
22 . . . a5 23.ttJd3 Wfg5 24.ttJ bc5 2"i:ad8 25 .�xb7
244
The Prime Years
h5 26.if3 Wf5 27.'it>g2 h4 2B.g4 Wg5
29.h3 Ei:d4 30.Wb3 g6 3 1 .e3 Ei:ddB 32.tLle4
a4 33.Wxa4 We7 34.g5 Ei:xd3 3 5 . tLl f6t 'it>fB
36.Wxh4 WdB 37.Ei:b7 1 -0 Karpov - Chandler,
London 1 9B4. The d2-pawn and e l -bishop
never again moved, and Karpov was able to
catch his opponent's king.
9 ... cxd4 10.�xd4 h6 1 l .J.e3 �e8
8
a) In their first encounter Karpov elected
to exchange the enemy bishop: 1 5 .tLlxe7t
Ei:xe7 1 6.Ei:ad l WeB 1 7.h3 ih5 I B .ixd5
ig6 1 9 .Wc l tLlxd5 20.Ei:xd5 tLl c4 2 1 .id4
Ei:ec7 Karpov - Kasparov, Moscow (7) 1 9B4.
Black has some activity and was able to
regain his pawn, but in return White took
over the initiative and was able to convert his
advantage.
b) Two games later Karpov deviated with
1 5 .id4 ic5 1 6.ixc5 Ei:xc5 1 7. tLl e3 ie6
I B .Ei:ad l WcB 1 9 .Wa4 Ei:dB 20.Ei:d3 Karpov
- Kasparov, Moscow (9) 1 9B4. Karpov got a
small but persistent edge and went on to win a
famous knight versus bishop endgame involving
the brilliant pawn sacrifice 47.tLlg2!!.
7
6
5
4
3
12 J.f8
2
.•.
a
b
c
d
f
e
h
g
12.Ei:cl
Karpov chooses the most common move.
Against Kasparov he twice played: 1 2.Wb3
tLl a5 1 3 .Wc2 ig4 1 4. tLl f5 Ei:cB (In 1 990
Illescas deviated with 14 . . .ib4 against
no less than Kasparov himself, but was
unsuccessful. )
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
13.a3!?
According to the database this move had
not been played before 1 993. Illescas was an
experienced Tarrasch player, so Karpov was
probably keen to steer the game away from the
most heavily analysed variations.
a
b
c
d
e
f
g
h
From this position Karpov was successful
with two different moves:
Usually White exchanges on c6 without delay.
Illescas faced this move a few times, including
a couple of particularly noteworthy encounters
against world-class opposition: 1 3 .tLlxc6
1 99 3
Anatoly Karpov - Miguel Illescas Cordoba
bxc6 1 4 . ttJ a4 id7 1 5 .ic5 ixc5 1 6.ttJxc5
.tg4 l S .h3 if5 From this position Kramnik
tried 1 9.Wa4 in Pamplona 1 992 and Illescas
held a draw, but in Linares 1 994 Kasparov
improved with 1 9 .Wd4! and defeated the
Spanish GM.
13 ...ig4 14.h3
Karpov elects to ease the pressure on the
e2-pawn, even though the h3-pawn might
become a target later.
245
Black's best option may well be 1 6 . . . ttJh7!?
which scores well, and passed a high-level test
when Grischuk used it to draw with Gelfand
at Biel in 200 1 .
17.'Wd3 tiJh7
Playing on the b-file deserved consideration:
17 . . . �bS l S .�c2 a5 ( 1 s . . .id6 1 9.�d1 We7
20.ixf6 Wxf6 2 1 .e4 Black is under pressure
in the centre.) 1 9 .�d1 ie6 20.e3 (20 . ttJa4
ttJ e4) 20 . . . �b3 It is not easy for White to make
progress here.
14 ...ie6
18.i.e3
After l S . ttJa4 ttJ g5 1 9 .<j;>h2 ttJ e4 Black has
counterplay according to Karpov.
18 ...i.d6?!
Another bishop move, and again it is on
the slow side. There were several alternatives
available.
a
b
c
d
e
f
g
h
15.ttJxc6 bxc6 16.i.d4
Karpov avoids playing ttJ a4 for a while, as
the black knight might then gain access to the
e4-square.
16 ...i.d7?
Too passive.
16 . . . c5 has been played a few times, but Black
just falls short of equality: 1 7.ixf6 Wxf6
18 .tiJxd5 Wxb2 1 9.Wa4 �adS 20.ttJf4 c4
2 1 .tiJxe6 �xe6 22.�c2 Wxa3 23.Wxc4 Despite
the level material and opposite-coloured
bishops, White has some advantage as his
bishop is stronger. Out of two games in the
database, White has scored one win and one
draw.
I S . . . ttJ g5 1 9 .h4! (Karpov's suggestion of
1 9 .ixg5 is not convincing as Black has a sweet
tactical finesse: 1 9 . . . hxg5 20.e4 [20.ttJa4 Wa5]
20 . . . dxe4 2 1 .ttJxe4 ixh3! This cute move
enables Black to equalize.) 1 9 . . . ttJe6 20.b4 a5
(20 . . . �cS 2 1 .ttJ a4) 2 1 .b5 ixa3 22.�c2 �cS
23.bxc6 ixc6 24.ttJxd5 White has a small but
long-lasting advantage.
l S . . . ttJ f6 looks odd, but seems playable:
1 9.�fd 1 �bS 20 .id4 a5 2 1 .�c2 (2 1 . ttJ a4
ttJ e4) 2 1 . . .ie6 Black should be okay.
Just as on the previous move, Black could
have gone for counterplay on the b-file:
l S . . . a5 !? 1 9. ttJa4 �bS 20.�c2 ttJ g5 The position
is roughly balanced.
19.�fdl
Karpov brings his last piece into action, and
White's pressure against the hanging pawns
increases.
246
The Prime Years
19 ...i.e6
Black could try to improve the knight with
1 9 . . .ltJ f8!?, with the point that after 20.ltJxd5
cxd5 2 1 .Wfxd5 �e5 22.Wfxa8 Wfxa8 23.�xa8
E1:xa8 White loses one of his extra pawns, so
Black is very much in the game.
However White has a significant improve­
ment in 20.b4!, after which he maintains a
clear advantage.
If 22.lLJxb5 �e5 Black survives for the time
being.
22.Wfxb5!? is quite promising, and after
22 . . . �xa3 23.Wfxd7 �xd7 24.E1:b l White
should be able to pick up both the a- and
d-pawns to reach a position with four pawns
versus three on the kingside, which offers
decent winning chances.
22 . . . �f5
8
7
6
5
4
3
2
1
a
a
b
c
d
e
f
g
h
20.b4!
This move not only solidifies White's control
over the c5-square, but also introduces the
possibility of b5 to undermine the d5-pawn.
20 tiJf8?!
•.•
Black improves his knight, but it takes too
much time.
2o . . . Wfd7
By attacking h3, Black tries to gain time to
organize his position.
2 1 .b5!
2 1 .'it>h2 tiJ f6 22.tiJa4 �f5 23 .Wfc3 E1:ac8
Black manages to get himself together.
2 1 .g4 takes away the f5-square from the
bishop, but gives an obvious target. 2 1 . . .h5!
22.b5 E1:ac8 23.bxc6 E1:xc6 24.�xd5 �xd5
2 5 .tLl xd5 hxg4 26.E1:xc6 Wfxc6 27.hxg4 Wfa4
Black should be okay.
2 1 . . .cxb5 22.tLl xd5
b
c
d
e
f
g
h
23 . lLJ b6
23 . lLJ e7t!? E1:xe7 24.Wfxd6 Wfxd6 2Hlxd6
�e4 26.E1:a6 lLJ f6 In this endgame White is
not a pawn up, but he arguably has better
winning chances as he managed to save his
a-pawn.
23 . . . axb6 24.Wfxd6 Wfxd6 25 . E1:xd6 E1:xa3
26.E1:xb6 �d7
White is likely to win the b-pawn and will
reach an endgame with four pawns versus
three on the kingside, which should offer him
decent winning chances with this many pieces
remaining on the board.
20 . . . a5!
Aggressively reducing the queenside pawns
is Black's best drawing attempt.
2 1 .b5! �xa3 22.bxc6! ?
After 22.E1:b l cxb5 23.E1:xb5 a 4 24.lLJxd5 if8
White's advantage is rather small.
22 . . . �xc 1 23.E1:xc 1
Karpov evaluates this exchange sacrifice as
clearly better for White. His compensation
1993
is certainly impressive, but Black has chances
to survive.
23 . . . liJf6 24. liJ a4
After 24 . .if4 We7 2 5 . liJ b 5 Ei:ec8 Black seems
to hold the position.
24 . . . Wc7 25 . .if4 Wa7
Black's
posmon
remains
unpleasant,
nevertheless if ! had to choose between White's
winning chances and Black's drawing chances,
I would put my money on the latter.
21.liJa4!?
2 1 .b5!? was also promlslOg: 2 1 . . . .ixa3
(After 2 1 . . . cxb 5 ? 22.liJxd5 White destroys
his opponent in the centre.) 22.Ei:b 1 cxb5
23.liJxd5 b4 24.liJxb4 Wxd3 2 5 . liJ xd3 White
wins the a7 -pawn and gets an unusual type
of semi-endgame, with four pawns against
three on the kingside, and a lot of pieces on
the board. Theoretically it may or may not be
winning, but in practice White would have
good chances to convert his advantage.
In the game Karpov decides to occupy the
c5-square. I noticed another game from the
same year in which he won by using the same
motif.
8
7
6
5
4
3
247
Anatoly Karpov - Miguel Illescas Cordoba
33 . . . .ixc5t 34.liJxc5 a5?!
Black hopes to get some play on the b-file,
but his position is not strong enough to
justifY such an ambitious approach.
The best chance was 34 . . . Ei:e8, tying White's
hands by forcing him to keep an eye on the
e5-pawn. 3 5 . liJ b3 The knight is headed for
d4. 3 5 . . . .id7 36.f4 (36.Ei:e3!?) 36 . . . Ei:g4 37.g3
h5 3 8 . liJ d4 h4 39.mg2 Black faces a tough
defence with little hope of counterplay.
35.bxa5 !
With Black's c- and d-pawns firmly block­
aded, there is no reason for White not to
accept the gift. His a-pawn soon becomes
extremely strong.
35 . . . Wa7 36.Ei:e3 Ei:b8
36 ... Ei:d8 37.a6 (Also after 37.Ei:e2 d4 38 .Wb4
d3 39.Ei:b2 the a-pawn is too strong.) 37 . . . d4
38 .Wa5 Ei:d5 (38 . . . We7 39 .Ei:e2 d3 40.Ei:d2
Black's passed pawn is securely blocked, but
White's is a different story. . . ) 39 .Ei:b3 d3
40.Ei:b7 d2 4 1 .Ei:d l Wxc5t 42.Wxc5 Ei:xc5
43.Ei:xd2 White wins as his a-pawn will
promote.
37.a4 Ei:d8 38 .Wd4
38.a6 is also good enough.
38 . . . Ei:b8
After 38 . . . Ei:a8 39.liJb3 We7 40.Ei:ec3 .id7
4 1 .f4 Black's position is cut in two.
Black can eliminate the front a-pawn with
38 . . . Wxa5 , but after 39.liJb7 c5 40.liJxc5
(but not 40.Ei:xc5? Wb6) Black is a pawn
down for no compensation.
V'='�'",O=
2
a
b
c
d
e
f
g
h
33.1c5 !
Earlier in the game Karpov left his knight
on c5, inviting an opposite-coloured bishop
middlegame. Black was not tempted, so
now Karpov goes for a different minor piece
exchange.
a
b
c
d
e
f
g
h
248
The Prime Years
39.tLlb3
39.:gb3 was also convincing.
39 . . . '.We7 40.f4 '.Wa3
40 . . . :gb4 is met by 4 1 .'.WcS .
4 1 .:ga l '.We7 42.a6
Black has little chance of surviving against
a pawn so close to the promotion square,
provided the pawn is securely defended,
which will be the case as soon as the knight
returns to cS .
42 . . . hS 43.tLlcS '.Wa7 44.h3 i.e4 4S .:ga2 :gb l t
46.Wh2 h4 47.:gd2
47.:gxe4 dxe4 4B.fS :gh6 49.:gd2 :gbB S O.'.Wc3
also wins.
47 . . . i.fS 4B.:gee2
White is spoilt for choice on every turn.
Here 4B.e6 fxe6 49.tLlxe6 '.Wxa6 SO.tLlgS was
good enough.
4B . . . :gg3
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
21 ...l'iJ g6?
Illescas probably wanted to put the knight
on eS, but he never gets a good chance to do
it.
2 1 . . .tLl d7 was a better try. From here the knight
can also head for eS, but also helps to defend
the cS-square in the meantime. 22.'.Wa6 (Also
after 22.i.d4 :gcB Black is worse, but for the
time being he is surviving.) 22 . . . tLleS 23.t2k5
'.WcB 24.'.WxcB :gexcB 2 S . f4 i.xcS 26.i.xcS tLlc4
Black has chances to hold.
22.WI'c3 .id7
a
b
c
d
e
f
g
h
49.e6!
After focusing on the queenside for most
of the game, Karpov seals the victory by
exploiting Black's vulnerable king.
49 . . . fxe6 SO.tLlxe6 '.Wxa6
The a-pawn perishes, but it served its purpose
and Karpov has no further use for it.
S I .tLlgS! '.WcB S2.:ge7 '.WfB S3.:gde2 :gbB S4.:gf7
'.Wxf7 S S .tLlxf7
1-0 Karpov - Ouer, Austrian Team
Championship 1 993.
Back to the game.
a
b
c
d
e
f
g
h
/f".\
23."Z..l
c5!
Karpov threatens to eliminate a key defensive
piece.
1 99 3
Anatoly Karpov - Miguel Illescas Cordoba
249
23 ...ixc5
26.ifl
Black does not have many moves that avoid
losing a pawn.
White wins an exchange, and the outcome
is decided.
23 . . . ie5 was playable but still unpleasant for
Black: 24.id4 ixd4 25.:8xd4 tD e7 (25 . . . :8xe2
26.tDxd7 'lMfxd7 27.'lMfxc6 is almost hopeless for
Black.) 26.e4 (26.b5 :8c8) 26 . . . 'lMfc7 27.:8cd 1
Black i s under huge pressure i n the centre.
26 ... :8xe3 27.'lWxe3 'lWxe3 28.fxe3
Apart from his material advantage, White
retains his pressure against Black's weak
queenside pawns. The rest is easy.
28 :8e8 29.@f2 tlJe5 30.:8c5 :8e7 3 1.e4!
..•
24.ixc5
White dominates the dark squares, and can
increase the pressure with moves like id4 and
perhaps e4 after suitable preparations.
24... :8xe2?!
Black knows he has been outplayed
positionally so he sacrifices an exchange, but
instead of changing the course of the game he
only accelerates his demise.
25.ie3 'lWe7
25 . . . 'lMff6!? saves the rook but leads to a
lost endgame for Black: 26.'lMfxf6 (Also after
26.�d3 :8a2 27.id4 'lMff5 28.'lMfxf5 ixf5 29.b5
l"lxa3 30.bxc6 White has an overwhelming
advantage.) 26 . . . gxf6 27.b5! tDe5 28.bxc6 ixc6
29.ixh6 With two bishops, a better structure
and several weakness to aim for, White should
win this endgame without too much trouble.
a
b
c
d
e
f
g
h
Karpov breaks up Black's pawns in the
centre.
31...ie6 32.b5 cxb5 33.exd5 id7 34.d6
:8e6 35.ixb5 :8f6t 36.@g2
1-0
This was a good win, yet it was only Karpov's
second of the tournament, and his final score
of 5 Y2/9 was only good enough to share third
place with Topalov and Leko, behind the winner
Yudasin and the runner-up Vyzmanavin.
Perhaps Karpov's busy tournament schedule
had taken its toll on his energy levels.
***
FIDE World Championship match
versus Timman
In 1 993, after various disputes and failed
negotiations, Kasparov and Short broke away
from FID E to contest their world championship
match under the auspices of the newly-formed
PCA. Ever since that time, there has never
been a FIDE World Championship in which
every top contender has participated. Kasparov
never took part in the various FIDE knockout
events. When Kramnik and Leko contested
their title match, neither Kasparov nor Anand
took part. Kramnik was not present in San
Luis, nor was Topalov in Mexico Ciry. And
even in the present day, as the book is about
to go to print, Magnus Carlsen has recently
withdrawn from the next championship cycle.
250
The Prime Years
Perhaps in an effort to re-establish its authority,
FIDE hastily organized a world title match
between the two players who went the furthest
in this and the previous championship cycle.
Karpov was the obvious favourite, but Timman
had raised his level over the previous few years.
The first twelve games were to take place in
three Dutch cities, and the second twelve were
initially slated for the Arab country of Oman
in southwest Asia, although as we will soon see,
the second part did not go according to plan.
Karpov had the black pieces in the first game.
Timman chose a sharp line against the 4 . . . lLl d7
Caro-Kann and later sacrificed a piece, but it
was not fully sound, and despite some minor
inaccuracies Karpov eventually converted
his advantage in the endgame. Despite his
disappointing start, Timman showed his
class in Game 2. Karpov surprised him with
a novelty in one of the main lines of the
4.g3 �a6 Queen's Indian, but it contained a
flaw which Timman exploited with expert
precision. He skilfully converted his advantage
in the endgame to level the match.
In Game 3 Timman improved on his previous
play against the Caro-Kann and obtained
some advantage, but he was not quite able
to convert it and Karpov gradually equalized.
In Game 4 the Dutchman switched to the
Vienna Variation. Karpov reacted timidly and
after twenty moves he stood marginally worse,
but Timman was content to take a draw. In
Game 5 Timman switched to the English
Opening and unveiled a novelty. Karpov
reacted badly and got into serious trouble,
but Timman misplayed his winning position
and in the end had to fight for the draw.
Up to this point Timman had demonstrated
superior preparation, but in Game 6 Karpov
showed that he had been hard at work on the
Vienna Variation as he unleashed a much more
ambitious counter than before. Timman soon
got into trouble and Karpov finished him with
a crushing kingside attack.
By this time it had transpired that the Dutch
organizers had pledged a much smaller sum
than had previously been announced. Worse
still, the Oman organizers completely withdrew
their offer to host and sponsor the second
half of the match. Nevertheless the players
continued, while FIDE officials frantically set
about making alternative arrangements for the
final twelve games . . .
I n Game 7 Timman came u p with another
new idea against the Caro-Kann, but this one
was toothless and a draw was soon agreed.
Before Game 8 Timman and his team carried
out some repair work on the Vienna, and this
time he drew comfortably.
In Game 9 Timman switched to l .d4 for
the first time, and after 1 . . . lLl f6 2.c4 e6 he
surprised Karpov with 3 .�g5 . There followed
3 . . . h6 4.�xf6 Wfxf6 5 .e4 d6 after which Karpov
castled long and Timman short. Part of the
game can be found on page 286 in the notes
to Game 37. For a while Karpov had the better
of the complex middlegame, but he played a
couple of bad moves and got into a terrible
position. But then it was Timman's turn to err,
and the game finally ended in a draw.
Life got worse for Timman in Game 1 0. He
equalized in an exchange Griinfeld, then won
a pawn although Karpov had a strong passed
pawn which gave him sufficient compensation.
Perhaps overestimating his position, Timman
went wrong and Karpov punished him.
Game 1 1 was agreed drawn in eleven moves.
It was rumoured to be in protest at the way
the match had been organized, although if
this was the case, it would seem a little unfair
on Timman to 'sacrifice' one of his white
games. Game 12 was a Fianchetto Griinfeld.
Timman came close to equalizing, but after an
inaccuracy he had to work hard for the draw,
nevertheless he eventually achieved it. Thus
Karpov led 7-5 at the halfway point.
1993
Following Game 1 2, no-one knew if the
match would even be finished. But a week
and a half later, it was announced that the
remaining twelve games would take place
in the Indonesian capital Jakarta. In Game
1 3 Timman tried a Reti but did not achieve
much and after some simplifications a draw
was agreed. In the next game Karpov tried a
different approach against the Vienna, but
soon got a worse position. But later Timman
made some mistakes and when the dust had
settled Karpov had a superior endgame, which
he eventually converted.
With the match situation becoming desper­
ate, Timman tried an ambitious treatment
of the 4.Wfc2 Nimzo-Indian, but played too
riskily. His king was caught in the centre and
he suffered a crushing defeat. In the next game
things went from bad to worse: Timman tried
the King's Indian, but Karpov quickly got a big
advantage and later won, despite a few minor
inaccuracies later in the game.
In Game 1 7 Timman switched to the 3.e5
if5 4.tt:lc3 Caro-Kann and the game naturally
became sharp. Karpov was somewhat worse,
but managed to hold it. Game 1 8 was another
Samisch King's Indian. Karpov went for safety
and the game was always roughly equal until
a draw was agreed before the time control. In
Game 19 Timman tried the Petrosian variation
against the Queen's Indian. He got a slight
plus, but Karpov defended well and held the
draw.
Karpov seemed to be coasting to victory,
but in Game 20 after an English Opening he
played for a draw too overtly and exchanged the
wrong pieces. Timman took over the initiative
and punished him. In Game 2 1 Timman had
the white pieces but played half-heartedly in
the 4.Wfc2 Nimzo-Indian, and a draw was
agreed on move 1 9.
Thus Karpov secured a match victory by the
score of 1 2Yz-8Y2 . It was a convincing win
on paper, although in all honesty the score-
25 1
line was rather flattering. Timman seemed
better prepared in the openings and missed a
number of wins. Karpov's immensely strong
endgame play is what made the difference in
the end. Overall Karpov's play was not on the
level of a world champion. He may have won
the FIDE title, but everyone knew Kasparov
was the stronger player. FIDE probably would
had done better to organize a round-robin
tournament along the lines of San Luis 200 5 .
***
Karpov's next event was the Tilburg knockout
tournament. He was seeded directly to the
second round, where he faced his former
compatriot Oleg Romanishin. In the first game
Karpov scored a fine victory which is referenced
on page 1 77 in the notes to Game 26. In the
return game Karpov was under serious pressure
and eventually found himself a pawn down in
a queen and knight endgame. Romanishin has
beaten all the great players he ever faced, with
the exception of Karpov. Perhaps this affected
him psychologically, and Karpov was able to
save the game thanks to his stubborn defence.
When I interviewed Romanishin he told me
that nowadays people do not understand j ust
how great a player Karpov really is/was.
In the next round Karpov faced Vyzmanavin
and started with the white pieces. He got
no advantage in the l .c4 e5 English, and
Vyzmanavin's fluent play resulted in a draw
right after the time control. The second game
was a quick draw. Vyzmanavin was considered
one of the best blitz players in the world at the
time, but Karpov managed to beat him in both
rapid games to advance to the next round.
Karpov's next opponent was Gregory
Kaidanov, who played solidly and drew both
of the main games, but Karpov eliminated him
in the rapid tie-breaker.
In the next round Karpov faced Artur
Yusupov, who played the main line of the
252
The Prime Years
4.e3 Nimzo-Indian in the first game but was
unable to achieve anything and Karpov drew
comfortably. In the second game Karpov was
at his best.
Yusupov played all the world champions
from Smyslov to Anand, with the exception
of Fischer. He scored nine victories, fifty five
draws and thirty four losses against them. He
found Karpov an especially difficult opponent:
up to this point, Karpov was ahead by eight
wins to one in their personal rivalry, with
twelve draws. Over the years Karpov managed
to increase his dominant record to twelve wins,
fifteen draws and one loss against Yusupov one of his most one-sided records against a
world-class opponent.
I Game 33 I
-
Anatoly Karpov Artur Yusupov
T ilburg
1 993
l .d4 lLlf6 2.c4 e6 3.lLla b6
Yusupov preferred the Queen's Gambit
Declined against Karpov, although he had
used the Queen's Indian against him once
before and made a draw.
4.g3 �b7 5.�g2 �e7 6.lLlc3 llJe4 7.�dl �f6
8.'lWc2
Despite having won four out of four games
with 8 .0-0 and drawn two out of two with the
queen move, he still opts for the latter.
8 ... lLlxdl 9.'lWxdl d6 10.d5!?
Ironically Karpov plays a move which
Yusupov introduced back in 1 980. According
to the database it had been repeated only once
before the present game. In 1 988 Karpov chose
1 0.0-0 and drew against Salov.
10 0-0
•.•
1 0 . . . e5 is possible, and would have prevented
White's next move.
8
7
6
i: _ S ��._
�!L _ _ • •
r
"
/,
" /'::-w.
i£'
'/'
?r
�
�wi£
�
;::
::-.;:: a m
�
: ��!.p;D
%%/' ''//' %%� %%�r�
��,
��/.�"� ',/.��/,.��J��,
�
� r�jLr�
� r�
�"
";/,.
=
""/,
.
:
�
3
2
1
a
b
c
d
e
f
g
h
1 l .lLld4!?
Karpov starts his play in the centre before he
castles. This policy is not only more energetic,
but it also maintains the useful possibility of
castling on either side.
1 l ... e5
Yusupov decides to keep his dark-squared
bishop. This piece certainly possesses great
potential, on the other hand it is restricted
behind Black's dark-squared pawn chain.
1 1 . . . �xd4!? 1 2.Wxd4 e5 looks safer, although
White maintains a slight plus after 1 3 .Wd2 f5
1 4 . f4! as pointed out by Karpov.
12.llJc6 'lWd7
Taking on c6 would open the d5-square for
White's pieces.
13.lLlxb8 :1Uxb8
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
14.h4!
Karpov gains space and prepares to put his
bishop on the h3-cB diagonal.
14 ... a5?
Yusupov acts in the same spirit as Kiril
Georgiev (see Game 9) - he tries to hold back
Karpov by keeping the position closed, but
only succeeds in obtaining a miserably passive
position.
Unusually for such a pawn structure, Black
actually had an opportunity to play actively on
the queens ide: 14 . . . a6! This is the most logical
way for Black to justify the position of his
pieces. 1 5 .�h3 ( 1 5 .a4 b5! Black exploits the
undefended rook on al and gets a good game.)
15 ...�eB 1 6. tt:l e4 �e7 1 7.0-0 b5 ( 1 7 ... g6
1 8.b3 f5 1 9. tt:l g5 is good for White) 1 B .:gfc 1
c 5 1 9. b 3 a 5 Black has a decent position.
15.a4 :gf8
a
b
c
d
e
25 3
Anatoly Karpov - Artur Yusupov
1 993
f
g
1 6 . . . We7!?
This looks critical. Black's idea is to play
. . .f5 .
Another interesting line is: 1 6 . . . WeB 1 7. tt:l e4
�e7 1 B .g4!?
a
b
c
d
e
f
g
h
I B . . . �xh4! ? (Safer is 1 B . . . �cB, intending
a gradual kingside advance, when Black is
only slightly worse.) 1 9 .95 f5 20.gxf6! Wg6!
(20 . . . �xf6? 2 1 .�e6t <j;JhB 22.�f5 Black is in
trouble.) 2 1 .�e6t <j;JhB 22.Wd3 �xf6 White
has promising compensation for the pawn.
1 7.tt:le4 h 5 !
1 7 . . .g 6 I B .h5 �g7 1 9.hxg6 hxg6 20.g4 f6
White has a small advantage.
h
16.e4!?
Karpov prefers to gain space instead of
leaving the e4-square free for his knight.
1 6.ih3
The strategy of playing with pieces is also
interesting, although Black should be able
to obtain a reasonable position if he plays
carefully.
a
b
c
d
e
f
g
h
1 B .g4
This seems like a principled move, but it
may not be the strongest one.
I B .O-O-O g6 1 9 .�d7!? Maybe White should
play less adventurously, but Black should be
okay in any case. 1 9 . . . �a6! (After 1 9 . . .�g7
254
The Prime Years
2o.ic6! White achieves a favourable minor
piece exchange.) 20.b3 Ei:ad8 2 1 .ic6 ic8
Black is in the game, and the bishop on c6
might turn out to be sidelined.
1 8 . . . hxg4 1 9 .ixg4 ia6!
1 9 ixh4 2o .if5 ic8 2 1 .ixc8 Ei:axc8
22.Ei:a3! f5 23.Ei:ah3 fxe4 24.Ei:xh4 is a bit
unpleasant for Black.
20.b3 ixh4
White maintains some compensation, but
with the third rank closed he is unable to
generate such a dangerous attack.
. . .
facilitate the smooth transfer of heavy pieces
between different parts of the board.
20 .. J�h8 21.Ei:h2 i.a6!
This move forces White to keep an eye on
the queenside, and may have contributed to
Karpov's decision to stop playing for b4.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
22.lLlb5!
Karpov avoids playing b3, as there could be
certain scenarios later in the game in which
one or more of his pieces might wish to use
that square to invade on the queens ide.
22 ... Ei:af8
17.0-0-0 g6 18.i.h3 Wfe7 19.@c2
Karpov was
probably contemplating
opening the queens ide with b4 at some point,
otherwise he would have put his king on b 1
immediately.
19 ... @g7
1 9 . . . ig7 is strongly met by 20.g4! hxg4
2 1 .ixg4 f5 22.ih3 intending h 5 , when White
gets lovely play on the kingside.
The opposite-coloured bishop position will
lead to a period of unpleasant passive defence
for Black: 22 . . . ixb5 23.cxb5 c;t>g8 24.c;t>bl
ig7 2 5 .W!e l ih6 26.Ei:c2 White can create
strong pressure by tripling on the c-file.
23.@bl!
Karpov gets ready for the capture o n b5. If
Black is able to organize a defensive formation
with the bishop on d8 and his rooks on the
kingside, it will be hard for White to make
progress.
23 ...Wfd8
20.£3!
Karpov opens the second rank in order to
Yusupov gives up on the plan of exchanging
on b5 , and instead prepares to exchange White's
1993
255
Anatoly Karpov - Artur Yusupov
other strong minor piece, although this brings
other problems as we will soon see.
23 . . .ixb5 runs into problems: 24.cxb5 WeS
25 .1Wc3! (White has to hurry; instead 2 5 .l'k l
idS enables Black to defend conveniently.)
25 ... idS 26.f4! White maintains his initiative
and will soon double his rooks on the f-file.
26.Ei:gl
Preparing to advance the f-pawn.
The immediate 26.f4!? may be even stronger:
26 . . . exf4 27.gxf4 Ei:eS (27 . . ..txh4 2S.e5 is
strong.)
24.�d3
Karpov opens the second rank, so that after
an exchange on b5 his rook will get to c2
quickly.
24...ie7
a
a
b
c
d
e
f
g
h
25.ge2!
b
c
d
e
f
g
h
2S.f5 (2S.Ei:g l transposes to the main
line below - see 26.Ei:gl Ei:eS 27.f4 exf4
2S.gxf4; 2S.tt:Jd4!? is also possible, and after
2S . . . .txd4 29.Wxd4t 'it>h7 White obtains
fine compensation after both 30.e5 Wxh4
3 1 .We3 and 30.b3 Wxh4 3 1 .We3 .) 2s . . . ixb5
29.cxb5 .txh4 30.fxg6 fxg6 3 1 .Ei:gl Ei:h6
32.if5 White has lovely play on the g-file.
26 . . . Ei:eS 27.f4 exf4
Black may as well open the diagonal for his
bishop, as 27 . . . We7 2S.f5 is depressing for
him.
2S.gxf4
With this subtle move, Karpov gets ready
to play f4 and sacrifice the h4-pawn. The
rook takes up a position in the centre where it
supports the e4-e5 push.
25 ...ic8?!
In principle the side with less space should
exchange pieces, but now White's knight
becomes the dominant minor piece.
try to
25 . . . if6
This would have been the lesser evil, although
Black's position remains unpleasant here
too.
a
b
c
d
e
f
g
h
2S . . . .txb5
2S . . . Ei:h6? 29.tt:Jd4 Black is in trouble.
25 6
The Prime Years
2B . . . 'it>h7 29 . .ie6! We7 (29 . . . fxe6 30.e5+-)
30.e5! White has a decisive attack, for
instance 30 . . . .ixh4 3 1 .exd6 cxd6 32 . .ixf7
Wxf7 33.lt:lxd6 and it is all over for Black.
29.axb5
This time White keeps his c-pawn in place,
in order to provide full support to his central
play.
29 . . . We7
29 . . . 1"le7 30.f5 is strong.
30 . .ie6! .ixh4 3 1 .e5
White has a powerful initiative to compensate
for the missing pawn, but Black is still alive.
26.,txc8 Wixc8
on the other hand it controls some important
squares and exerts pressure against the a5- and
e5-pawns.
28 Wih3 29.Wic2
•••
Karpov wants to be able to expel the queen
with 1"lh2 at any moment.
29 J3e8 30 .!tJc6 i.f6 3 1 .1"lfl Wid7 32.Wfe2
1"lhf8 33.1"lh2
••
•
Before pushing his f-pawn, Karpov makes
sure the enemy queen will not be able to
invade.
The immediate 33.f4 was also possible:
33 . . . exf4 (33 . . . Wh3! ? could be considered)
34.gxf4 (34.1"lxf4? .ie5) 34 . . . Wh3 (34 . . . .ixh4?!
3 5 .1"lh2 .if6 36.1"lxh5 1"lh8 37.1"lxh8 1"lxh8
3B.e5 [3B.'it>a2!?] 3B . . . Wf5 t 39.'it>a2 White
has a clear advantage.) 3 5 .� a2 'it>h7 36.e5
Wf5 There is no question that White has some
advantage, but it is not clear how serious it is.
33 1"lh8 34.Wifl 1"lef8
•..
a
b
c
d
e
f
g
h
27J3g1 !?
Karpov decides to prepare more pieces
before advancing his pawns. He may even have
intentionally delayed this action until shortly
before the time control in order to maximize
his opponent's discomfort.
Nevertheless, the immediate pawn advance
was objectively not weaker at all: 27.f4!? exf4
(27 . . ..if6 2B.f5) 2B.gxf4 .ixh4 29.1"lgl White
has excellent compensation for the pawn.
a
b
c
d
e
f
g
h
35.1"lfl !
27 i.d8 28 . .!tJ a7!
After 3 5 .g4 .idB 36.g5 f5 37.gxf6t 1"lxf6
3B .Wg2 'it>h7 39.1"lh3 1"lgB Black can resist
stubbornly.
Karpov transfers the knight to c6. Unlike b5
it will not attack anything from that square,
35 'iflg8 36.£4!
•.•
.••
Finally Karpov executes the long-awaited
pawn break. As in many of his games, he takes
this troublesome action shortly before the time
control. The fact that he kept his king on b 1
shows that he anticipated a queen exchange,
otherwise he would have surely have moved
the monarch to a2 first.
36.. �g4
Waiting passively was virtually hopeless:
36 . . . Wh7 37.f5 ig7 3S .Wff3 ih6 39.l3hf2
E1hg8 40.Wfd3! Preparing to triple the heavy
pieces. 40 . . . WfeS 4 1 .l3f3 Wfd7 42.l3 1 f2 <J?g7
43.'?Nfl WfeS 44. tLl a7! The knight comes to
b5, after which Black will have a hard time
defending all his weaknesses.
.
37.Wff3!
Karpov exchanges queens in order to make
progress.
37.fxe5? ixe5! solves all Black's problems
thanks to the trick 3S.tLlxe5 Wfxe4t!.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
37 .. J:�e8?
Yusupov plays for a fortress, but it can be
cracked. Black could have obtained better
drawing chances by keeping some active
possibilities alive for his pieces. Two alternatives
deserved attention.
37 ... '?Nxf3 3S J::!:xf3 l3eS
257
Anatoly Karpov - Artur Yusupov
1 993
Compared with the game continuation,
Black does not have a weak pawn on g4 and
White is unable to focus all his power on the
queenside. Nevertheless his winning chances
are excellent.
39.l3hf2 l3h7 40.f5
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
40 . . . g5
40 . . . ihS? 4 1 .f6 White traps rwo pieces
and will easily overpower Black on the
queens ide.
40 . . .gxf5 4 1 .l3xf5 ihS 42.l3g5 t (42.Wc2 f6
is ugly, but Black is surviving for the moment
at least.) 42 . . . l3g7 43.l3ff5 l3g6 44. tLl a7!
White will invade one way or another.
4 1 .hxg5 ixg5 42.f6!
42.E1h2 f6 43.E1fl <J?f7 44.E1fh l E1ehS Black
resists.
42 . . . h4 43.E1f5 ie3 44.E1h2! h3
44 . . . hxg3? 4 5 . tLl e7t wins.
45.E1fl ig5 46.Wc2 E1h6 47.tLl e7t WhS
47 . . . <J?fS 4S.tLlf5 does not change much.
4S.E1fh l
White has a large advantage thanks to his
vastly superior minor piece.
However, Black had an alternative which
offered more stubborn resistance:
37 . . . exf4!
Opening the diagonal for the bishop and
setting up some additional chances for
White to go wrong.
3S .Wfxg4
The Prime Years
258
This is probably what Yusupov feared. White
targets the g4-pawn, but he will have to
be careful as his rook might end up in an
awkward position.
It is entirely possible that White's most
promising course of action involved a simple
central advance rather than trying to win
a pawn. The following continuation offers
him decent winning chances: 38.�xf4
�xf4 39.gxf4 (39.Ei:xf4 i.g7 40.Ei:hf2 f6 is
also better for White, but the position will
not be easy to win.) 39 . . . Ei:e8 40.Ei:e l �g7
4 1 .e5 �h6 42.�c2! The king will soon
defend the rook, and Black will find it hard
to live with his opponent's central pawn
wedge.
38 . . . hxg4 39.Ei:xf4
47.Ei:f5 White is in full control, and has good
chances to win the game.
40.Ei:hf2 i.e5 4 1 .Ei:xg4 � h6
Also after 4 1 . . .Ei:h6 42. �c2 Ei:e8 43.'Jid3 f6
White has problems freeing his rook.
42. ttJ e7 �h5 43.Ei:g5t �h6
I do not see an effective way for White to
free his rook. Therefore it looks better for him
to deviate with 3 8 .�xf4 as analysed above.
38JWxg4 hxg4 39.5
If 39.Ei:hf2 Ei:h7 Black can resist.
a
a
b
c
d
b
c
d
e
f
g
h
39 .. g5?
.
e
f
g
h
39 . . . �g7!
This subtle move practically forces White to
take the g4-pawn, after which the rook will
be out of play.
39 . . . i.e5 allows White to switch to a different
plan: 40.ttJxe5! (40.Ei:xg4 �g7 4 1 .�c2 Ei:h6
42 .�d3 f6 43.ttJd4 � f7 44. ttJ e6 Ei:c8 Black
is living dangerously, but I do not see how
White can free the g4-rook or win the game
without it.) 40 . . . dxe5 4 1 .Ei:f6 Ei:d8 42.Ei:c6
Ei:d7 43.c5 The endgame is unpleasant
and possibly losing for Black, for instance:
43 . . . �g7 44.cxb6 cxb6 45 .Ei:f2 (45 .Ei:xb6 f5 !
The outcome is not clear but Black certainly
has drawing chances.) 45 . . . Ei:f8 46.h5! gxh5
Yusupov plays for the fortress too directly.
3 9 . . . �g7!?
This would have given White more chances
to go wrong.
40.�c2!
In order to win the game White must look
towards the queenside.
40.Ei:hf2? Karpov gives this move and
evaluates the position as clearly better for
White, but in fact it is a serious mistake in
view of 40 . . . g5! 4 1 .hxg5 i.xg5 42.f6t mf8!
when Black solves his problems.
40 . . . Ei:ef8
40 . . . Ei:h5 4 1 . fxg6 fxg6 42.Ei:hf2 Ei:f8 43.c5!
(Best, although 43.ttJ b8 is also good enough.)
43 . . . dxc5 44.�d3 White's king will walk to
b7 with decisive effect.
1993
a
b
c
d
e
f
g
a
h
41 .c5! dxc5
4 l . . .bxc5 42.l'!hf2! This move works with
the black rook on f8 instead of e8, as 42 . . . g5
loses to 43.hxg5 ixg5 44.f6t with a mating
attack.
42J!hf2 l'!h6
42 . . . g5 43.hxg5 ixg5 44.f6t! wins again.
43.fxg6 l'!xg6 44.l'!f5 l'!e8 45 .md3 l'!h6 46.mc4
E1g6 47.\t>b5
The king invasion decides the game.
a
b
c
d
e
259
Anatoly Karpov - Artur Yus upov
f
g
h
40.h5!
Now White is clearly winning as he can open
the queens ide and crack his opponent's fortress.
Let me show you another interesting game in
which Karpov managed to break a stronger
fortress than the one Yusupov has constructed
in the present game. His opponent was Jon
Speelman, who closed the position almost
hermetically.
b
c
d
e
f
g
h
27.l'!b l
It is not easy to find a way into Black's
position as there are no open files. Karpov
threatens to open the b-file in the future.
27 . . . l'!b5 28.l'!cc l
Karpov transfers this rook t o the kingside.
28 . . . l'!h8 29.l'!n l'!h7 30.l'!f3 tLl b6 3 1 .l'!h3
The rook comes to the semi-open file and
threatens to win an exchange.
3 1 . . .mg8
Speelman defends the rook.
With 3 1 . . .a3!? Black prepares to sacrifice an
exchange. The problem is that White can
target the daring a-pawn: 32.id l ! (Instead
after 32.if2 ixf2 33.l'!xh7 ie3 34.l'!n tLl a4
35 .l'!h3 id2 36.l'!ff3 tLl b2 it will be hard
for White to win the position, and he will
probably have to give up a pawn or two
just to create chances.) 32 . . . mg8 33 .ic l
mf7 34.ixa3 White's extra pawn should be
enough to decide the game.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
260
The Prime Years
32.id l !
With this move Karpov paralyses both the
knight and the rook.
32 . . . �h8 33.<Jifl
The king threatens to march to a3 to win
the a-pawn.
33 . . . �h7 34.<;t>e2 �h8 3 5 .ic2 �h7 36.�fl
�h8 37.<Jidl �h7
44 . . . �a6 are good enough for Black to hold
the position.
8
7
6
5
4
3
8
2
7
6
a
5
4
3
" =��-=�,=, �='�,,, ,,,F=�
2
a
b
c
d
e
f
g
h
38.<Jic l ! ?
Karpov continues approaching the a4-pawn
and sacrifices a pawn in the process.
3 8 .�ff3 was by no means clear: 38 . . . �h5
39.<;t>c l a3 (39 . . .ie l ? 40.�xh5 gxh5 4 1 .ig l
Black's bishop will soon perish; 39 . . . <Jih7
40.<Jib2 <Jig8 4 1 .id2 <Jih7 42.<Jia3 <Jig8
43.ixa4 ltJ xa4 44.<;t>xa4 �b7 4 5. <;t>a3 �a7t
46.<;t>b2 <Jif7 Black is still resisting despite
the pawn deficit.) 40.<Jidl ltJ a8 4 1 .ic l �b6
42.ib 1 �a6 Black has managed to organize
his defences.
38 . . . if2!
Speelman wins the pawn by the help of a
small tactic.
39.�f3 ixe3t 40.�xe3 �xh2
Taking the pawn should be enough to draw
the game, but it was not the only way.
40 . . . a3! ? Closing the queens ide was also
good enough: 4 1 .�f2 (4 1 .�ff3 �xh2
42.�h3 �xh3 43.�xh3 ltJ a8 Black holds.)
4 1 . . . ltJ a8 42.<;t>d2 ltJ c7 43 .ia4 �b6
(43 . . . �h5 !? 44.ixb5 ltJ xb5 is a reliable
fortress.) 44.<;t>e2 Now both 44 . . . ltJ b 5 and
b
c
d
e
f
g
h
4 1 .a3!
Karpov fixes the a4-pawn.
4 1 . . . �h4?!
With this and his next move Speelman just
waits, but it soon becomes clear that his
fortress is not watertight at this stage.
Grandmaster Groszpeter found the most
effective defensive formation for Black:
4 1 . . .<;t>f7! 42.�ff3 �h8 43.�h3 �a8 44.�h7
ltJ d7 4 5 . �eh3 �bb8 White cannot invade.
42.<;t>b2 �h2 43.�ef3
8
7
6
5
4
3
"=j'��""O=,/'="
2
a
b
c
d
e
f
g
h
43 . . . �xc2t!
The English grandmaster sacrifices an
exchange in order to eliminate the pressure
on his a4-pawn, thus liberating his two
queens ide pieces. White only has one open
file, so it was not unrealistic to put his faith
in this plan.
1 993
Passive defence was not good enough, as
Black must either allow a rook invasion
or lose his a-pawn: 43 . . . E!:h4 44.E!: l f2 E!:h l
45.E!:g3 E!:h4 46.E!:gg2 'iit f7 (46 . . . E!:h5 47.E!:h2
It>h7 48 . .id l Black has to take the rook.)
47.E!:h2 E!:xh2 48.E!:xh2 'iit g8 49.E!:h4 Black is
in zugzwang and must give up the a4-pawn.
Please remember that zugzwang is often the
key to unlocking a fortress.
44.lt>xc2 ttJ d7!
Speelman plans to block White's rooks on
the eighth rank.
45J!h3 ttJ f8 46.E!:fh l � f7
8
7
6
5
4
3
',=j' ��" ,,;;;;, /'="
2
a
b
c
d
26 1
Anatoly Karpov - Artur Yusupov
Black repeats his previous mistake, and this
time Karpov punishes him.
5 1 . . . 'iit e 7! With this superior move Black
challenges his opponent to come up with
a convincing winning plan. 52.�d2!? The
subtle plan involving the king on c2 was
suggested by Ashot Nadanian. (52.'iit e2
�f7 5 3 . � e3 This looks like the natural
square for the king, but matters are not so
simple. 53 . . . � e7 54.E!:bh l E!:b8! [54 . . . �f7?
5 5 .E!:xf8t �xf8 56.E!:h8t 'iit e 7 5 7.E!:g8 'iit f7
58.E!:c8+-] 5 5 .E!:g8 [ 5 5 .b5 E!:xb 5 !] 5 5 . . . � f7
56.Elhh8 E!: e 8 I d o not see how White can
crack Black's fortress.) 52 . . . �f7 53.�c2
'iit e7 54.E!:bh l E!:b8 5 5 .b5 Elxb5 56.E!:xf8!
'iit xf8 57.E!:h8t 'iit e 7 5 8.E!:a8 c5 5 9. dxc5 E!:xc5
60.Elxa4 According to Nadanian White
has reasonable winning chances. His ideas
include placing his rook on a6 to tie Black's
king to the defence of the e6-pawn, followed
by walking his own king to d4.
8
e
f
g
h
47.Elh8
Now the knight must remain on f8 to
prevent a rook invasion.
47 ... Elb8 48.E!:b 1
48.El l h7 E!:e8 gets nowhere for White.
48 . . . Elb5 49.'iit d 2 Elb8?
Black should have left his rook in place and
preferred 49 . . . �e7!, when the play resembles
the note to Black's 5 1 st move below.
50.lt>c2?
Karpov misses a clear win with 50.b5! cxb5
5 1 .Elb4 �e7 52.E!:h 1 ttJ d7 53. 'iit e 3 'iit f7
54.�f3 �e7 5 5 .E!:hb l Elh8 56.'iit g3, as
pointed out by Groszpeter. This was a rapid
game, so it is likely that both players were
short of time by this point.
50 . . . Elb5 5 1 . 'iit d l !
Karpov opens the second rank.
5 l . ..Elb8?
7
6
5
4
3
L=j'��",,;;;;J' = "
2
a
b
c
d
e
f
g
h
52.b5!
White opens the b-file and his rooks invade.
52 . . . cxb5 5 3.E!:h2! ttJ d7 54.E!:hb2
The rest is simple.
54 . . . E!:h8 5 5 .E!:xb5 E!:h l t 56.�e2 E!:h2t 57.'iit f1
Elh3 5 8 .E!:b7 'iit e7 5 9 . El 1 b4 E!:xc3 60.E!:xa4 E!:b3
6 1 .E!:xb3 cxb3 62.E!:b4 �d8 63.E!:xb3 �c7
64.a4 ttJ b8 65.�e2 ttJ c6 66.'iit d3 ttJ a5 67.Elb5
ttJ c4 68.a5 'iit c6 69.E!:c5t 'iit b 7 70.E!:xc4 dxc4t
7 1 .'iit xc4 1 -0 Karpov - Speelman, Roquebrune
(rapid) 1 992.
262
The Prime Years
Back to the game.
a
a
b
c
d
e
f
g
h
40 1"la8!
..•
Forced, otherwise White manoeuvres his
knight via a7 all the way back to the kingside
and gobbles the g4-pawn. Unfortunately for
the defender, White has other ways to improve
his position.
b
c
d
e
f
g
h
50.f6 (Even the flamboyant 50.!'lg6!? works:
50 . . . fXg6 [50 . . .f6 5 1 .!'lxg5 !]
5 1 .tLle7 +-)
50 . . . ixh6 5 1 .!'lxh6 !'le8 52.tLle7 !'lxe7 (If
52 . . . !'lbb8 5 3 . tLl f5 White collects the g4-pawn
with tLl e3xg4, then invades with his king.)
5 3 . fXe7 !'lb8
4l.'i� c2 'iif g7 42.'iif c3 1"la6 43.1"lal 1"lha8
44.b4 'iif f8
If 44 . . . axb4t 4 5 .'iif b 3 <;t>h7 46.tLlxb4 !'la5
47.tLlc6 !'lc5 48.a5 bxa5 49.!'lha2 White
invades and wins.
45.h5!?
Karpov must have been feeling particularly
sadistic. Opening the b-file would have won
more quickly: 45.bxa5 bxa5 46.!'lb l !'lc8
(46 . . . 'it>e8 47.h6) 47.!'lb5 Black can resign.
a
b
c
d
e
f
g
h
54.c5! This lovely breakthrough seals White's
victory.
45 1"l6a7 46.h6!
.•.
Black will not be given time to seal the
kingside completely.
46 .ih8
.•.
46 . . . 'it>g8 does not help: 47.h7t <;t>h8
48.!'lah l !'lb7 (48 . . . ig7 49.f6 ixf6 50.!'ln
wins.) 49.!'lh6 ig7
a
b
c
d
e
f
g
h
1993
Anatoly Karpov - Artur Yusupov
47.f6!
Karpov makes sure he will be able to invade
on the f-file.
47... ixf6 4SJ'Ul ihS 49. ttJxa7 gxa7 50.gh5
rile7 5 l .gxg5 gaS 52.h7 f6 53.ggS gf8
a
b
c
d
e
f
g
h
54.c5!
This cute sacrifice opens the path for a
decisive king invasion.
54 ... dxc5 55. <t!l c4 <t!l f7 56.d6 cxd6 57.gxf8t
1-0
With this fine win Karpov ensured his
passage to the next round, where he met
Beliavsky. In the first game he equalized with
the 4 . . . ttJ d7 Caro-Kann and proceeded to
outplay his opponent. He eventually won a
sublime endgame which can be found on page
297 in the notes to Game 3 8. In the second
game Karpov played solidly in the Orthodox
Queen's Gambit, and Beliavsky was unable to
stir up complications. A draw was agreed on
move 33.
263
In the final Karpov faced Ivanchuk. In the first
game he equalized with the Caro-Kann and
drew a long game. In the next game Ivanchuk
played a Stonewall set-up. Karpov played f4
himself, but never got a serious advantage
and Ivanchuk held a draw without much
difficulty. Karpov managed to win the first
rapid tie-break game with black, and he safely
drew the second one to secure first prize in the
tournament.
Karpov eliminated six opponents in Tilburg.
He played twelve regular games, winning three
and drawing nine. He also had six rapid games,
winning four and drawing two.
1 993 was a successful year for Karpov. He
captured the FIDE World Championship and
won four tournaments, playing some great
chess along the way.
264
1993 Summary
Wijk aan Zee KO ( l st place) : 5/8 (+3 =4 - 1 ) (regular games)
Linares (2nd-3rd place) : 8 Yzl 1 3 (+6 =5 -2)
Austrian Team Championship: 1 Yzl2 (+ 1 = 1 -0)
French Team Championship: 4/5 (+3 =2 -0)
European Cup final: 1 Yzl2 (+ 1 = 1 -0)
Dortmund ( l st place) : 5 Yz17 (+5 = 1 - 1 )
Dos Hermanas ( l st place) : 7Yzl9 (+6 = 3 -0)
Leon (3rd-5th place) : 5 Yzl9 (+2 =7 -0)
FIDE World Championship match versus Timman: Won 1 2Yz-8Yz (+6 = 1 3 -2)
Tilburg ( l st place) : 7Yz/ 1 2 (+3 =9 -0)
Total 67.0% (+36 =46 -6)
II Wins • Draws • Losses
1994
Rating 2740 (2 in the world)
Karpov started the new year with a six-game match against the number one Chilean player, Ivan
Morovic Fernandez. In the first game he played an unambitious set-up and got no advantage
against the Queen's Gambit Declined, but exploited his opponent's subsequent errors to win with
a lovely rook sacrifice. In the second game Karpov firmly held a 4 . . . tLl d7 Caro-Kann. In Game
3 Karpov repeated his harmless opening but managed to outplay his opponent, who resigned
before the time control.
In Game 4 Karpov once again equalized with the Caro-Kann, but this time he played ambitiously
and outplayed his opponent to secure the match victory with two games to play. In the next game
Karpov switched to the Catalan and won after Morovic Fernandez made a serious mistake in the
early middlegame. The final game was a quick draw, so Karpov won the one-sided match by the
score of 5-1 .
Karpov's next event was Linares, which would prove to be a historic one. Karpov's first opponent
at the "Wimbledon of chess" was Lautier, whom he outplayed on the black side of a 1 .c4 e5
English. Karpov had a slice of good fortune in Round 2, as Bareev equalized in a French Tarrasch
and exchanged off his isolani, only to commit an inexplicable blunder, allowing Karpov to capture
a free rook and deliver an instant checkmate at the same time. In Round 3 Karpov equalized in a
Queen's Indian against Illescas, who proceeded to make some mistakes which Karpov punished.
In Round 4 Karpov extended his perfect start, defeating Topalov in spectacular style. The game
offers a convincing reminder that Karpov was much more than a strategic maestro, and could
easily morph into a tactical wizard when the situation on the board demanded it:
l.d4 tLlf6 2.c4 c5 3.tLlf3 cxd4 4.tLlxd4 e6 5 .g3 tLl c6 6.ig2 ic5 7.tLlb3 ie7 8.tLlc3 0-0 9.0-0 d6
1O.if4 tLlh5 1 1 .e3 tLlxf4 1 2.exf4 id7 1 3 .1Wd2 1Wb8 1 4.Ei:fe l g6 1 5 .h4 a6 1 6.h5 b5 1 7.hxg6 hxg6
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
The Prime Years
266
I S .tLlc5! dxc5? 1 9.Wxd7 :gcS 20.:gxe6!! :ga7
2 1 .:gxg6t! fxg6 22.We6t Wg7 23.�xc6 :gdS
24.cxb5 �f6 2 5 . tLl e4 �d4 26.bxa6 Wb6
27.:gd l Wxa6
a
b
c
d
e
f
g
h
2S.:gxd4! :gxd4 29 .Wf6t WgS 30.Wxg6t � fS
3 1 .WeSt �g7 32.We5t WgS 33.tLlf6t Wf7
34.�eSt WfS 3 5 .Wxc5t Wd6 36.Wxa7 Wxf6
37.�h5 :gd2 3S.b3 :gb2 39.Wg2 1 -0 Karpov ­
Topalov, Linares 1 994.
soon obtained a slight plus in the middlegame.
Kasparov gave up a pawn in order to eliminate
the queens ide pawns, and held the resulting
endgame comfortably thanks to his bishop
pair.
In Round S Karpov once again had the blaek
pieces, this time against Gelfand. For a while
he was worse and Gelfand had some attacking
chances, but he failed to make the most of his
position and Karpov punished him to score
another excellent win. In the next two games
Karpov was pressing against both Shirov and
Kamsky, both of whom were able to draw.
In Round 1 1 Karpov faced Vladimir
Kramnik, who by that time had already
broken the 2700-barrier. They had drawn
twice previously. Their lifetime score is two
wins apiece, with nine draws.
I Game J4 1
Anatoly Karpov - Vladimir Kramnik
In Round 5 Karpov faced Ivanchuk. He
equalized with a Queen's Indian which quickly
turned into a Hedgehog-type position. His
position remained comfortable and perhaps
marginally better, when suddenly Ivanchuk
overlooked a simple tactic and blundered a
pawn. Then he gave Karpov a gift by resigning
instantly, even though a great deal of work
was required to convert Black's advantage.
The last victim of Karpov's incredible winning
streak was Judit Polgar. Karpov surprised her
with 2.c3 against the Sicilian, and soon got
a pleasant edge in the French-type position
which arose after 2 . . . e6 3.d4 d5 4.exd5 exd5.
Judit played the opening rather poorly for a
highly ranked grandmaster, and Karpov won
convincingly.
In Round 7 Karpov faced the ultimate test,
having the black pieces against Kasparov,
but here too things went his way. Unusually
for him, Kasparov did not play the opening
especially well; Karpov equalized easily and
Linares
1 994
l .d4 d5 2.c4 c6 3.ltH'3 lLlf6 4.lLlc3 e6 5.e3
tD bd7 6.i.d3 dxc4 7.i.xc4 b5 8.i.d3 a6
Karpov scored better against this variation
than he did against the other main line S . . . �b7.
A year later Kramnik switched to that move
against Karpov and drew in eighteen moves.
9.e4 c5 10.d5 c4
Nowadays many players prefer l O . . . WEe7
here.
1 1 .dxe6
Karpov also played 1 1 .�c2 at once.
1 l ... fxe6
Karpov encountered the 1 1 . . . cxd3 gambit
variation three times. 1 2.exd7t Wxd7 1 3 . 0-0
�b7 1 4.:ge l Karpov faced two different moves:
here:
1994
a
b
c
d
e
267
Anatoly Karpov - Vladimir Kramnik
f
g
a) 14 . . . .tb4 1 5 .tZJe5 Wi'e6 1 6.tZJxd3 .txc3
17.ctJf4 Wi'd7 I B .bxc3 ctJxe4 1 9 .Wi'xdlt c;t>xd7
20.1a3 Ei:heB 2 1 .Ei:ed l t White has a strong
initiative and Karpov later managed to catch
his opponent's king, Karpov - Tal, Bugojno
1 980. It was his only win against the magician
from Riga.
b) Lutz tried 1 4 . . . .te7 twice against Karpov.
1 5 .e5 ctJd5 1 6. tZJ e4! (Improving over their
first encounter: 1 6.Wi'xd3 tZJxc3 1 7.Wi'xc3 0-0
18 .1g5 Ei:acB 1 9.Wi'e3 .txg5 20.tZJxg5 Ei:c2
2l .b3 Wi'g4 22.Wi'g3 Wi'xg3 23.hxg3 h6 24.tZJ e4
1xe4 25.Ei:xe4 Ei:dB Black managed to become
very active and he drew the endgame without
much difficulty, Karpov - Lutz, Bundesliga
1 994.) 1 6 . . . 0-0 1 7.Wi'xd3 Wi'g4 I B . tZJ fg5 Ei:adB
1 9.a3 f5 20.exf6 gxf6 2 1 .Wi'h3 Wi'xh3 22.tZJxh3
Black has insufficient compensation and White
went on to win, Karpov - Lutz, Dortmund
1 994.
12 ..tc2 .th7 13.0-0 Vlfc7 14.tLlg5
Four years later Karpov used 1 4.Wi' e2 to good
effect against Anand: 1 4 . . . .td6 1 5 . tZJ d4 tZJc5
1 6. f4 e5 1 7. tZJ dxb5 ! ? This was an important
novelty at the time. 1 7 . . . axb5 I B .tZJxb5 Wi'b6
19.ctJxd6t Wi'xd6 20.Exe5 Wi'xe5 2 1 .Ei:f5 Wi'e7
22.�xc4 Ei:cB 23.Wi'b5t tZJ cd7 24.Wi'xb7 Ei:xc2
25 .1g5 (25.Ei:a5!?) 25 . . . Wi'd6 26.Wi'aBt c;t>m
27.�xhB Wi'd4t 2B.c;t>h l Wi'xe4 29.Ei:f3 Ei:xg2
30.c;t>xg2 tZJe5
c
b
a
h
d
e
f
h
g
3 1 .Wi'xglt! With this beautiful tactical shot
White simplifies to a winning endgame which
he eventually converted, Karpov - Anand,
Lausanne ( 1 ) 1 99 B .
14... tLlc5
a
b
c
d
e
f
g
h
15.e5
Karpov sacrifices a pawn to open the e-file
towards Black's king. It has been White's most
popular choice in the position.
15 ...Vlfxe5 16J:�el Vlfd6 17.Vlfxd6 hd6
18 ..te3
White develops instead of taking back the
pawn at once. According to the database
Kramnik played this move in the 1 993-94
Bundesliga season, but the exact dates are not
given so it is hard to work out which game took
place first. I would guess that Kramnik adopted
it after his defeat in the present game.
The Prime Years
268
IS ... 0-0
This is Black's most popular move. Kuczynski
played 1 8 . . . tLld3 against Kramnik and drew
after a long battle.
Black is slightly better but he was content to
take a draw here in Kramnik - Shirov, Novgorod
1 994.) 23 . . . if3 24.l"1d2 b4 25.tLla4 ia7 Black
has done well from here; see for instance the
game Nikolic - Shirov, Horgen 1 994.
19.:aadl Jie7 20.fixe5 fixe5 2 1 .tlJxe6 :afeS
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
22.h3!!
This is a great idea - White prepares g4
followed by a general expansion on the
kingside. This move was played in several other
games in 1 994; the database does not give the
precise dates, but it is likely that Karpov played
it first and other players followed his example.
Despite the ingenuity of White's concept,
Black should be okay and over the years he
has scored more than fifty percent from this
position. But of course it is much harder to
deal with an idea when it comes as a surprise,
even for such a gifted player as Kramnik.
22 ...Jif8?!
Nowadays Black has
playing this slow move.
virtually stopped
22 . . . l"1ab8! has become the main line. It has been
championed by Shirov, who used it a few times
in 1 994, including one game against Kramnik
himself. The main line continues: 23.g4
(23 .tLlxc5 l"1xc5 24.l"1e6 b4 25.tLla4 l"1g5 ! 26.g3
l"1h5 27.h4 if3 28.l"1d4 id5 29.l"1e7 if7 30.f4
a
b
c
d
e
f
g
h
23.g4!
Karpov continues his plan. The advancing
pawns, combined with the powerful knight on
e6 and White's control over the d-file, could
easily cause problems for the defender.
23 ... h6?!
This weakens the black kingside. Kramnik has
a strong bishop on b7, and perhaps he thought
White's king would become vulnerable. Black
would do better to avoid moving his h-pawn.
23 . . . if3 24.l"1d4! (24.l"1d2 l"1e8 25 .g5 [25 .l"1d
ic6] 25 . . . tLlh5 26.l"1e3 ic6 It is not easy for
White to increase his pressure.) 24 . . . l"1e8 25 .g5
tLlh5 26.l"1e3 (26.g6 h6 27.id 1 ixd 1 28.l"1dxd l
tLl f6 Black must pay attention to the strong e6
knight, nevertheless his position should be
playable.) 26 . . . l"1ab8 27.if5 ic6 2 8 .ig4 g6
29 .ixh5 gxh5 30.tLle2 White dominates the
centre.
23 . . . l"1e8! Black's most straightforward plan
is to get rid of the unpleasant knight on e6.
24.g5 (24.a3 can be met by 24 . . . �f7!? or
1 994
24 . . . ic8 .) 24 . . . tLlh5 2 5 .1"i:e3 g6 26.ie4 (26.a3
tLl g7) 26 . . . ixe4 27.1"i:xe4 1"i:a7 28.a4 (28 . tLl d5
Ei:d7) 28 . . . bxa4 29.tLlxa4 tLl g7 Black should be
able to survive.
a
b
c
d
269
Anatoly Karpov - Vladimir Kramnik
f
e
g
h
24.f4!
Karpov intends to plough forwards with g5 ,
which will weaken the enemy kingside and
drive away the defending knight.
24 if3
..•
Another possibility was:
24 . . . ib4
Active play on the queens ide is no longer
enough to neutralize White's initiative on
the kingside.
25.g5 hxg5 26.fXg5 tLl e8
26 . . . tLld5?! 27.1"i:e5! (27.tLlxd5 ixd5 28.1"i:xd5
ixe 1 29.1"i:d7 )f;>h8 White will have no more
than a perpetual check.)
27 . . . ixc3 (27 . . . tLlxc3 28.1"i:d7!; 27 . . . tLl c7
28 .if5 !) 28.bxc3 1"i:e8 29.if5 )f;>f7 30.1"i:de 1
g6 (30 . . . tLlxc3 3 1 .ih7! catches Black's king.)
3 1 .ic2 tLl xc3 32.1"i:fI t )f;>g8 33 .ixg6 1"i:e7
34.1"i:ef5 White's attack decides the game.
27.if5
There is also 27.a3 ia5 28.if5 if3 29.1"i:d2
tLl c7 30.)f;>f2 ic6 3 1 .1"i:e5 when Black is
under some pressure.
27 . . . ic5 t
If 27 . . . if3 28.1"i:d4! the rook joins the
attack.
28.tLlxc5
After 28.)f;>fI ?! tLl d6 29.tLlxc5 1"i:f8! Black
obtains good counterplay.
28 . . . 1"i:xc5 29 .ie6t cj;Jh7 30.h4
Despite the reduced material, White
maintains strong pressure.
25J!d2 ic6
Black gave up a tempo to lure the rook to
d2, perhaps hoping to gain a tempo with . . . b4
and . . . c3 later.
26.g5
The pawn drives the knight away and exerts
a choking effect on Black's kingside.
26 ... hxg5 27.fxg5 tLld7
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
a
2S.tLlxf8!
b
c
d
e
f
g
h
270
The Prime Years
Karpov makes a difficult but correct decision,
exchanging his powerful knight in order to
invade on the seventh rank.
2BJ'!f2 looks tempting, intending i.g6-f7t and
l'!f4, but Black can defend: 2B . . . l'!eB (2B . . . lLl c 5?
loses to 29.l'!xfBt; 2B . . . i.c5 i s possible though,
and after 29.lLlxc5 lLl xc5 30.l'!e5 lLl d7 3 1 .l'!e6
lLl c5 Black can neutralize his opponent's
initiative.) 29.i.g6 i.c5 ! 30.i.xeB l'!xeB 3 1 .l'!ee2
i.xf2t Black has solved all his problems.
28 lDxf8 29J�d6!
...
Karpov makes a characteristic prophylactic
move, taking away the g6-square from the
knight and tying the cB-rook to the defence
of the bishop.
29 b4
...
Kramnik has said that in his childhood
he investigated Karpov's play a lot by using
a book from his local library. Therefore it is
a bit ironic that he now goes on to lose in a
similar manner to several of Karpov's previous
victims: he strives for counterplay but loosens
his position, and Karpov goes on to punish
him on the flank where he was supposed to
be stronger.
Having said that, it is important to clarifY
that the text move is not a mistake in itself Black really does have to do something on the
queenside. Other moves were possible and it
must have been hard for Kramnik to choose
between them, as he is somewhat worse in all
variations and has to try to find the one that
gives the best drawing chances. Here is one
such example:
29 . . . l'!abB
Black begins his queenside play in a slightly
different way.
30.l'!e7 b4 3 1 . lLl e4
Alternatively after 3 1 . lLl e2 l'!eB 32.l'!xeB
i.xeB 33 .h4 a5 34.i.e4 l'!cB 35.<;t>f2 White is
somewhat better in the ending.
3 1 . . .b3!?
3 1 . . .i.b5 32.h4 l'!eB (the evaluation is
similar after 32 ... l'!c6 33.l'!xc6 i.xc6
34. c;t>f2) 33.l'!xeB l'!xeB 34. <;t>f2 lLl d7 Black
is somewhat worse, but he should be able to
hold the endgame.
32.axb3 cxb3 33 .i.d l
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
33 . . . i.aB
Alternatives include 33 . . . i.eB 34.<;t>f2 l'!c2t!?
3 5 .i.xc2 bxc2 36.l'!c7 i.g6, and 33 ... i.d7
34.h4 l'!c2!?, with counterplay in both cases.
34.lLlg3!
Threatening to go to f5 or h 5 .
34.lLlf2 l'!c 1 3 5 .l'!e l (or 3 5 . l'!xa6 l'!b l )
35 . . . l'!c6 36.l'!d3 l'!cb6 37.l'!ee3 l'! b 5 38.h4
lLl g6 Black obtains reasonable counterplay.
34 . . . l'!dB 3 5 . l'!xdB l'!xdB 36.i.xb3t i.d5
Black is a pawn down, but he has decent
drawing chances due to the limited number of
pawns remaining.
a
b
c
d
e
f
g
h
1 994
27 1
Anatoly Karpov - Vladimir Kramnik
30 ...ie8
33.lLlxd6
Black had other moves as well.
30 ... aS 3 1 . l2l d2 .ibS 32.l:'i:eS l:'i:dB 3 3 .l2l e4 .ieB
34.l:'i:xdB l:'i:xdB 3 5 .l:'i:xa5 l2l e6 Black has some
compensation for the pawn, and the limited
number of pawns also helps his drawing
chances.
30 . . . .ib5 3 1 . a4!? (3 1 .h4 l2l d7 32.'it>f2 l2le5
33.\t>g3 l:'i:fB Black remains alive.) 3 1 . . .bxa3
32.bxa3 l:'i:eB 33.'it>f2 .id7 34.h4 l:'i:abB 3 5 .l:'i:d4
l'!b2 Black obtains some counterplay.
31.lilg3 l:'i:d8
Kramnik exchanges the strong rook in the
centre.
Reducing the number of pawns would not
have guaranteed a draw: 3 1 . . .c3 32.l2lf5 'it>hB!
33.bxc3 (33.l:'i:d4 b3; 33. l2le7 l:'i:c5) 33 ... l:'i:xc3
34.ie4 l:'i:bB 3 5 .h4 a5 (35 . . . b3 36.axb3 l:'i:cxb3
37J'!xa6 .id7 Black may be able to draw this
position, but he will have to suffer for quite some
time.) 36 . .id5 l:'i:c7 Ultimately Black should be
able to hold this position, nevertheless White
can continue to set him problems with 37.l2le7
or 37.l:'i:a6 a4 3B.l:'i:b6.
32.lilfS :Sxd6
If32 . . . g6 33. l2l h6t i> g7 (33 . . . i> hB? 34.l:'i:xdB
l'!xdB 35 . .ia4 White wins.) 34.l:'i:e7t 'it>hB
35J�xdB l:'i:xdB 36.h4 White has good winning
chances.
Thanks to Karpov's effective use of the
knight, it replaces the rook on a fine square
and conveniently attacks the c4-pawn.
33 ...ig6?
Kramnik was probably short of time by now,
and he hastily sacrifices a pawn in order to
exchange White's strong bishop. He probably
planned something different here, but then
changed his mind after spotting something
that was not to his liking. Black's defensive task
is certainly not easy, but his position should
still be tenable. There were two alternatives
which deserved attention.
33 . . . .if7 34.l2lxf7
34.l:'i:e7 .ie6 (34 . . . .id5 ? 3 5 . l2l f5) 35 ..if5 c3!
36 . .ixe6t l2l xe6 37.l:'i:xe6 c2 3B.l:'i:el l:'i:dB
39.l:'i:c1 l:'i:xd6 40.l:'i:xc2 i> f7 White is a pawn
up in the rook ending, but he has virtually
no winning chances.
34 . . . 'it>xf7
White has the superior minor piece, but
Black has chances to become active and there
are fewer pawns on the board than White
would ideally like.
35 .l:'i:e4 l:'i:cB 36.b3
36.l:'i:xc4 l:'i:xc4 37 ..ib3 a5 3B . .ixc4t i> g6 is
equal.
a
b
c
d
e
f
g
h
36 . . . l2l e6
36 . . . a5 37.l:'i:xc4 l:'i:xc4 3B.bxc4 l2l e6 39.h4
l2l c5 40. i> f2 White has decent winning
a
b
c
d
e
f
g
h
The Prime Years
272
chances, as his king will dominate in the
centre.
36 . . . cxb3 37.l"i:f4t! rj;JeS 3S . .txb3 l"i:c5 39.h4
a5 40.g6 Black is struggling.
37.h4 c3 38 .l"i:xb4 a5 39 .l"i:e4 l"i:d8 40.l"i:c4 l"i:d2
Black can also consider 40 . . . l"i:d4 4 1 .l"i:xc3
l"i:xh4 when he has reasonable drawing
chances.
4 1 .l"i:xc3 ttJ d4 42.l"i:c7t rj;Je6 43 . .te4 i>e5
Despite being two pawns down, Black has
good chances to survive as his pieces are so
active.
33 . . . .td7 34.i>h2!?
White patiently improves his king.
If 34.ttJxc4 .txh3 35 . .tb3 l"i:d8 36.ttJ b6t
rj;Jh7 37.ttJd5 .te6 Black is active enough to
hold.
Alternatively, after 34.h4 .te6 3 5 .l"i:e4 b3
36.axb3 cxb3 37 . .td3 a5 3S.i>f2 a4 39.ttJb5
.to 40.l"i:b4 ttJ d7 Black is not completely
home and dry, but he is not far from it.
34 . . . .te6
The bishop stands well here.
3 5 .l"i:e4 b3 36.axb3 cxb3
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
37 . .td l
After 37 . .td3 l"i:d8 38.l"i:d4 a5 Black holds.
37 . . . a5 38 . .te2 a4 39. ttJ c4 .txc4
39 . . . l"i:dS 40.i>g3 i>h7 should also be good
enough.
40 ..txc4t rj;Jh7 4 1 .l"i:f4 g6
Black has nothing to worry about.
34.,LgG tlJxg6 35.tlJxc4
Karpov collects the pawn which was
weakened by Black's own pawn advance from
j ust six moves ago. In additional to his material
advantage, his pieces remain more active.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
35 .. J:�d8?!
Kramnik improves the rook, but that could
and should have been done later.
3 5 . . . ttJ h4!
Improving the knight was stronger, as White
does not have an ideal way to deal with
the check on f3 . Black's situation remains
precarious, but he could certainly have
offered stubborn resistance.
36.l"i:e6
36.l"i:fl l"i:c8 37.l"i:f4 (37.b3? l"i:c5!) 37 . . . ttJf3t!
Exchanging the knights serves Black's
interests, and the rook ending is likely to
end in a draw.
36.i>f2 l"i:fSt 37.rj;Jg3 ttJ f3 38.l"i:fl (After
3 8 .l"i:e3!? ttJxg5 39.h4 White keeps some
initiative, but it will be hard to turn it into
a win with so little material remaining.)
38 . . . ttJxg5 39.l"i:xf8t i>xf8 40.ttJe5 Black
will lose a pawn, but will do it by playing
. . . b3 to ensure that White's pawns will be
doubled. White has chances to win, but it
will not be easy.
36 . . . ttJ f3t 37.i>g2 ttJxg5 38 .l"i:b6 a5 39.l"i:b5
1 994
Anatoly Karpov - Vladimir Kramnik
White wins a pawn, but winning the game
will require skilful technique as there are so
few pawns left on the board.
a
b
c
d
e
f
g
h
36J:�e4!
This superb move restricts the enemy knight,
prepares h4 and defends the c4-knight.
36 h3 37.axh3
...
Karpov allows Black to exchange a pair of
pawns, as in return he gains time.
Nevertheless 37.a4!? intending <.t;g2 and h4
also wins, and may even have been stronger
than the game continuation.
37 �d3 38.<.t;g2 �xh3 39.h4! lbf8
...
39 . a5 40.h5 ctJ f8 4 1 .g6! paralyses Black's
kings ide, and after 4 1 . . .a4 42.�e5 �b4 43 .�c5
Elb3 44. Wf2 White is winning.
. .
273
40J:�e8!
1-0
Kramnik overstepped the time limit. It was
a happy ending for Karpov, although in a way
it is a pity that we did not get to see these two
great endgame maestros play to the end.
40 . . . �f7!
After this move Black's position remains
difficult, but he certainly has chances to
resist.
4 1 .�d8
4 1 .ctJd6t �g8 42.h5 looks scary for Black,
but he should be able to snatch the b-pawn
without losing his knight. (If 42.g6 �b6!
Black defends.) 42 . . . �xb2t 43.Wg3 �b3t
Black's position is ugly, but it looks like he
can survive.
Alternatively, after 4 1 .�e3 �b4 42.b3 White
is a pawn up but the win is still a long way
off.
4 1 . . .ctJ e6
4 1 . . . ctJ g6 42.�d4 maintains control.
42.�d6 ctJ f4t 43.�f1
White has reasonable chances to convert
his extra pawn, but Black's pieces are fairly
active so the win is certainly not a foregone
conclusion.
In the penultimate round Anand tried the
main line against the Caro-Kann, but Karpov
neutralized his play and it was not long before a
draw was agreed. Karpov's opponent in the final
round was Beliavsky, who was absolutely out of
form and had one of the worst tournaments of
his life, scoring only fours draws out of twelve
games. He played disastrously against Karpov
as well, as he allowed one of his knights to
become trapped and he had to resign after j ust
twenty moves.
Thus Karpov won the tournament with an
incredible score of 1 1 / 1 3 , two and a half
points clear of Kasparov and Shirov who
a
b
c
d
e
f
g
h
The Prime Years
274
were equal second. The average rating of the
participants was 268 5 , the highest ever at
the time. Karpov's Tournament Performance
Rating of 2985 was the highest of all time,
although in 2009 Magnus Carlsen surpassed it
by winning the Pearl Spring tournament with
8/ 1 0 and a TPR of 3002. Nevertheless taking
into account rating inflation I would still rate
Karpov's achievement as the more impressive.
Kasparov suggested that luck played a
role, pointing out that thanks to a somewhat
favourable drawing of lots, almost all of
Karpov's opponents met him immediately after
facing Kasparov in the previous round. The
world number one was also keen to emphasize
that a number of Karpov's wins came as the
result of blunders by his opponents. It would
seem fair to say that Karpov benefitted from
the rub of the green, but equally, one cannot
achieve such a result by luck alone! Karpov
won two truly great games against the future
world champions Kramnik and Topalov, and
played excellent chess throughout most of the
tournament.
Karpov reached his highest ever rating
of 2780 on the July list of 1 994, and the
supersonic Linares result had a lot to do with
it.
***
Karpov's next tournament was Dos Hermanas.
Following his other-worldly display in Linares,
there must have been a great deal of pressure
on him to perform. In the first round he failed
to equalize on the black side of a l .c4 e5
English against Rivas Pastor. He managed to
turn the tables and gained the upper hand in
the middlegame, but the Spanish grandmaster
found a way to give a perpetual check. In
Round 2 Karpov got no opening advantage
against Gelfand's King's Indian. Later he
managed to get his rook to the seventh rank
and won a pawn, but Gelfand was able to hold
the ending.
Karpov's next opponent was Morovic
Fernandez, who played seventeen games
against the world champions, scoring one win,
nine draws and seven defeats. His lifetime score
against Karpov is five draws and five losses.
I Game J� I
Ivan Morovic Fernandez - Anatoly Karpov
Dos Hermanas
1 994
l .c4
Morovic rarely opens with his c-pawn, but
he had no joy against Karpov's Caro-Kann in
their match earlier the same year.
1 ...lLlf6 2.lLlc3 e5 3.lLlf3 lLlc6 4.a3
Karpov likes to develop his bishop to b4
in this variation, so the Chilean grandmaster
prevents it.
4 ... g6
Previously Karpov had played 4 . . . d5 once
and 4 . . . d6 twice in this position.
5.g3 i.g7 6.i.g2 0-0 7.0-0 d6 8.d3 lLld4
This move had rarely been played before
the present game, although the Hungarian
grandmaster Ivan Farago had obtained a couple
of draws with it.
9.lLld2
Avoiding the exchange is White's most
challenging approach, and has been the usual
choice.
9 c6 10.b4
.•.
Morovic decides not to send the knight back
j ust yet.
Here is an interesting game featuring active
play from Black: W.e3 lLl f5 (The way Karpov
plays in this game suggests that he may well
1 994
275
Ivan Morovic Fernandez - Anatoly Karpov
have chosen this retreat square, although
10 . . . 'Ll e6 has scored better.) 1 1 . 'Ll f3 E&e8 1 2.e4
I2l d4 1 3 .'Llxd4 exd4 1 4 . 'Ll e2 d5 1 5 .exd5 cxd5
1 6.'Llxd4 ig4 1 7. 'Ll f3 dxc4 1 8 .dxc4 Wb6
1 9.�a4 'Ll e4 Black had nice compensation for
the pawn and a draw was soon agreed, Ruck Groszpeter, Austria 2004.
8
7
6
5
4
3
2
1
i. � � ��.m
� lm�_',Y-.r
',��
/- " " /-U rEI;�
�
�� . �
�WJ �"ti'� ��
� ��,���
'''//- ��,
'� '''', ./-�:
:fl�
�/-�
��
"
0%
%'0%
iO%
"",z�/"""U��.f�
�'''''�� �.!=""
a
b
c
d
e
f
g
h
15 lLlcS!
••.
Black defends d6 but more importantly
goes after the c4-pawn. Karpov evaluated the
position as clearly better for Black, which seems
a trifle optimistic, although I would certainly
agree that his position is to be preferred.
a
b
c
d
e
f
g
h
10 ...ie6!?
This was a novelty at the time; previously
Farago played 1 0 . . . d5 here. Karpov wants to
complete development before taking action in
the centre.
1 1.ib2 Wd7 12.e3 lLlf5 13.tiJde4 lLlxe4
There was a positional threat of 1 4 . 'Ll g5
eliminating an important bishop. Karpov
prevents it by exchanging the knight, although
1 3 . . . h6 was a valid alternative.
14.dxe4?
White hopes to develop pressure on the d-file,
and underestimates the dangers associated with
his slightly weak c4-pawn.
After 1 4.'Llxe4 Black can choose from several
plans, including 1 4 . . . h6 followed by . . /iJ e7
and .. .f5 .
14... lLle7 15J�'e2
16.E&fdl :gdS 17.llJa4
After 1 7.E&d2 lLl b6 1 8 .c5 'Ll c4 1 9 .E&d3 We7
Black has the advantage.
17 ...WeS lS.:gd2?
1 8 .ic3! was necessary, and after 18 . . . a6
1 9 . 'Ll b2 b5 20.a4 White is not worse.
276
The Prime Years
The immediate 1 8 . . . bS only leads to equality
after 1 9.cxbS cxbS 20.tLlc3, so Karpov prepares
to recapture with the a-pawn.
. . . Ei:d8 followed by . . .i.c4-d3 and/or . . . tLl c4.
Alternatively he might j ust push the
c-pawn.
19.i.c3
19 ... b5 20.tLlb2 lLl b6 2 1.cxb5
The Chilean grandmaster prepares to retreat
his knight to b2, in order to provide some
defence of the weak c4-square.
If 2 1 .cS dxcS 22.bxcS tLl c4 23.tLlxc4 i.xc4
24.Ei:xd8 Ei:xd8 2S .'lMrc2 'lMre6 Black will press
with his pawn majority on the queenside.
1 9.cS!?
This was worth considering, although it does
not solve White's problems entirely.
1 9 . . . i.b3! 20.tLlc3
After 20.tLlb6 tLlxb6 2 1 .cxb6 as Black gets
the upper hand.
20 . . . dxcS
Black can consider a different way of
changing the pawn structure: 20 . . . bS!? 2 1 .a4
dxcS 22.bxcS b4 23.tLla2 as The protected
passed pawn gives Black some advantage,
despite the fact that it is securely blockaded
for the time being.
2 1 .bxcS Ei:xd2
2 1 . . .b6 22.Ei:xd8 'lMrxd8 23.cxb6 tLlxb6 is
similar.
22.'lMrxd2
21. .. axb5 22.Wfdl
8
7
6
5
4
3
22 . . . 'lMre7 is less effective: 23.a4 bxa4 24.tLlxa4
tLl c4 2 S .Ei:c2! (After the more obvious 2S .Ei:da2
Ei:a7 26.i.fl Ei:da8 the pin on the a-file is a
problem for White.) 2S . . . 'lMrb7 (Now 2S . . :1'\a7
is met by 26.Ei:cc l Ei:da8 27.tLlb2 and White
eases Black's grip on the queenside.) 26.Ei:ccl
'lMrbS 27.i.fl Ei:d7 28. tLlb6 (28 .i.e l also looks
alright) 28 . . . Ei:xa l 29 .i.xa l White seems to be
okay.
.
,,,, , ju.'m.,,
2
a
b
c
d
e
f
g
h
22 . . . b6!
After 22 . . . 'lMre7 23.a4! White is out of the
woods.
23.cxb6 tLlxb6
Black's superior pawn structure gives him
some advantage. His plans may include
23.a4
White must have pinned his hopes on
this move which aims to neutralize Karpov's
structural advantage, but it arrives a bit too
late.
1 994
Worse is 23.exd5 ? ttJ xd5 ! 24.Wc2 ttJ xc3
25.Wfxc3 e4 26.Wc2 Elxd2 27.Wxd2 id5 and
White is lost.
23 ... ttJc4?!
Karpov plays strategically and obtains a
strong passed pawn for himself. He keeps an
advantage this way, but there was an even
stronger continuation.
23 ... bxa4! The tactics in the centre yield a clear
advantage for Black: 24.ttJxa4 ttJ c4 2 5 .Eld3 d4!
(Karpov mentions 25 . . . dxe4 26.ixe4 f5 with a
slight plus, but Black can do better than this.)
26.exd4 exd4 27.ixd4 (If 27.ie 1 Ela6 Black
doubles on the a-file.)
8
7
277
Ivan Morovic Fernandez - Anatoly Karpov
8
7
6
5
4
3
2
1
a
b
c
d
f
e
g
h
25 .if8!
.•.
Karpov wastes no time in preparing . . . c5.
26.a6?
L-;-;-;:O ////·///'
6
5
4
3
2
a
b
c
d
e
f
g
h
27 . . . ig4!! This lovely tactical shot highlights
the fact that White's pieces are overloaded.
28.f3 (28.Wxg4 ixd4-+) 28 . . . Elxd4! 29.Elxd4
ttJe3 30.Wd2 ixd4 3 1 .Wxd4 ttJ c2 Black
obtains a decisive material advantage.
Morovic must have hoped that the a-pawn
would provide him with counterplay, but it
becomes a target which Karpov will collect.
Exchanging queens would have led to a
passive ending, but it was the lesser evil.
26.Elxd8 Wxd8 27.Wxd8 Elxd8
Let's see how White can try to hold.
28 .ifl
28 .ixe5? ixb4 29.a6 E1a8 30.a7 ic5 3 1 .id4
(3 1 .Ela6 id7) 3 1 . . .ixd4 32.exd4 b4 Black's
passed pawns are irresistible.
28.'i!?fl f6 29.'i!?e2 (Or 29.a6 Ela8 30.f4
Ela7 and the a-pawn falls.) 29 . . . c5 30.bxc5
ixc5 3 1 .Elb 1 b4! 32.ixb4 ixb4 33.Elxb4 c3
34.Elb l ig4t 3 5 . f3 Eld2t Black wins.
28 .. .f6
24.ttJxc4 dxc4
Black's protected passed pawn is a major
asset, despite being securely blockaded for
the time being. Black has reasonable chances
to win or exchange the b4-pawn, in order to
obtain connected passed pawns.
25.a5
After 2 5 . axb5 Elxa 1 26.ixa 1 Elxd2 27.Wxd2
cxb5 White faces a difficult endgame.
a
b
c
d
e
f
g
h
278
The Prime Years
29.:B:b l
Holding up the advance of the c-pawn.
29.a6 i> f7 30.f4 :B:a8 3 1 .fxe5 fxe5 32.a7 i> e8
and Black's king will collect the a-pawn.
29 . . . i> f7 30.f4 �d6
Black has time to stabilize his centre before
slowly preparing . . . c5.
There is no sense in playing 30 ... exf4?
3 1 .gxf4 f5 32.e5 when White may be able to
keep his position together.
3 1 .�e2 11e7 32. i> f2 �d7 33.�f3 :B:a8 34.11e2
c5 3 5 . bxc5 �xc5
Black has good winning chances.
Instead of this move the tricky 28.a7!
would have given Black a chance to go wrong:
28 . . . :B:xd2? 29.Wxd2 We7
a
26... f6
Karpov defends the pawn and restricts the
c3-bishop.
27.f4
With the pawn on a6 the endgame is even
less favourable for White: 27.:B:xd8 Wxd8
28 .Wxd8 :B:xd8 29.�f1 :B:a8 30.a7 11f7 3 1 .:B:a6
l1e7 (3 1 . . .�d7 32.f4) 32.:B:xc6 :B:xa7 33.:B:b6
:B:a3 And Black is winning.
8
b
c
d
e
f
g
h
30.�h3!! White can improve his passive
bishop with this sweet tactical stroke. 30 . . . if7
3 1 .�d7! Attacking the b5-pawn. 3 1 . . . exf4
32.bxc5 Wxe4 (32 . . . Wxc5 33 .�d4) 33.�d4
Thanks to his energetic play White manages to
stay in the game.
However, after 28.a7! Black has a stronger
reply: 28 . . . :B:dc8! 29.Wb l :B:c7 30.:B:da2 �e7
Now Black's pressure against the b4- and a7pawns is too much to bear.
28 ...�xd8 29.�el
7
After 29.Wxd8 :B:xd8 30.bxc5 �xc5 Black is
winning.
6
5
8
4
7
3
6
2
5
a
b
c
d
e
f
g
h
27 ... c5!
This move not only ensures Black connected
passed pawns, but also usefully opens the sixth
rank to attack the a6-pawn.
28J�xd8
4
3
2
1
a
29 ...�b6
b
c
d
e
f
g
h
279
Ivan Morovic Fernandez - Anatoly Karpov
1 994
Now the a-pawn falls, and White's position
quickly becomes hopeless.
hopelessly lost after 38 . . . :lih3 . (The alternative
38. <;t>f2 could have been refuted in the same
way.)
30.fxe5 fxe5 3 1 ..be5 :!:ha6 32.gdl
32.l"1xa6 'l&xa6 33 .'I&f2 :lie7 34.bxc5 b4-+
32 ...ig4
By this stage Karpov was in time trouble, but
his position is comfortably winning and he had
no problem finding the right moves. 32 . . . cxb4
33 .'.Wf2 :lie7 was also good enough.
33.l"1bl '?;Ve6
33 . . . cxb4 wins as well.
34.ic3
a
b
c
d
e
f
g
h
In Round 4 Karpov used the English Opening
against Illescas, and soon got the advantage of
the two bishops. He neutralized his opponent's
activity and won a nice endgame. Then he
made a quick draw with Lautier on the black
side of a Nimzo-Indian. In Round 6 Karpov
repeated the 2.c3 Sicilian against Judit Polgar,
who once again failed to equalize and later
gave away a free pawn. Karpov had no trouble
converting his advantage.
In Round 7 Karpov's forty four game
unbeaten streak came to an end against Boris
Gulko. In a 3 .e5 Caro-Kann Karpov went for
artificial castling and tried an unorthodox rook
manoeuvre on the kingside, but it turned out
badly. He struggled on past the time control
but was unable to save the game. In Round 8
he had the black pieces again, and was able to
take a convenient rest by means of a very quick
draw with Epishin.
In the ninth and final round Karpov faced
Topalov. The Bulgarian grandmaster was
able to equalize, but later he made a mistake
which allowed Karpov to showcase his tactical
alertness. Here is the key moment of the
game:
34 ... ga2!
Invading on the second rank comes naturally
to Karpov, even with little time on the clock.
White is unable to defend his king.
35.bxc5 ixc5 36.i.d4
Now Karpov finishes the game with a
flourish. In the event of 36.l"1xb5 White would
have suffered the same fate.
36 ... gxg2t! 37.@xg2 '?;Vxe4t 3S.@gl
0-1
As he played this move Morovic Fernandez
overstepped the time limit, but his position is
a
b
c
d
e
f
g
h
30.ttJf6!! @xf6 3 1 .:lie5 t!! <;t>xe5 3 2.'I&xe4t
<;t>xe4 33.l"1e I t <;t>f5 34.l"1xe8 Black soon had
280
The Prime Years
to resign, Karpov - Topalov, Dos Hermanas
1 994.
Karpov's final score of 6/9 was enough for
second place, half a point behind Gelfand and
half a point ahead of Epishin who was third.
Karpov's next tournament was Las Palmas
in the Canary Islands, where he started with
White against Topalov. Having suffered two
recent spectacular losses in tactical combat
against the former champion, Topalov opted
for solidity in the Queen's Gambit Accepted,
and managed to draw without too many
problems.
In the second round Karpov faced Alexei
Shirov. The Latvian-born grandmaster has,
at the time of writing, contested 1 63 games
against world champions, scoring twenty two
wins, eighty nine draws and fifty two losses.
Before this meeting Karpov had beaten him
twice and drawn three times. Their lifetime
score is a lopsided seven wins to Karpov, with
twelve draws and no defeats - a truly dominant
record against a world-class opponent.
I Game 36 I
Alexei Shirov - Anatoly Karpov
Las Palmas
1 994
l .e4 c6
On the one previous occasion when Shirov
opened with 1 .e4, Karpov replied with
1 . . .e5 . Nevertheless the switch to the Caro­
Kann could not have come as a surprise.
2.d4 d5 3.e5 i.f5 4.tlJf3
One would expect the 4 . lLl c3 variation from
Shirov.
4 ... e6 5.i.e2 c5
Karpov and Shirov contested several games
in this variation, but the following was the
only one which occurred at a regular time
control. Karpov used a different move order,
and Shirov a different plan: 5 . . . lLl e7 6. 0-0 c5
7.c4 lLl bc6 8 . lLl c3 dxc4 9.dxc5 lLl d5 1 0.lLld4
lLlxc3 1 1 .bxc3 .ixc5 1 2.lLlxf5 exf5 1 3 . .ixc4?!
.ixf2t 1 4.'it>h 1 Wxd 1 1 5 .l'!xd 1 lLlxe5 White
had insufficient compensation for the pawn
and Black went on to win this game as well,
Shirov - Karpov, Vienna 1 996.
6.0-0
Nowadays the more dynamic 6 . .ie3 is
considered critical, and over the course of his
career it was this move that gave Karpov the
most problems.
6 ... tlJc6 7.c3 cxd4
Karpov also delayed this capture with 7 . . ig4
a couple of times.
.
8.cxd4 tlJge7 9.a3
9.lLl c3 is the main line, although the text move
has also been popular. Please remember that
the Short Variation had only been recognized
as a serious line for a few years, and at the time
players were still experimenting to find the best
move orders and piece placements.
9 ... tlJc8
In Game 4 of the Short match Karpov
played 9 . . . .ig4 followed by . . . lLl f5 , but failed
to equalize and eventually lost. The plan of
deploying the knight on the queens ide was
first seen in 1 993, and has since become the
main line.
10.tlJbd2
1 O. lLl c3 is also quite playable.
10 ...i.e7!
Karpov remains flexible and puts his bishop
on its only decent square, in order to hide his
intentions with the c8-knight.
1 l.b3
1994
28 1
Alexei Shirov - Anatoly Karpov
White takes away the c4-square from the
knight, but he is playing rather timidly. The
more ambitious I l . b4 would have transposed
to a more popular position.
16 ...'iNb6! 17.idl
It must have been unpleasant for White to
put the bishop on such a passive square, while
also breaking the connection between the
rooks.
17 ... �c7! lS.h4 h6 19.�e2 �fcS
After making room for his bishop, Karpov
increases his control over the only open file.
20.'iNf4 tLlbS!
The knight vacates the c-file and prepares
two important piece manoeuvres.
21.�e3
a
b
c
d
e
f
g
h
l l ... a5!
Karpov clears the a7-square for the c8knight. Later his queen can go to b6, from
where she can support the further activation of
the knight via b 5 .
12.ib2 tLlSa7 13.�el 0-0
It is interesting that Karpov waited to find
good squares for his minor pieces before
castling.
14.�f1 gcs 15.tLlg3 ig6 16.'iNd2?!
White should have preferred 1 6.id3
although Black has no problems here either.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
21 ...tLlb5!
Karpov makes great use of his knights; the
position is closed, which is why such long
manoeuvres are possible. In some variations
the knight may even end up in the centre on
e4. It is not unusual for Black's g8-knight to
arrive on the e4-square, but it seldom makes
the journey via such a long route.
22.tLle2
Shirov stops . . . tt:l c3 .
22 ... tLld7!?
A prophylactic move for the sake of safety.
Karpov probably j ust wanted to bring a
a
b
c
d
e
f
g
h
282
The Prime Years
piece closer to his kingside to reduce White's
attacking chances.
Karpov advances with the knight in order to
increase his advantage on the queenside.
Acting on the queens ide with 22 . . . i.c2 was
possible, despite the fact that Black has almost
no pieces around his king, as White still cannot
hurt it. 23.a4 i.a3 24.i.xa3 lt'lxa3 2 5 . lt'l e l Wb4
Black has the initiative on the queenside, but
he is far from breaking in.
27.i.xa3?
23.ilJh2
Shirov gets ready to launch an attack on the
kingside.
23 ... �c6!
Karpov pays attention to what his opponent
is doing, and clears a path for the queen to get
to the kingside.
24.ilJg4 Wd8 25.�h3
a
b
c
d
e
f
g
Shirov has to give up one of his bishops
for the enemy knight, but he chooses an
unfortunate way to do it.
27.h5 i.h7!
Better than 27 . . .i.c2 28.Elg3 mhB 29.Elc3
i.b4 30.Elxc6 Elxc6 3 1 . lt'l e3 i.xd l 32.Elxd l
It'l c2 33 .i.c3 when White manages to keep
his queens ide together.
Instead Black should play for . . . It'lc2 in
order to exchange his knight for the enemy
bishop.
2B.Elg3!
The rook sends the king to the corner, tying
Black's queen to the defence of the f7-pawn.
If2B.lt'lc3 ? i.g5 29.Wf3 Wb4 Black is able to
invade on the queenside as f7 is protected.
White can also try 2B.Elc3 It'l c2 29.ixc2
i.xc2 30.Wd2 Elxc3 3 1 .i.xc3 i.xb3 32.ixa5
when he is surviving, although Black remains
somewhat better.
2B . . . mhB
h
25 ...W!'f8!
From this square Karpov's queen defends the
kingside while at the same time exerting pressure
on White's queenside. The queen already did
an excellent job by driving the bishop to d l
and supporting . . . It'l b 5 , but once those tasks
had been accomplished Karpov was able to find
a better square for his strongest piece.
26.a4 lba3!
a
b
c
d
e
f
g
h
29.lt'lc3!
29.lt'le3 enables Black to penetrate in a
most unusual way: 29 . . . lt'l b l ! 30.Wg4 tLl b6!
Stopping any It'lxd5 tricks. 3 1 .lt'l f4 tLlc3
White is under pressure, and the knight
may cause further problems from e4. It is
extraordinary for the gS-knight to make a
journey from e7-cS-a7-b5-a3-b l -c3 to get
to the e4-square.
29. . . tD c2
If 29 . . . j,b4 30.tDa2 White keeps his
queens ide together.
3o.ixc2 j,xc2 3 1 .tDb5
White continues to resist on the queenside.
27...ixa3
283
Alexei Shirov - Anatoly Karpov
1 994
stronger, as it is here that Black's unopposed
dark-squared bishop can make the biggest
difference. 30.lt>h2 (30.Ei:a2 iWd8 wins the h4pawn.)
28J�� g3
a
b
c
d
e
f
g
h
30 . . . iWd8 (There is also 30 . . . iWeS! ? to defend
the g6-bishop and prepare . . . f6. 3 1 .tDgl f6
White is in trouble here as well.) 3 1 .Ei:h3 j,e4
With . . . tD fS coming next, White is in trouble.
30,Ei:al i.c1 3 1.Y;Yg5
a
b
c
d
e
f
g
h
28 ... h5!
Karpov drives back the knight in order
to ease the pressure on his kingside, having
correctly j udged that his opponent is in no
position to exploit the weaknesses created by
this pawn advance.
29.tD e3!
Shirov chooses the best retreat square. It
was important to block the c 1 -h6 diagonal, as
shown by the line 29.tDh2? j,b2! 30.Ei:a2 j,c 1 !
3 1 .tDxc 1 Ei:xc 1 32.Ei:d2 iWb4 33.tDf1 tDc5! and
Black wins.
29 ,j,b2?!
Karpov tries to invade on the queenside. His
plan prevails, but only after some help from
his opponent.
••
29 . . . j,e7! Returning to the kingside was even
After 3 1 .lDxc 1 Ei:xc 1 32.lt>h2 Ei:Sc3 33.Ei:d2
iWb4 34.iWg5 tD fS Black's pieces stand
menacingly.
31...'iMb4
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
32.tlJf4??
Shirov is a great attacker, but here he ignites
the fire naively and blunders badly.
284
The Prime Years
32.li>h2!
By removing his king from the vulnerable
first rank, White could have kept the game
interesting. Black may still be better here,
but he would have had to play accurately to
prove anything.
32 . . . lLl fB
After 32 . . . �xe3 33 .Wxe3 �b l 34.2'�b2 �f5
3 5 .Wd2 Black will have to work hard to
achieve anything real.
33.lLlxc l
After 33.Wf4 �d2 34.!'lg5 lLl h7 3 5 . !'lg3
White survives for the moment, but Black
has several ways to try and increase his
advantage.
33 . . . !'lxc l 34.!'lc2!
White exploits the pin on the g6-bishop to
force the exchange of one of the opponent's
strong rooks.
8
7
6
5
4
3
2
1
a
b
c
d
32 i.xe3! 33.fxe3 !'leI
e
f
g
h
•.•
Karpov does not pass up the chance to win
the bishop. It is a pity that a game so rich
in strategic ideas was decided by a horrible
blunder, although I'm sure Karpov did not
mind.
34.tiJxg6 gxdl t 35.c.flh2 fxg6 36.Wxg6 V!1e7
The queen defends the g7-pawn, and White
has nothing for the piece.
37J'U2?
In a hopeless situation White blunders
rook as well.
a
b
c
d
e
f
g
h
34 . . . lLl h7
After 34 . . . Wxd4 3 5 . !'lxcB !'lxcB 36.�xh5
lLl h7 37.Wg4 Wxe5 3B .�xg6 Exg6 39.Wxg6
lLlf8 40.Wd3 White is in the game.
3 5 .!'lxc8t !'lxcB 36.Wf4 We I 37.!'lf3 !'lfB! ?
Th e most ambitious move. Black's kingside
pieces are passive, but his troublesome queen
can still cause plenty of problems.
37 . . . !'lcl 3B .�c2! White is out of danger.
37 . . . !'lc7 3B .�c2! White should be able to
hold the position.
3B .�c2 Wd2
Black maintains some pressure.
8
7
6
5
4
3
2
1
a
b
c
d
e
37...Wxh4t 38.gh3 Wxf2
0-1
f
g
h
a
1994
285
Michael Adams - Anatoly Karpov
In the next round Karpov had the white
pieces against Morovic Fernandez, who was
turning into a frequent opponent for Karpov
in 1994. This turned out to be one of the rare
occasions when Karpov failed to defeat the
Chilean grandmaster, who played a solid Semi­
Slav and was never in real trouble. In Round 4
Lauder played the 4.Wc2 line against Karpov's
Nimzo-Indian. In a tense middlegame Karpov
weakened his kingside with . . . g5 , in return for
some active play. But then he made a serious
mistake and allowed his king to be driven to
the centre, and the game was soon over.
In Round 5 Karpov faced Epishin, who used
the solid . . . c6/ . . . d5 line in the Fianchetto
Griinfeld. For a while the position was
absolutely equal, but Karpov persevered and
was able to capitalize on some mistakes from
his opponent. In the next round Karpov had
the white pieces again, but was unable to make
any headway against Kamsky, who handled the
Semi-Slav similarly to Morovic, with the same
end result.
In Round 7 Karpov equalized easily against
Illescas in a Caro-Kann, and the game never
looked like being anything other than a draw.
Karpov's next opponent was Judit Polgar, who
he had defeated twice with the 2.c3 Sicilian
in recent tournaments. This time he switched
to l .d4 followed by the Fianchetto System
against the King's Indian. Once again he got
some advantage from the opening, and Judit
never recovered.
Karpov's opponent in the ninth and final round
was Michael Adams. The English grandmaster
has so far played 1 50 games against the world
champions. He won eighteen of them, drew
ninety two and lost forty times. This was their
second meeting; Karpov had won the first.
Their lifetime score is two wins to Karpov, one
win to Adams and six draws.
I Game 31 I
Michael Adams
-
Anatoly Karpov
Las Palmas
1 994
l .d4 lbf6 2 ..ig5
Adams picked up the Trompowsky from
Julian Hodgson. By 1 994 it had lost some of
its surprise value.
2 ... e6
Karpov always played this variation, which
offers White the chance to occupy the centre
at the expense of the bishop pair. He achieved
several nice wins, but also some losses in this
line, which reflects its double-edged nature.
3.e4
Adams occupies the centre at once. Karpov
scored an excellent win after his opponent
postponed it for a move: 3. c3 b6 4.e4 h6
5 .�xf6 Wxf6 6.e5 We7 7.Wf3 ttJ c6 B.Wg3 �b7
9.ttJd2 0-0-0 1 O. f4
a
b
c
d
e
f
g
h
1 0 . . . g5! Karpov opens the position as much
as he can for his bishops. l 1 .fXg5 ?! ( 1 1 .0-0-0
was better) 1 1 . . . Wxg5 1 2.Wf2 �e7 1 3 .ttJ h3?!
( 1 3 .ttJ gf3) 13 ... Wg7 1 4.�e2 f6! Having picked
a plan, Karpov follows it through with his usual
consistency. His strategy is to open the game for
his dark-squared bishop. 1 5 .exf6 �xf6 1 6.�f3
2"lhfB Having eliminated his opponent's pawn
286
The Prime Years
wedge Black stood better and went on to win,
Qi Jingxuan - Karpov, Lucerne 1 98 5 .
3...h6 4.i.xf6 '!Wxf6 S.lDc3 d6 6.'!Wd2
This move started to become the main line
in the early nineties. White often follows it by
playing f4.
Karpov faced 6.tt:lf3 on two occasions. 6 . . . g5 ! ?
( I n Game 1 9 o f the 1 974 match against
Korchnoi, Karpov played 6 . . . g6 and eventually
lost a long game. Later he pushed his pawn
further, hoping his strong bishop would help to
keep his slightly weakened kingside together.)
7.e5 We7 8 .ib5t id7 9.0-0 d5 1 O.id3 tt:l c6
1 1 . tt:l b 5 0-0-0 1 2.c3 h5 1 3 .a4 �b8 1 4 . b4 ig7
1 5 . tt:l a3 E!:df8 1 6.We2 g4 1 7. tt:l d2 f6 1 8 .exf6
ixf6 1 9 .b5 tt:l a5 20.f4 Wg7 At this stage Black
stands better, but later he was completely lost,
and in the end the game was drawn, Timman
- Karpov, HollandlIndonesia (9) 1 993.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
8 dxeS!
•.•
Karpov can afford to open the position, and
it is worth it to create a weakness on e5.
9.dxeS '!We7 10.f4 lLlc6 l 1 .lDa
White can also consider 1 1 .g3 id7 1 2 ig2
0-0-0 with equal chances.
.
1 l i.d7 12.h4
•••
6 ... gS!?
Perhaps the successful opening from the
Timman game inspired Karpov to adopt this
ambitious approach again.
7.0-0-0
If 7.tt:lf3?! g4! is strong.
In 1 99 8 Anand introduced a new scheme of
development involving 7.ic4 followed by
short castling and a quick f4. Karpov got an
unclear position out of the opening, but later
he went wrong and the Indian grandmaster
punished him.
7...i.g7 8.eS!?
Adams opts for a risky and ambitious
approach. In 1 996 he switched to 8 . g3 against
Topalov but lost that game too. Funnily
enough, Topalov switched sides and played the
same move against Karpov in a rapid game in
1 997, but Black won that encounter as well.
Adams decides to force the issue on the
kingside. It may have been worth directing
his attention towards the other flank with
1 2. tt:l e4!? 0-0-0 1 3 .a3 - but not 1 3 .Wc3?!
because of 13 . . . Wb4!.
12 ...gxf4
Naturally Karpov chooses to eliminate the
more central pawn, and in doing so he gets
closer to the weakness on e 5 .
13.'!Wxf4 0-0-0 14.lDe4
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
287
Michael Adams - Anatoly Karpov
1 994
14 �b8!!
•.•
The point of this great move is to improve
the d7-bishop, and to a lesser extent the knight
as well. The position is closed enough that
Black has j ust enough time to carry out the
intended manoeuvre.
15.�f6
This is not a bad move, but it is a step in the
wrong direction. Adams probably misj udged
the ensuing position.
The present position was reached in two
subsequent games:
1 5 .g4 ic6 1 6.ig2 � d7 1 7.g5 hxg5 1 8 .hxg5
�xh l 1 9 .:B:xh l
1 6.ic4! White has enough pieces to stop Black
setting up pressure on the e5-pawn. ( 1 6. ltJ f6?
h5! [Karpov evaluates 16 . . . Wc5 as slightly
better for Black.] 1 7.<j;Jb l ixf6 I S .exf6 Wd6
Black's advantage is bigger here.) 1 6 . . . ixe4
( 1 6 . . . h5 1 7. ltJg5) 1 7.Wxe4 White is not worse
as he has a good reply to all Black tries: If
1 7 . . . c6 I S . ltJ f3, if 1 7 . . . Wc5? I S .ltJxe6 or
1 7 . . . ltJ d7 I S . ltJ c6.
15 ...i.c6 16.i.e2 lLld7!
Karpov naturally wants to exchange the
pieces around the weakness on e5, and he
attached an exclamation mark to this move.
The reason is that as well as being tactically
j ustified the move requires fine j udgment of
the position in a few moves time.
From d7 the knight not only attacks the e5pawn but also covers the f6-square.
17.�h5?
Adams must have missed a detail in the forced
tactical sequence that follows this move.
a
b
c
d
e
f
g
h
Now 1 9 . . . Wb4?? 20.ltJ d6t 1 -0 was the
unfortunate end to Ionescu - Leskur, Bucharest
1 998, but Black could have obtained a slight
plus with 1 9 . . . ltJ fS .
1 5 . ttJ d4 ic6!? (More enterpnsmg than
1 5 . . . ltJ c6 1 6. ltJ f3 ltJbS 1 7. ltJ d4 ltJ c6 I S . ltJ f3
Y2-Y2 Gorevlov - V. Ivanov, Moscow 1 996.)
Going after the other bishop would also have
led to trouble for White: 1 7. ltJ d4 ? ixg2
I S .:B:hg l ( 1 S . ltJxd7 ixh l -+) I S . . . ltJxe5!
1 9.1tJh5 ltJ g6-+
White should have opted for a policy of damage
limitation with 1 7.ltJxd7 when he should be
able to live with his small disadvantage.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
The Prime Years
288
17 i.xf3!
Karpov seizes the tactical opportunity to
take over the initiative.
In his analysis Karpov showed that 25 .. .\Wb5t
26.Wb3 Wa5 is ineffective due 27.Wc3! - but
not 27.Ei:e3? when 27 . . . Ei:4d5 ! wins.
18.i.xf3 �xe5!
26.Wxc5 Ei:xc5 27.Ei:e2
This allows White to obtain two pieces for
a rook, but Karpov has correctly evaluated the
resulting position as being in his favour.
The English grandmaster stops Black's rook
from invading on the second rank, but perhaps
he could have resisted more stubbornly with
27.ttJf4 ! ? Ei:d2 28.i.d l (28 .i.e4) 28 .. .Eld4,
although of course Black has decent winning
chances here too.
•.•
19JWe4 c6 20Jhd7
White has to take the knight, otherwise he
will be a pawn down for nothing.
27 Ei:dl !
.•.
20 ... ,bb2t! 21 .<ti>xb2 Ei:xd7
The dust has settled. Black has not only
picked up a second extra pawn, but he has also
exposed White's king in the process.
Curiously in this position the black rooks
can do more damage on the first rank than the
second. Karpov has such a fine sense of when
to go after the opponent's king in endgames.
22.�e3 Ei:hd8
28.lLlf4
Karpov sensibly brings the other rook into
play. It was too early to go after the king with
22 . . . Wb4t 23 .Wb3 Wd4t 24.Wc3 when White
easily parries the direct threats.
Swapping the c-pawn for the e-pawn weakens
White's king too much: 28.Ei:xe6 Ei:d2 29.Eie3
Ei:dxc2t 30.<j;Jb3 (30.�b l Ei:2c3-+)
23.a3
8
7
6
5
4
a
3
2
1
a
b
c
d
e
f
g
h
23 ... Ei:d4!
Karpov brings his pieces closer to the king.
24.g3 Wc5 25J�el Ei:c4!
Forcing a favourable endgame.
b
c
d
e
f
g
h
30 . . . a5 !! Out of the blue, Black can create
decisive threats against White's king. 3 1 .ttJ f6
(3 1 .ttJ g7 a4t 32.�xa4 Ei:b2 catches the
king) 3 1 . . .a4t! 32.<j;Jxa4 Ei:b2! 33.i.g4t
(33.Eie8t �c7 34.Ei:e7t <j;Jd6 3 5 . ttJ e4t �xe7
36.ct:lxc5 <j;Jd6 37.ct:ld3 b5 t 38 .<j;Ja5 Ei:b3-+)
33 . . . �c7 34.Ei:e7t <j;Jd6 3 5 .Ei:e4 Ei:c3! (But not
35 . . . b6? 36.ct:le8t when White can draws via
perpetual check.) 36.h5 b6 37.ct:le8t �c5
Black wins.
1994
289
Michael Adams - Anatoly Karpov
28 .. J:�b5t 29.c;ta2
a
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
29 .. J:�bbl !
35 ... a5!
When i t comes to endgames, Karpov is
second to none in the history of chess. In this
position he spots a weakness on a3 and goes
after it immediately.
The a-pawn is a powerful force when all the
white pieces are far from it.
36.grz ge3!
3o.ih5?
Karpov throws the centralized bishop off
balance.
Adams goes after his opponent's pawns, but
the queens ide is what really matters.
37.gf'7t c;tb8!
30.tt:J d3 was also inadequate, as after 30 . . . E1:al t
3 1 .'it>b3 E1:db l t 32.tLl b2 E1:g1 33 .�g2 E1:ae 1
Black will win another pawn.
White could have offered sterner resistance
with: 30.E1:e3! E1: e l 3 1 . tLl e2 (3 1 .�e2 a5 ! 32.c3
a4 33.tLld3 E1:h l 34.�f3 E1:al t 3 5 .'it>b2 E1:hb l t
36.'it>c2 E1:f1 37.'it>b2 'it>d8 Black has decent
winning chances.) 3 1 . . .E1:al t 32.'it>b3 E1:eb l t
33.'it>c4 E1:a2 34.�e4 White is clearly worse,
but not necessarily losing.
30 .. J::� a l t 31 .c;tb3 gdb l t 32.c;tc4 gxa3
Mission accomplished.
33.ixf'7 gxg3 34.i.xe6t
After 34.tLlxe6 a5 3 5 .E1:d2 and Black can
match easily White's threats with 35 . . . E1:b4t or
35 . . . 'it>b8.
After 37 . . . 'it>b6 White would get some
counterplay with 38 .�c8 .
38.ge7 ge4t 39.c;tc3
If 39.'it>c5 E1:b5t 40.'it>d6 E1:d4t the king is
caught.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
290
The Prime Years
39 J::g b 5!
0-1
..
Karpov does not even bother to take the
h-pawn but instead restricts the enemy knight
and prepares to push the a-pawn. It impressed
Adams so much that he resigned.
This win gave Karpov a final score of 6/9,
which was enough to finish in second place,
half a point behind Kamsky and half a point
ahead of Topalov and Lautier.
Karpov's next tournament was Dortmund,
where coincidentally his first opponent was
Adams, who this time opted for 1 .e4. Karpov
obtained a solid position with the Caro-Kann
but failed to equalize fully, and the Englishman
gradually increased his advantage and took his
revenge for the previously analysed game.
In Round 2 Karpov played very dryly against
Leko, but managed to outplay his young
opponent. Karpov drew solidly against Epishin
with the Queen's Indian, then he defeated Lutz.
Part of the game can be found in the note to
Black's eleventh move in Game 34.
In Round 5 Karpov chose the Hubner
Variation against Yusupov's Rubinstein Nimzo­
Indian. A long manoeuvring game ensued, but
neither player was able to obtain the upper
hand and a draw was eventually agreed. In
the next game Karpov had the black pieces
again, and was able to take a quick draw with
Timman.
Karpov's next game against Korchnoi was
a remarkable one. He chose the Petrosian
Variation against the Queen's Indian, and
a complex middlegame ensued, with the
position mostly blocked. Gradually Korchnoi
took over, and later he sacrificed a pawn
for the initiative. Karpov looked to be in
trouble, but found some hidden resources
and the position became wildly complicated.
After some inaccuracies on both sides, it was
Korchnoi who finally showed his class and won
brilliantly, after allowing Karpov to promote to
a second queen. (Ironically it was the losing
mistake - had Karpov promoted to a knight he
would have been able to draw!) The win must
have meant a lot to Korchnoi.
Unlike the previous tournament, Karpov
was unable to finish strongly in the late
rounds, and his final two games resulted in
quick draws with Dreev and Piket. Perhaps the
titanic struggle against Korchnoi took most of
his energy away. And more generally, it looks
as though Karpov simply played in too many
tournaments; Dortmund was his ninth event of
the year (including rapid tournaments) , which
is a lot for any professional player, especially
one in his forties. Comparing the quality
of his play at the Linares and Dortmund
tournaments, the difference is enormous.
***
The next event of Karpov's calendar was the
Tilburg elimination tournament. His first
opponent was Kurajica. Karpov was unable to
get a serious advantage in either of the slow­
play games, so they went to a rapid and blitz
playoff. This too was competitive, but Karpov
finally prevailed after winning two games,
drawing two and losing one.
In the next round Karpov started with the
black pieces against Antunes of Portugal,
who played a Catalan. Karpov undertook
tripled pawns on the c-file and managed to
keep one of them as a passed pawn. Despite
a few inaccuracies he was able to convert his
advantage. The return game was a Semi-Slav.
Karpov was a bit better for the whole game, but
at one moment he erred and allowed an out­
of-the-blue tactical shot. Fortunately Antunes
missed his chance and Karpov converted his
advantage smoothly.
In the next round Karpov faced Shabalov. An
unusual line of the Semi-Slav soon resulted in
an IQP position. Karpov sacrificed his d-pawn
and got into trouble, but managed to get the
1994
better of the complications and later won.
The return game was a sharp Caro-Kann with
castling on opposite sides. Shabalov threw all
his forces into the attack, but Karpov kept his
cool and refuted his opponent's play to secure
a 2-0 victory.
Karpov's next opponent was Kiril Georgiev,
who played a 4.Wc2 Nimzo-Indian creatively
and sacrificed a piece for two pawns and
attacking chances. Nevertheless it was not
fully sound, and Karpov defended skilfully
and won. In the second game Georgiev
somewhat surprisingly played for a win with
the Tartakower System against the Queen's
Gambit. A tense middlegame ensued, but
Karpov got the upper hand and later capitalized
with a wonderful double piece sacrifice, which
led to a crushing victory.
Karpov's opponent in the final was Evgeny
Bareev, who in the first game chose 4.e3 against
the Nimzo-Indian, and got some advantage in
the position with two bishops and hanging
pawns. Karpov kept himself in the game, but
later he made an uncharacteristic endgame
mistake which led to his defeat. In the second
game Karpov opened with 1 .e4 and played
his favourite Tarrasch Variation against the
French. For a while it looked like Black was
under pressure in the IQP position, but Bareev
eventually managed to hold an inferior rook
endgame, thus securing overall victory.
Despite the disappointing finish, Karpov's
overall performance in Tilburg was a good
one. In the slow-play games he scored a total
of six wins, three draws and one loss.
***
Karpov played in a few more events to round
off an extremely busy year. He took part in a
Sicilian-themed tournament in Buenos Aires,
but in view of the unusual terms of play I will
not discuss it in detail. Karpov scored three
wins, nine draws and two losses to finish on
29 1
6Y2/ 1 4, which was enough for fifth place out
of a field of eight.
Karpov also played a short two-game match
against the Argentine grandmaster Campora.
In the first game he made a solid draw with
the Caro-Kann, and in the second he got
some advantage with a minority attack in the
Carlsbad structure, and won without too many
problems.
Karpov's final event of 1 994 was a six-game
match against Laurier in Ubeda, Spain. In
the first game Karpov equalized against the
Catalan, and the game never looked like being
anything other than a draw. In Game 2 the
Frenchman ventured the Schliemann Variation
against the Ruy Lopez. He got a reasonable
position but made a mistake in the endgame,
and Karpov was able to simplifY to a favourable
rook endgame which he converted to a win.
In the third game Lautier got some advantage
with the Catalan, but failed to capitalize and
Karpov managed to secure a draw.
In Game 4 Karpov played a timid opening,
but Lautier reacted strangely and soon got a
worse position. Karpov later won a pawn but
was unable to convert it. Game 5 was another
Catalan. Karpov tried a fresh approach and
was able to generate some pressure on the
queenside. Lautier had chances to hold the
endgame, but eventually he succumbed. In the
final game Karpov switched to 1 .e4 and got
a slight plus against the Sveshnikov, but was
unable to generate serious winning chances
in the middlegame with opposite-coloured
bishops, and the game ended in a draw. Thus
Karpov won the match 4-2.
292
1994 Summary
Bundesliga 1 993-94: 3/4 (+2 =2 -0)
Match versus Morovic Fernandez, Las Palmas: Won 5-1 (+4 =2 -0)
Linares ( l st place) : 1 1 1 1 3 (+9 =4 -0)
Dos Hermanas (2nd place) : 6/9 (+4 =4 - 1 )
Las Palmas (2nd place) : 6/9 (+4 =4 - 1 )
Dortmund (5th-7th place) : 4Y2/9 (+2 = 5 -2)
Tilburg knockout (losing finalist) : 7l1z1 l O (+6 =3 - 1 ) (regular games)
Match versus Campora, San Nicolas: Won 1 Y2-Y2 (+ 1 = 1 -0)
Match versus Lautier, Ubeda: Won 4-2 (+2 =4 -0)
Total 7 1 .3% (+34 =29 -5)
II Wins • Draws • Losses
199 5
Rating 2765 (2 in the world)
In 1 995 Karpov played in numerous rapid tournaments but fewer team events. In the Spanish
league he drew three games against Speelman, Franco Ocampos and Cramling. According to
the database he played one game in the final of the EU Cup against Boris Alterman. The Israeli
grandmaster only played a few games against the world champions, but has an unusually good
record with two wins, four draws and j ust a single loss. The following is his only game against
Karpov.
I Game 38 I
Boris Alterman - Anatoly Karpov
EU
Cup final, Tynisre
1 995
l.d4 tlJf6 2.c4 e6 3.tlJa b6 4.g3 Aa6 5.b3 b5
This line, which was invented by the Polish international master Dobosz, was an occasional
surprise weapon for Karpov. He only played it three times, scoring one win and two draws. Many
other Queen's Indian experts use it as a secondary system.
6.cxb5 hb5 7.i.g2
Th ree years later 011 played differently but got no advantage: 7.'tJc3 ib4 8 .id2 ic6 9 .ig2
'!!N c8 1 0.0-0 ixc3 1 1 .ixc3 Wfb7 1 2.Ekl 0-0 1 3 .E\e l �c8 1 4.Wfc2 h6 1 5 .Wfb2 id5 1 6.'tJh4 ixg2
17.'tJxg2 d6 1 8 .f3 'tJ bd7 1 9 .e4 c5 20.'tJe3 cxd4 2 1 .ixd4 �xc l 22.�xc l e5 23 .ic3 d5 Black has
equalized and the players soon agreed a draw, 011 - Karpov, Polanica Zdroj 1 998.
7,..i.c6
Karpov deviates from a previous encounter with Epishin in which he played 7 . . . a5 . In that
game both players were too ambitious.
Karpov encountered 7 . . . ib4t twice against Christiansen. He won both games, although on both
occasions the American grandmaster was able to equalize in the opening.
By the way 7 . . . d5 is the most common move in the position.
8.0-0 i.e7
8 . . . Wfc8 and 8 . . . a5 are also played. Karpov prefers to castle first and only then decide how to
play on the queenside.
294
The Prime Years
9.�g5 0-0 10.i.xf6
Alterman decides to exchange his bishop in
order to facilitate a central advance.
In the event of 1 O. tZl c3 Karpov might have
responded with 1 0 . . . a5 1 1 .'lWc2 tZl a6!? 1 2.e4
h6 1 3 .ixf6 ixf6.
10 ...i.xf6 l I .ltlc3 d5
Karpov does not want his opponent to get
an ideal pawn centre.
12.�c2
The immediate pawn advance would not
be too threatening: 1 2.e4 dxe4 1 3 .tZle5 ixe5
1 4.dxe5 e3! Black has good chances, as pointed
out by Karpov.
12 ... a5 13.e4
16J3acl E:d6 17.E:c4
1 3 .l:l:fc 1 was possible, although the e4advance will now be harder to carry out as
the d4-pawn may hang. 1 3 . . . tZl a6!? ( 1 3 . . . tZl d7
1 4.e4) 1 4.a3 ib7 ( 1 4 . . . l:l:bB 1 5 .e3 'lWd6 1 6.ifl
l:l:b6 1 7. tZl a4 White has the initiative.) 1 5 . tZl a4
l:l:cB 1 6.b4 axb4 1 7.axb4 c6 White is j ust
fractionally better.
Alterman probably visualized the ensuing
endgame and evaluated it as completely
drawn. Instead he could still have played for an
advantage. My advice in such situations is to
play the position and not the opponent. Here
this strategy would have yielded at least a draw
for the Israeli grandmaster.
13 ... dxe4
1 7.l:l:fd l
With this simple move White holds the
position in the centre and forces Black
to play accurately to avoid an unpleasant
squeezing.
1 7 . . . tZl a6
Alternatives do not equalize:
1 7 . . . c6 I B .tZle5! (stopping . . . 'lWb6) I B . . . ixe5
( 1 B . . . 'lWb6?? 1 9. tZl c4) 1 9.'lWxe5 'lWc7 20.l:l:c4
l:l:fdB 2 1 .l:l:dc 1 (2 1 .l:l:a4 c5) 2 1 . . .'lWa7 22.l:l:a4
l:l:xd4 23 .'lWxa5 White's bishop is stronger
than Black's knight.
17 . . . tZl d7 I B .ifl g6 1 9 .ib5 The c7-pawn is
a headache for Black. The pain is not sharp,
but it may last a long time.
I B .ifl !
After 1 8 .tZle5 c5! 1 9. tZl c6 'lWb6 20.dxc5
Karpov had a choice between improving
his rook or his knight. He goes for the
more powerful piece, by means of a creative
manoeuvre that will be revealed shortly.
The alternative was: 1 3 . . . tZl a6 1 4.l:l:fd l ! ( 1 4.a3
dxe4 1 5 .tZlxe4 ixe4 1 6.'lWxe4 l:l:bB Black has
counterplay according to Karpov; 1 4 . tZl xd5
exd5 1 5 .'lWxc6 tZl b4 1 6.'lWc3 [ 1 6.'lWc5 dxe4
1 7.tZle5 'lWd6=l 1 6 . . . dxe4 1 7.tZle5 c5! I B .'lWxc5
l:l:eB 1 9 .1:l:fd l l:l:cB Black has no problems thanks
to his fluent piece play.) 1 4 . . . tZl b4 1 5 .'lWe2 dxe4
1 6.tZle5 id5 1 7.tZlxe4 Black is a bit worse due
to his backward c-pawn.
14.ltlxe4 .L:e4 15.�xe4
1 99 5
295
Boris Alterman - Anatoly Karpov
l'!xd 1 t 2 1 .l'!xd 1 �xc5 22.�c4 �xc4 23.bxc4
tiJ b4 Black frees his position.
18 ... tiJb4 1 9 .a3 tiJd5 20.l'!c5
Black has still not equalized fully.
17...c6 18 . .!iJe5
1 8 .�f4 prevents Black from deploying
his queen on b6, but there is another good
plan available: 1 8 . . . tiJ a6! Black has time to
improve his knight. 1 9.tiJe5 ( l 9 .l'!fc l tiJ b4;
1 9.a3 tiJ c7 20.tiJe5 ttJd5 2 1 .�e4 �b6 Black
is comfortable.) 1 9 . . . tiJ b4 20.a3 ttJ d5 2 1 .�f3
'iWb6 Black has a pleasant position.
Karpov tries to keep as much material on
the board as possible. Another option was
2 1 . . .l'!xc6 22.l'!xc6 �xd4 23 .�xd4 l'!xd4
24.l'!c8t i.d8! when Black also succeeds in
keeping a pair of rooks on the board.
22.Wlxd4
22.l'!xd6 �xd6 is evaluated by Karpov as
clearly better for Black, but this seems too
optimistic: 23 .l'!c6 Exchanging the rooks
should lead to a draw even if Black wins the
a2-pawn. (23.i.f3 �e7) 23 . . . �d l t 24.i.fl g6
2 5 .�c2 White should be able to hold.
22 ... �xd4
This endgame should of course be drawn,
but Black has a couple of advantages as his
king is safer and his pawn majority is more
mobile. It is remarkable that Karpov was able
to make these subtle factors count against a
strong grandmaster.
8
7
a
b
c
d
e
f
g
h
18 ...�b6!
Karpov improves his queen and prepares to
bring the other rook into play.
19.1'!fc1
I f 1 9 .1'!dl l'!fd8 20.l'!d3 i.xe5 2 1 .�xe5 ttJ a6
Black is comfortable.
Switching to the a5-pawn with 1 9.1'!a4!? may
have been best: 1 9 . . . �a7 ( l 9 . . . l'!d5 20.�e3
l"IbS 2 1 .ttJg4 i.e7 22.l'!c4 White is fractionally
better) 20Jkl l'!fd8 2 1 .�e l (2 1 .l'!c5 l'!d5)
2 l . . .l'!xd4 22.l'!xaS �c7 23.ttJxc6 ttJxc6
24.i.xc6 �d6 It is unlikely that White can do
much with his extra pawn, but he is certainly
not in any danger.
19 ... l'!fd8 20 . .!iJxc6 .!iJxc6 2 1 .�xc6 Wlxd4
6
5
4
3
2
a
b
c
d
e
f
g
h
23.� lc2
The solid approach should be enough to
hold the position, although active play would
have been more principled:
23.l'!a6!
Eliminating Black's last queenside pawn
is a useful insurance policy. Even if White
were to lose both his queens ide pawns, the
resulting endgame with four pawns versus
296
The Prime Years
three on the kingside will usually be drawn not that White has to resort to that.
23 . . . :gd2 24.:gxa5 i.d4
Karpov stops here, mentioning that Black
has compensation. He is right, but White
should certainly still be able to hold the
position.
2 5 .:gn
Also after 2 5 .:gaB!? i.xf2t 26.W h l :gxaB
27.i.xaB h5 2B.a4 White should not be
worse.
25 . . . :gcB 26.i.b7!
The bishop stops Black from doubling his
rooks on the second rank.
26 . . . :gbB 27.i.a6
White is safe.
It was also worth considering 23.h4!? in order
to exchange a pair of pawns in the event that
Black decides to expand on the kingside.
23 J�dl t 24.i.f1 g6
.•
Black could have considered 24 . . . g5 , but I
think it is a little premature here as in some
lines White may be able to transfer his bishop
to h5 and attack the f7 -pawn.
Karpov transfers the bishop to a spot from
which it will securely defend Black's a-pawn.
26.:gc7 i.b4
The bishop also covers the d2-square. It can
become important.
27J3cS
There is nothing wrong with this move, but
once again Alterman shows that he is only
looking for a draw.
27.:gb7!? Doubling the rooks on the seventh
rank often accomplishes a lot. In the present
position it would have led to an interesting
fight. 27 . . . :gBd5 (27 . . . :g 1 d5 2B.:gcc7 :gf5 29.f4
[There is also 29.:gd7 :gxd7 30.:gxd7 when
compared with the game White's rook is more
favourably placed.] 29 . . . g5 30.rj;Jf3 gxf4 3 1 .g4
:gf6 32.i.c4 White has nice compensation for
the pawn.) 2B.:gcc7 :gf5 29.f4 :gd2t (29 ... g5
30.i.c4 gxf4 3 1 .:gxf7) 30. rj;Jf3 i.c5 3 1 .i.b5
WfB (3 1 . . .:gxh2?? 32.i.eB White invades
decisively.) 32.:gcBt rj;Jg7 33.:gcc7 White has
enough play on the seventh rank.
27 .. J3xcS 2SJ3xcSt c;f;g7 29.13c2 c;f;f6
25.c;f;g2
Karpov improves his king before he starts
pushing his kingside pawns.
25 .h4 can be played here as well.
8
29 . . . f5 ?! would have been premature: 30-Ei:c7t!
rj;Jf6 (30 . . . rj;Jh6 3 1 .i.c4 :gel [3 1 . . .e5!?] 32.f4
:gc2t 33.Wf3 It is hard for Black to make
further progress.) 3 1 .:gxh7 :gd2 32J�c7!
White prevents . . . i.c5 and no longer has any
problems.
7
6
5
4
29 . . . :gd7 stops the white rook from invading
the seventh rank, but after 30.i.b 5 ! the rook
must either relinquish its defensive duty or
vacate the open d-file.
3
2
1
a
25 ...i.e7!
b
c
d
e
f
g
h
In his excellent Learn from the Legends book,
Mihail Marin recommends 29 . . . g5 ! ? intending
297
Boris Alterman - Anatoly Karpov
1 99 5
... <;t>g6 and . . . h 5 . The Romanian grandmaster
even evaluates this plan as an improvement
over Karpov's play. He mentions that the
attempt to block the kingside with g4 would
give Black's bishop excellent prospects on the
dark squares.
30.ie2
Alterman forces Karpov to decide whether
to keep his rook on the d-file or on the first
rank.
30 .. J�d7
a
b
c
d
e
f
g
h
3 1 .h4
Another idea was 3 1 .�b 5! ? when the black
rook has to decide whether to vacate the
seventh rank or the d-file.
Alterman's move is also sensible. The Israeli
grandmaster knows that the h-pawn might
eventually become a target on its starting
square, so he prepares to exchange it for an
advancing enemy pawn. Alterman is a strong
and gifted player who even went on to work
for Kasparov towards the end of the nineties,
so I can imagine that he had a good chess
education and may well have been familiar
with the following Karpov masterpiece from
1 993 against Beliavsky.
a
b
c
d
e
f
g
h
29.ctJe4?!
When one's position has a minor defect,
one should think carefully about exchanging
pieces as the significance of that defect may
grow. In the present position White has
an inferior bishop and a less mobile pawn
majoriry.
Another example of this theme can be found
in Karpov's win over Browne in San Antonio
1 972. In that game Browne had doubled
pawns, and Karpov kept exchanging pieces
until he was left with a winning knight
endgame.
29 . . . ctJxe4! 30.E!:xe4 <;t>f8 3 1 .<;t>f1 <;t>e7 32.E!:e l
According to Jacob Aagaard, the more active
placement of the rook on c4 was stronger, but
White must take care to time it correctly.
The immediate 32.E!:c4?! runs into 32 . . . E!:d8!
when White has nothing better than
returning with his rook to e4, as after 33.b4?
�xf2! Black wins a pawn.
Instead White can improve the rook more
patiently: 32. <;t>e2 f5 (32 . . . E!:d8 33 .�b3)
33.E!:c4 E!:d8 According to Aagaard and
Marin White can hold the position with
34.f3 or 34.f4 g5 3 5 .b4.
Finally, White can also think of solving the
problem of the h-pawn with 32.h4!? and after
32 . . . E!:d8 33 .�b3 White should be okay.
32 . . . �d6
Karpov's sryle in the endgame is to begin by
improving his pieces, and then focus on his
king and pawns.
The Prime Years
298
32 . . . g5 ! ? was an interesting alternative.
33.l'l:d1
33.h4! Beliavsky should have pushed his
h-pawn to the fourth rank now or on the
next move, to prevent it from becoming a
target.
33 . . . l'l:c5! 34.id3 ?!
a
b
c
d
e
f
g
h
34 . . . g5 !
Karpov gains space and fixes the h2-pawn as
a potential weakness.
3 5 .ib5 h5 36.mg2 l'l:f5 37.id3 l'l:e5 38.mf1
Marin's 38.h3!? was worth considering.
a
b
c
d
e
f
g
42 . . . l'l:h5 43.l'l:e4?
A losing blunder. Beliavsky must have
overlooked Black's clever reply.
According to Marin White could still have
saved the game with the accurate 43.l'l:hl!
l'l:xh4 44.ib5! f5 (44 ... l'l:h3 45 .ie2 f5 46.f3)
45 .h3!, exchanging his weakness.
43 . . . l'l:xh4 44.mf1
44.f4 gxf3t 45 .'kt>xf3 l'l:h3t wins.
h
3 8 . . . g4!
Fixing White's kingside pawns.
39.l'l:e 1 l'l:d5 40.ic4 l'l:g5 4 l .mg2 h4! 42.gxh4
Beliavsky decides not to allow the pawn
to get all the way to h3. It is hard to say if
White would have been able to hold that
endgame, but it certainly would have been
unpleasant.
a
b
c
d
e
f
g
h
44 . . . e5!
Thanks to this clever move Black can defend
his g4-pawn before taking on h2.
45 .id5 f5 46.l'l:c4 l'l:xh2
Black has won a pawn for no compensation,
and he converted his advantage without too
many problems:
47.b4 e4 48.l'l:c6 g3 49.fxg3 e3 50.ic4 E1m
5 l .m e 1 axb4 52.cxb4 ixb4t 53.md1 l'l:d2t
54. mel 'kt>d7 5 5 .l'l:xb6 ia3t 56.'kt>b l e2
57.ixe2 l'l:xe2 5 8 .a5 id6 59.a6 l'l:e5 60.l!?c2
l'l:a5 6 l .l'l:b7t 'kt>e6
0- 1 Beliavsky - Karpov, Tilburg ( 1 ) 1 993.
Before we return to the main game, I would
like to show another instructive example to
highlight the significance of a weak rook's
pawn. Ironically the winner was none other
than Beliavsky himself! Having won a game
in this fashion, it is rather surprising that he
allowed his h2-pawn to become fixed in the
above game with Karpov. The following game
took place in 1 990.
1 99 5
Boris Alterman - Anatoly Karpov
299
have a securely defended pawn so close to its
promotion square.
44.�d2 �c7 4 5 .�e3 �a5 46.�f4 �e 1 47.�e3
a
b
c
d
e
f
g
h
34.lt>e2?!
White should have preferred 34.h3 intending
g4.
34 . . . g5 3 5.lt>d3?!
Once again 3 5 .h3 was safer.
35 ... g4!
Beliavsky fixes White's kingside pawns on
the same-coloured squares as his bishop - a
serious accomplishment.
36.lt>c4 It>e7 37.�d2
Either here or on the following move White
should have played 37.d5 !. It may or may
not be good enough for a draw, but he
simply had to try it.
37 . . . lt>d7 38.�c l It>c6
From here I have not been able find a defence
for White; zugzwang will be his downfall.
39.�e3 �c7 40.�d2 �d8 4 1 .�b4 f5 42.�d2
h4 43.�f4
a
b
c
d
e
f
g
h
43 ... h3!
In endgames it is tremendously useful to
a
b
c
d
e
f
g
h
47 . . . lt>d6! 48 .�f4t �d7! 49.�e3 �c6!
Thanks to the triangulation White IS In
zugzwang and must give way with his king.
50.lt>d3 �d5 5 1 .�e2 �c3 52.�d3 �b2
White finds himself in another zugzwang
and must shed material. The rest is easy.
53.f4 gxf3 54.�f2 �c l 5 5 .�c2 �g5 56.lt>d3
�f6 57.�e3 �d8 5 8 .�d2 f2 59.�e2 It>xd4
60.�e3t It>e4 6 1 .�xf2 �c7
0- 1 Pritchett - Beliavsky, Novi Sad (01)
1 990.
Back to the game.
a
b
c
d
e
f
g
h
31...e5
Karpov sets his kingside pawns in motion.
300
The Prime Years
32.h5?!
When one is almmg to draw a slightly
worse ending, exchanging pawns is usually a
good strategy. Nevertheless the text move is a
mistake, as White misses a chance to improve
his position, while also giving up control over
the g5-square.
It is possible that Alterman had seen the
Beliavsky - Karpov game and was content in
the knowledge that he had played h4. However,
he may not have appreciated the fact that his
queenside is weaker than Beliavsky's was.
Marin shows the correct path for White, which
involves active defence: 32J''l c 6t! <jrf5 33.g4t!
<jre4 34.Ek4t <jrd5 3 5 .i.f3t �e6 36.Ek6t
�e7 37.g5 ! White equalizes as the f7-pawn has
become a fixed target.
Apart from the above line, 32.i.b5!? was also
an improvement over the game continuation.
32 ...�g5 33.hxg6 hxg6 34. �f1
Alterman decides to sit and wait.
34.i.b5!?
Forcing the decision is not a bad idea here
either. In some lines the bishop might also be
used for attacking Black's kingside pawns.
34 .. .1'h7
34 . . . l'!d6 35 .i.e8 f6 (35 . . . f5 36.l'!c6 White
exchanges rooks and draws immediately.)
36.�f3 (36.f3 l'!d8 37.i.b5 is also fine.)
36 . . . e4t 37.<jre2 Black has little chance of
making progress.
35 .l'!e2
White can also wait with 3 5 .<jrf3 , for instance
35 .. .f5 36.l'!c6 e4t 37.�e2 l'!h7 38 .i.c4 l'!h l
39.f4t exf3t 40.�xf3 and White holds the
position easily.
35 . . . f6 36.l'!e4 l'!c7
36 . . . l'!e7 37.f4t draws easily.
37.f4t exf4 38.gxf4t �h6 39 .i.c4
White has freed his position and has no
problems.
34.. JM6
Karpov defends the g6-pawn in advance.
35.�g2
Once again, activating the bishop with
3 5 .i.b5! would have made it harder for Black
to progress: 35 . . . f5 36.i.e8 (With his bishop
away from the passive e2-square, White should
also be able to draw with 36.f3!? e4 37.�e2.)
36 . . . l'!e6 Black must avoid a rook exchange.
37.l'!c8 (37.i.d7 l'!e7) 37 . . . e4 38.�e2 �g4
(38 . . . l'!d6 39.l'!c6) 39 .l'!d8 I do not see how
Black can improve his position.
35 ...£5
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
36.�f1?!
Alterman continues to wait passively, but
it was time to take countermeasures against
Black's advancing pawns.
36.f3!
This improvement was suggested by Marin.
36 . . . l'!d5
This is Marin's idea, intending to shuffle the
rook to d4 before advancing with . . . e4.
The immediate 36 . . . e4 seems threatening,
but in fact it only relieves the pressure:
37.fxe4 fxe4 38.<jrf2 l'!d5 39.i.a6 i.c5t
40.�e2 l'!f5 (40 . . . e3 4 1 .i.b7 l'!f5 42.1f3
Black cannot invade.) 4 1 .l'!c4 l'!f2t 42.�e l
i.b6 43.l'!xe4 White can easily hold.
30 1
Boris Alterman - Anatoly Karpov
1995
37.ic4 E!:d4
37... e4 38.<'!?f1 <.!?e5
Karpov instructively centralizes his king.
a
b
c
d
e
f
g
h
38.if7!?
Marin does not consider this move, but it
seems to me that targeting the g6-pawn is
White's most convincing drawing plan.
3 8.ie2 e4 39.fxe4 E!:xe4 (39 . . . fxe4 40.E!:c4)
4o.if3 According to Marin the weakness
on g3 gives a stable advantage for Black;
nevertheless I think White should be able to
live with it.
3 8 .@f2 e4 39.@e3 ic5 Marin stops here,
evaluating the position as unpleasant for
White. He is right, but the position is still
tenable: 40.f4t!? (40.\t>e2 exf3t 4 1 .@xf3 is
also possible) 40 . . . @f6 4 1 .@e2 g5 (4 1 . . .id6
42.E!:d2) 42.ia6 id6 43.E!:c6 gxf4 44.gxf4
We7 45.@e3 E!:d 1 46.ic8 White can hold
the position.
38 . . . e4
3 8 . . . @f6 39.ie8 e4 40.fxe4 fxe4 4 1 .\t>f2
White is out of danger.
3 8 . . . id6 39.@f2 e4 40.fxe4 E!:xe4 4 1 .E!:c6
E1d4 42.@e3 E!:d1 43.\t>e2 White forces a
repetition.
39.fxe4 fxe4 40.ie8 E!:d3 4 1 .ic6 @f5 42.E!:e2
Eld4 43.ib7
Black is tied to the defence of the e4-pawn,
so White should have little trouble drawing.
a
39 ... g5!
b
c
d
e
f
g
h
Karpov prepares a further pawn advance.
40.<.!?f1 E:h6 41.<.!?g2 E!:d6 42.<.!?f1 E:d8
43. <.!?g2 f4
After tiring his opponent with a few shuffling
rook moves, Karpov continues with his main
plan. At the same time he creates a threat
which forces Alterman to abandon his policy
of waiting.
44.0
36 ...\t>f6 37.\t>g2?
This was White's last chance to play 37.f3!
under good conditions.
a
b
c
d
e
f
g
h
The Prime Years
302
47.i.dl
44... e3!
Karpov creates a mighty protected passed
pawn. It is an obvious idea of course, but the win
is not trivial as White is covering all the entry
squares. It is likely that Karpov had already seen
the winning method he uses in the game.
The bishops has very few squares available, a
fact which enables Black to win by zugzwang.
47".i>d4 48.i>f2 i>c3 49.i>e2 i>b2!
AI; Marin pointed out, the black bishop has
to remain on the b4-square. If 49 . . . .ta3? 50.b4!
White liberates his bishop and obtains a draw.
45.g4
Exchanging pawns and winning a tempo
with 45.gxf4t makes no real difference:
45 . . . gxf4 46.c;:t>fl l:'!:d2 47.l:'!:c8 (After 47.l:'!:xd2
exd2 Black wins j ust as in the game.) 47 . . . l:'!:xa2
48.l:'!:e8t 'it>d4 49.l:'!:e4t c;:t>c3 50.l:'!:xf4 l:'!:a 1 t
5 1 . 'it>g2 'it>d2! The e-pawn is unstoppable, as
pointed out by Marin.
50.i>d3
50.a4 c;:t>c l is an immediate zugzwang, and
50.a3 also loses quickly after 50 . . . c;:t>xa3 (or
50 . . . c;:t>c l 5 1 .a4 .tc5 52.b4 axb4-+) 5 1 .'it>d3
c;:t>b2 and Black wins.
45"J:�dl!
The rook exchange will lead to Black having
a securely-defended pawn just one square
away from promotion. Pure opposite-coloured
bishop endgames are notoriously drawish, and
are frequently drawn even when one side has
one or more extra pawns. In the present case
Black does not even have an extra pawn, yet
he is winning as his positional advantage is so
great.
a
46J:�xdl exdl
It is possible that Alterman overlooked the
fact that the e-pawn would turn into a d-pawn,
or at least underestimated its consequences.
50".i>bl !
b
c
d
e
f
g
h
Karpov sets u p a zugzwang, which brings his
opponent down almost instantly. 50 . . . 'it>xa2?
5 1 .c;:t>c2 and 50 . . . c;:t>c l ?! 5 1 .c;:t>e2 .te7? 52.h4!
both allow White to escape with a draw.
5 1.a3
After 5 1 .a4 'it>b2 White's right to move
once again becomes a desperately painful
obligation.
5 1".i>c1
a
b
c
d
e
f
g
h
Karpov takes the pragmatic option. If, dear
reader, you happen to reach the same position
against somebody whom you strongly dislike
and wish to obtain your victory in the most
Boris Alterman - Anatoly Karpov
1995
humiliating way possible, you might prefer the
flashy 5 1 . . . 'it>b2!? 52.axb4 (52. 'it>e2 'it>xa3-+)
52 . .'.!j>cl ! 53. 'kt>e2 axb4 when the zugzwang is
fatal for White, despite his extra piece.
303
59.ie6 'it>a3 60.if5 'kt>a2 6 1 .ie6 and White
continues to resist.
.
56.ic6
56.'kt>xd2 'kt>a2 is equally hopeless for White.
52.'kt>e2 ha3 53.b4
Alterman keeps hoping for a miracle.
53 ... axb4!
Black does best to give himself a b-pawn,
as his king will be able to walk around it and
support it from the a-file. I noticed a similar
theme in Game 4 of Karpov's rapid match
against Macieja from Warsaw 2003, in which
the Polish grandmaster missed a drawing
opportunity based on giving Karpov an a-pawn
instead of a b-pawn.
54.ia4
It is useless for the bishop to stay on the
other diagonal, as the d2-pawn becomes too
strong: 54.ib3 'it>b2 55 .id5 'it>c2-+
54... 'i!?b2 55. 'i!?dl
If 55.'kt>xd2 b3! wins.
a
55 ... b3!
b
c
d
e
f
g
h
Karpov gives his opponent no chance,
and simply wins the bishop. There was no
sense in 55 . . . 'it>c3 56.ic2 b3? 57.ie4 'kt>b2
(57 . b2 58.ic2 ib4 59.ib l ) 58.id5 ib4
. .
56 ... 'i!?al
0-1
In 1 995 Karpov took part in an elimination
tournament in Baden-Baden. In the first
round he equalized and drew on the black side
of a Reti against Hick!. In the return game
Karpov got the advantage of the two bishops
against the Wade Defence, then increased his
advantage and won smoothly.
In the next round Karpov started with the
white pieces against Vaisser, who played a
Dutch Stonewall. Karpov got a slight plus,
but Vaisser equalized and later took over the
initiative. At one point he missed a win, then
he over-pressed and Karpov finally took control
and won. In the return game Karpov played
the variation of the Nimzo-Indian which is
named after him. He equalized and later got
an edge, but was content to play for safety and
the game was later drawn.
Karpov's next opponent was Yusupov, who
got an edge in the Leningrad. Karpov managed
to ease the pressure and later took over and won
thanks to some mistakes from his opponent.
In the second game Yusupov employed the
Lasker Defence against the Queen's Gambit a strange choice for a must-win game. Karpov
obtained a slight plus in a quiet position, and
later managed to get an extra pawn in a rook
ending. Ironically, the way he converted his
advantage was reminiscent of an early Yusupov
win over Tal.
Karpov's opponent in the final was Gavrikov,
who played the 4.'Wc2 system against the
Nimzo-Indian. White had a small edge with his
bishop pair, but after castling long he allowed
Karpov to build up a menacing attack on the
304
The Prime Years
b-file. The end of the game was dramatic:
Karpov blundered a rook in a winning position,
but Gavrikov captured it with the wrong piece
and lost the game because of it. The second
game was a Chebanenko Slav. Karpov went for
safety, and Gavrikov sacrificed a pawn to stir
up complications. But Karpov was equal to the
challenge, and when the fireworks had died
down he had reached a winning endgame with
a queen versus rook and knight. He made no
mistake and wrapped up the win convincingly.
Apart from winning the overall tournament,
Karpov's score of 7/8 was most impressive,
even if he did have a slice of good fortune in
two of the games.
FIDE candidates match versus Gelfand
Karpov was seeded straight to the semi-final
where he met Boris Gelfand. In Game 1 the
Israeli grandmaster opened with 1 .e4 and
played the Short System in the 3.e5 Caro­
Kann. The opening resulted in a double-edged
position and Gelfand sacrificed a piece. Both
players played well and the game soon resulted
in a perpetual check. The second game was a
symmetrical Fianchetto Griinfeld. Gelfand
got close to equality although later he had to
find some accurate moves to avoid suffering.
He was up to the task and a draw was agreed
shortly after the time control.
In Game 3 Gelfand came up with a new idea
in the Short System. Karpov played carelessly
after the opening and his position deteriorated.
He did a reasonable job of getting back into the
game, but then became too ambitious and soon
found himself in a lost ending, which Gelfand
converted. In Game 4 Karpov immediately
struck back. The players repeated the Fianchetto
Griinfeld, but a different type of middlegame
ensued, with White having an IQP. Gelfand
was a bit worse, and subsequent inaccuracies
allowed Karpov to build up a powerful attack,
which was enough to decide the game.
In Game 5 Karpov switched to 1 . . .e5 and
drew comfortably with the 9 . . . lLJd7 variation
in the main line of the Ruy Lopez. The sixth
game was his best of the match.
I GameJ9 1
Anatoly Karpov - Boris Gelfand
Sanghi Nagar (6) 1995
l.d4 tLlf6 2.c4 c5 3.d5 b5
Gelfand takes a serious risk - according to
the database he had never played the Benko
Gambit before. On the other hand it is hard to
imagine that Karpov and his team would have
prepared for it. Karpov himself had only faced
the gambit three times previously. Over his
career he played eight games against the Benko
Gambit, winning half of them and drawing
the others.
4.cxb5 a6 5.bxa6!
Karpov plays a new line for the first time
at the age of forty four - it shows he was still
making the effort to learn and improve. In
1 993 he won two games with 5.b6, so Gelfand
would surely have prepared thoroughly for
this.
5 ... g6 6.tLlc3 .ba6 7.e4hfls.'it>xf1 d6 9.g3
i.g7 10.'it>g2 tLlbd7 1 1 .tLlO 0-0 12.h3
Karpov prevents the knight from coming to
g4. In 1 996 he played 1 2.�e 1 twice, but only
managed to draw against Georgiev and Salov.
12
•.
J:!: a6
This move was played first in by Sievers in
1 984, although the plan involving . . . Was
only really became popular when Fedorowicz
used it to beat Rajkovic in 1 987 (although the
American grandmaster preferred 1 2 . . . �a7).
In the future Karpov faced 12 . . .tZl b6 twice:
he defeated Matamoros Franco in 200 1 , and
drew a rapid game against Adianto in 2002.
1 995
Anatoly Karpov - Boris Gelfand
13.ig5!?
30 5
Karpov aims to provoke . . . h6. The idea was
new at the time, and has seldom been repeated.
13.E:e 1 is the most popular move by far.
similar with his queen.) Play might continue
1 7.Wc l Wh7 I B.tLla4 !'1xa4 1 9.bxa4 Wxa4
and Black has some compensation for the
exchange.
13 ... h6 14.id.2 Wl'a8 15J�el
17J�e2
The current position has been reached in
over sixty games, most of them featuring an
earlier E:e 1 from White.
1 7.a4 was also possible, although after
1 7 . . . !'1ab6 I B.!'1bl tLl c7 White does not get a
chance to put his knight on b5.
a
b
c
d
e
f
g
h
15 . J�b8
In the majority of games Black has preferred
to crack the centre with 1 5 . . . e6, with mixed
results.
.
16.b3
Karpov secures his b-pawn, having ascer­
tained that Black is unable to exploit the
weakening of the long diagonal. From here
White has two main plans. One idea is to play
for a queenside blockade with a4 and tLlb5,
which can work well as long as the b3-pawn
does not become weak. The other plan involves
gaining space in the centre with e5. This will
usually require careful preparation in order to
prevent Black from exploiting the opening of
the centre to develop counterplay.
16... tDe8
16 ... E:a3!? was worth considering. (In the
1 2 ... Wa5 line Black often does something
a
17 tDc7
b
c
d
e
f
g
h
••.
1 7 . . . !'1a3!? is reasonable here as well: I B .!'1b 1
( I B.Wc l ? !'1bxb3! or I B.!'1c l .ixc3!) I B . . . .ixc3
1 9 ..ixc3 !'1xa2 20.!'1xa2 Wxa2 2 1 .tLld2 Wa6
Black is in the game.
18.!'1c1
The rook wisely vacates the long diagonal.
18 �a7
..•
Gelfand prepares a knight manoeuvre to b4.
Black has some other ideas as well.
I B . . . .ixc3?! looks questionable, as Black
remains a pawn down: 1 9.!'1xc3 !'1xa2 20 . .ixh6
!'1al It is hard to say just how big White's
advantage is, but the black kingside certainly
looks vulnerable.
I B . . . !'1a3!? Once again this move was worth
considering. 1 9.!'1c2 tLlb5 20.tLla4 (20.tLlxb5
The Prime Years
306
�xb5 2 1 .w"c 1 �a6 White has a small edge,
but Black retains his typical Benko-style
compensation.) 20 .. :Wa6 (Also interesting is
20 . . . g5! ? 2 1 .\Wc 1 e6 with counterplay.) 2 1 .w"c 1
�xa4 22.bxa4 tLlc3 23.�xc3 W"xc3 24.w"xh6
�b4 Black's compensation is enough for a
roughly balanced game.
20 ... llJb4
Otherwise Karpov would have played W"c3
next, exchanging Black's strong bishop and
preparing e5 followed by a kingside attack.
21 .hb4
2 1 .\Wb l ? was not a serious option in view
of 2 1 . . .�xa4! 22.bxa4 tLlxd5 when Black gets
lovely play for the exchange.
21..Jhb4
a
20.llJa4!
b
c
d
e
f
g
h
Karpov blocks the queens ide. It slows his
own play, but it is more important to shut
down Black's counterplay first.
In the event of the more obvious 20.a4 Black
has a few reasonable plans, the choice between
them being large a matter of taste:
a) 20 . . . tLlc7 2 1 .�b l �a6 (Or 2 1 . . .�ab7 22.�ee l
\Wa6 23.tLldl e6 24.dxe6 tLlxe6 and White is
just a bit better.) 22.w"e3 \Wb7 23.tLld2 �b6
24.tLla2 f5 25.f3 fxe4 26.fxe4 Both players
must play carefully, but I would take the extra
pawn if given the choice.
b) 20 . . . c4! ? Sacrificing a second pawn is a rea­
sonable option. 2 1 .bxc4 tLlac5 22.�b l (22.e5
tLlxe5 23.tLlxe5 W"xe5 Black should be okay.)
22 . . . �xb l 23.\Wxb l �b7 24.\Wa2 White keeps
his two-pawn advantage, but after 24 . . . �b3 or
24 . . . �b4 he remains rather tied up.
a
22."Wd3!
b
c
d
e
f
g
h
Karpov clears the way for his rook to get to
c4.
22 .. J�a5 23J�ec2 h5?!
This move has pros and cons, but
unfortunately for Gelfand the latter outweigh
the former. Black wants to open the h6-c1
diagonal for his bishop, but the additional
weakening of Black's kingside proves more
important as the game goes on. Gelfand's main
defences at the time were the King's Indian
and the Sicilian Najdorf, both of which require
Black to play actively and energetically instead
of merely waiting and defending. Perhaps this
contributed to his decision-making on the
present turn.
23 . . . �b7 was possible, although White has
Anatoly Karpov - Boris Gelfand
1995
a nice way to improve his position: 24.tLl d2!
'iWb8 25.tLlc4 :8:a8 26.f4 White is better; his
knight is ideally placed on c4, and he has good
prospects in the centre.
23 ... tLle5! Black's best chance was to exchange
the knight and then wait and see. 24.tLlxe5
�xe5 25 .l"k4 :8:b8 Since the knight is unable
to return from a4, it will not be easy for White
to make anything of his small edge. On the
other hand he could play on for a long time
with practically no risk of losing.
29.We3 .td4 30.Wf3 Black will have to worry
about a kingside attack based on f5 .
26J'Hc2 :8:5b7 27.�e2 :8:a7
In the next phase of the game Karpov plays
several fairly nondescript moves, which was
probably a deliberate ploy to tire his opponent.
He is in control of the position so he can afford
to do it.
a
a
b
c
d
e
f
g
h
24.:8:c4!
Karpov kills Black's aspirations to seize the
initiative on the queenside. He does it in a
highly economical way, as his rook and knight
are able to keep three of the opponent's heavy
pieces at bay.
307
b
c
d
e
f
g
h
28.:8:dl �a5 29.�dl �a6 30.�dl i.h6
31.�dc2 i.g7 32.�e3 �a8 33.�dl �a7
34.�dl �b5 35.�e2 �b8
Over the last eight moves White has moved
his rook from c2 to d l , while Black's pieces are
in identical positions.
24. :8:b8 25.�dl :8:ab5
..
Gelfand decides to wait for a while. It would
be interesting to know how Karpov would have
proceeded after a knight exchange: 25 . . . tLle5
26.liJxe5 .txe5 A possible continuation is:
27.f4 (There is also 27.g4!? hxg4 28.hxg4 when
White keeps an edge and can look to play on
the h-file.) 27 . . . .tf6 28.:8:n !? White can play for
an attack. (If 28.:8:e l then 28 . . . :8:d8! ? restrains
the e5-push.) 28 . . . <;t>g7 (28 . . . 'it>h7 29.:8:f3 :8:d8
30.We2 Wb8 3 1 .e5 White gains useful space.)
a
b
c
d
e
f
g
h
The Prime Years
308
36.tlJd2!
After a period of manoeuvring Karpov
embarks on a small but significant plan,
namely to retreat his rook and park his knight
on the c4-square. Just as in many of his other
games, he times his direct action to take place a
few moves before the time control, when it will
create maximal problems for the opponent.
merely waits he could soon find himself in
trouble, for instance: 39 . . . WbS 40.E!:b l 'tt>h7
[40 . . . .th6 4 l .f4 h4 42.a5! E!:b5 43.E!:a4 hxg3
44.Wxg3 E!:a7 45.lt:lc4±J 4 1 .a5! E!:a6 42.!'la4
E!:b5 43.lt:lc4 White has made significant
achievements on the queenside.) 40.E!:b 1 !'lb8
4 l .f4 .tf6 42.lt:lf3 Wb7 (42 . . . .tg7 43.e5±)
8
36 ... tlJb6
Gelfand is not interested in finding out how
strong the knight could become on the c4square.
7
6
5
4
37.tlJxb6lhb6 38.a4 �h6?!
Gelfand embarks on a faulty plan to force
matters on the side where he is weaker.
Piling up the heavy pieces on the b-file was also
not the right plan: 3S . . . Wb7 39.f4 E!:aS 40.Wd3
E!:bS 4 1 .E!:b 1 WfS 42.a5! E!:b5 43.E!:a4 White
pushes his opponent back on the queenside.
Black's best bet was to continue waiting:
3S . . . E!:ab7
At some point White will have to decide
whether to take action on the queenside, on
the kingside, or in the centre.
a
b
c
d
e
f
g
3
2
a
b
c
d
e
f
g
h
43.E!:c2! (43.e5 .tg7 is less clear) 43 .. !'lb4
(43 . . . E!:xb3 44.E!:xb3 Wxb3 45 .Wxb3 !'lxb3
46.E!:a2 The endgame is excellent for White
as the a-pawn has become strong.) 44.'iJd2
E!:d4 45 .We3 White keeps some advantage.
39 . . . Wa6
39 . . . WeS 40.e5 WaS 4 1 .exd6 exd6 42.!'le4
(42.f5) 42 . . . Wh7 43.f5 Black's kingside is
weak.
39 . . J�bS!? 40.Wf3 Wb7 4 1 .E!:b l (4 1 .!'lfl
White can also consider playing for f5.)
41 . . . E!:aS 42.Ek2 E!:b4 it is hard to progress
with White.
.
h
39.f4
The queens ide plan is also quite promising:
39.Wd3 Wa6! The queen keeps an eye on
the a-pawn and pins the c4-rook. (If Black
a
b
c
d
e
f
g
h
Anatoly Karpov - Boris Gelfand
1995
40.e5!
White should not shy away from direct
action.
Further preparation with 40.l'!e l allows
40 ... l'!xb3! 4 1 .tLlxb3 l'!xb3 when Black has
good compensation for the exchange.
Manoeuvring does not lead anywhere
special: 40.Wfd3 if6 4 1 .l'!b l l'!b8 42.tLlf3
iWb7 43.:1'k2 l'!b4 44.tLld2 l'!d4 45 .Wfe2
1'%b4 It is not easy for White to make
progress.
40 ... dxe5
If the pawn is allowed to advance to e6,
Black's whole position will be split in two.
4 1 .fxe5 ih6!
Black's position is not entirely safe, but he is
still fighting.
42.e6
42.tLle4 l'!xb3 43.l'!c2 Wfxe2t 44.l'!xe2 c4
Black has counterplay.
42.l'!e4 Wfxe2t 43.l'!xe2 ixd2 44.l'!dxd2
1'%xb3 45 .d6 l'!d7 46.a5 White is somewhat
better in this double-rook endgame, but the
win is far from certain.
8
7
6
5
4
309
Black cannot afford to enter an endgame, as
the threat of a breakthrough with d6 would
paralyse him.
44.Wff2
The strong e6-pawn gives Black considerable
problems.
8
7
6
5
4
3
2
1
a
39.f4 h4?
b
c
d
e
f
g
h
Black decides to force matters, but the
opening of the kingside is in White's interest,
as he can quickly bring considerable firepower
to bear on that part of the board. Perhaps
Karpov was short of time and Gelfand wanted
to force him to make a decision, but if that
was the case then the gamble proved wholly
unjustified.
40.�g4
3
Karpov brings the queen to the attack and
forces his opponent to open the h-file.
2
a
b
c
d
e
f
g
h
42 .. .f6!?
This looks ugly, but Black has chances to get
away with it.
43.1'%e4!?
43.tLle4 l'!xb3 44.l'!c2 Wfa8 45.tLlxc5 l'!e3
46.Wfc4 l'!c7 Black is living dangerously, but
he is alive.
43 ... '.Wa8
There was a second strong continuation:
40.l'!fl !? hxg3 4 1 .l'!c3! l'!b4 42.l'!xg3 With the
makings of a powerful attack.
40 ...hxg3 41 .h4
4 1 .l'!c3!? is also promising. In the coming
variations, we will see that the successful
transfer of this rook to the kingside is often
integral to White's success.
The Prime Years
310
43 .. .l''1 ab7 44.';!;>xg3 Wa6 (44 . . . Wa5 45.�g4
E!:c7 46.\t>h2+-) 45 .Wg4 E!:xb3t (45 ... e6
46.E!:gl +-) 46.';!;>h4! e6 47.E!:gl f5 4S.exf5
exf5 49.Wxf5 White's attack is about to
break through.
44.dxe6 fxe6
a
41. .. �h7?
b
c
d
f
e
g
Gelfand bases his defence on transferring his
queen to the kingside. An unfortunat� sid�­
effect of this plan is that it involves placmg his
king on a more dangerous file.
More resilient was:
4 1 . . .�g7!
The point of this move is to prevent the c4rook from joining the attack.
42.h5
After 42.Wxg3 WeS 43.Wg4 e6 44.dxe6 fxe6
Black is struggling, but it will not be easy to
break his resistance.
42 . . . gxh5 43.Wxh5
a
b
c
d
e
f
g
a
b
c
d
e
f
45 .Wg6!
The queen gets closer to the enemy king and
attacks the e6-pawn, but more importantly,
enables the rook to use the c3-square.
45 . . . E!:e7 46.E!:c3 E!:b4 47.E!:h l !
47.E!:xg3?! E!:d4! sets up an awkward pin.
47 . . . E!:d4
47 . . . Wb7 is too slow: 4S.E!:e3! White defends
the e4-pawn in order to transfer his knight
to g5 . 4S . . . Wa6 49.ttJf3 d5 50.ttJg5 With
decisive threats.
4S .Wh7t \t>f8
4S . . . \t>f7? 49.ttJf3 is crushing.
h
43 . . . e6!
Black opens the seventh rank, allowing his
rooks to join the defence.
The queenside counterattack is too slow:
a
b
c
d
e
f
g
h
49.WhSt! �xhS 50.E!:xhSt c;!;>f7 5 1 .E!:xaS E!:xd2t
31 1
Anatoly Karpov - Boris Gelfand
1 995
52.lt>xg3
Black has survived the attack, but White's
extra pawn gives him decent winning chances
in the ending.
4S . . . 8:b8
45 . . . 8:b4 meets
refutation:
with
a
spectacular
42.h5
Now is not the right time to transfer the
rook with 42.2"k3? due to 42 . . .f5! .
42 ...%Yg8?
This allows a convincing refutation.
42 ... 1Wh8!
This surprising queen move would have
posed more problems. The point is to stop
the c4-rook joining the attack. White can still
win, but considerable accuracy is needed.
43.hxg6t
After 43. <;t>xg3 Wb2! White is better, but he
is tied up to a certain extent.
43 ... fxg6 44.Wxg3
44.We6 Wb2 4S .Wf7t Wg7 46.Wxg7t <;t>xg7
47.<;t>xg3 White's extra pawn gives him
excellent winning chances but the game is
far from over.
a
b
c
d
e
f
g
h
44 ... 1Wb2!
The queen attempts to create maximum
disruption in White's camp.
45.2"k3
If45 .We3 8:b8 46.e5 (46.Wg3 8:ab7) 46 . . . 8:f8
47.8:e4 dxe5 Black once again manages to
stay in the game.
c
d
e
f
g
h
46.8:g 1 !! White goes after the g6-pawn. (After
46.f5 ? g5 Black is still alive.) 46 . . . Wxd2t
(46 . . . .ixf4 47.Wxf4 Wxc3 48.Wf7t+-)
47.<;t>h l g5 48.8:e3! White cuts off the queen
with this lovely move. 48 . . . 8:xb3 49.8:xb3
Wxf4 50.Wxf4 gxf4 5 1 .8:a3 White wins the
endgame easily.
a
a
b
b
c
d
e
f
g
h
46.<;t>h l !
46.8:d3 c4! Black creates counterchances.
46.8:f3 Wc2! (46 . . . c4 47.Wf2! wins as White
breaks the pin on the second rank while
attacking the a7 -rook.) 47.8:ffl 8:f8 Black's
queen is annoying.
46 . . . 8:g8 47.lLlc4 We2 48.8:e l WhSt 49.<;t>g2
Finally White is ready to begin the decisive
assault on the h-file, and there is not much
Black can do about it . . .
The Prime Years
312
Black resigned as his kingside is about to
disintegrate.
8
7
6
5
4
3
2
1
a
43Jk3!
b
c
d
e
f
g
h
Karpov brings the rook into the attack.
Black has nowhere near enough pieces on the
kingside to deal with White's overwhelming
firepower.
43
.•.
5
44.hxg6t VNxg6 45.VNh4 VNf6
8
7
6
5
4
3
***
2
1
In Game 7 Gelfand returned to his normal
l .d4, but chose a sideline in the form of 4.e3
against the Queen's Indian. Karpov equalized,
and for a long time the game was roughly
balanced. Gradually the pieces came off the
board and an equal minor piece endgame
ensued. Gelfand tried to play for a win, but it
was Karpov who gradually obtained the upper
hand, and he eventually secured the win after
a magnificent performance in a same-coloured
bishop endgame.
Game B was a Slav. Karpov went for
safety, Gelfand was unable to stir up any
complications, and the game never looked like
it would end in anything other than a draw.
The ninth and final game was a 1 . . .e5 English,
in which Karpov played the defensive system
involving . . . tt:l f6 and . . . �c5 which bears his
name. Later he exchanged his light-squared
bishop in order to gain space in the centre, and
was able to develop threats against Gelfand's
king. The Israeli grandmaster made a mistake
and Karpov punished him to win the game
and the match. Karpov's final score of 6-3,
with four wins, four draws and one loss, was
highly impressive.
a
b
46J�hl! fxe4
c
d
e
f
g
h
Black is simply defenceless. Taking the
rook would not have helped him: 46 . . .'IMfxc3
47.VNxh6t <j;JgB 4B.Whlt rj;JfB 49.Wxf5t rj;JgB
50.WcBt <j;Jf7 5 U'1h7t Black loses his queen
at least.
47J3xg3 �b4 48.VNg4 �a8 49.�gh3
1-0
Karpov's next classical tournament was Linares,
although before that he played in the famous
Amber tournament. He made a respectable
6/ 1 1 in the blind part of the event, but in
the rapid part he scored a spectacular 10/ 1 1 ,
which much have given him a great deal of
confidence.
Karpov's first opponent in Linares was
Shirov, who played a Pire. Karpov got a slight
plus with his usual Classical set-up, but Shirov
played well and held the draw. In Round 2
Karpov played the Bogo-Indian against Dreev,
1 995
313
Anatoly Karpov - Nigel Short
and after a brief tactical skirmish the players
agreed a draw in a level ending. In the next
round Karpov faced Tiviakov. An unusual
Queen's Indian line resulted in a blocked
position, and Karpov sacrificed a piece for
three pawns in order to fight for the advantage.
But Tiviakov defended well and after some
simplifications a draw was agreed in an
equal rook ending. In Round 4 Karpov faced
Akopian, who played the Reti. Karpov played
uncharacteristically with an early . . . h5 and left
his king in the centre. His position was shaky
and at one point Akopian could have obtained
serious pressure, but instead he simplified to
an opposite-coloured bishop endgame with an
extra pawn, which Karpov held easily.
After drawing his first four games, Karpov
faced Nigel Short. The English grandmaster
has so far played 1 59 classical games against all
the world champions from Smyslov onwards,
with the exceptions of Fischer and Petrosian.
He won nineteen of those games, drew eighty
six and lost the other fifty four. Before this
game Karpov had a narrow lead in their head­
to-head meetings, with seven wins to the
Englishman's six, and fourteen draws. Their
lifetime score is nine wins to seven in Karpov's
favour, with seventeen draws.
may also have been that none of Short's
opponents had chosen this line.
3 ... a6 4.e3 tlJf6 5.hc4 e6 6.0-0 c5
7.i.d3
Karpov chooses a slightly unusual line, which
had been tested by a few strong players such as
Eingorn. He never repeated it.
A year later he won a nice game with a
different line, which remains theoretically
critical to this day: 7.i.b3 l2J c6 B.l2Jc3 i.e7
9.dxc5 'lWxd l 1 O.2"i:xd l i.xc5 1 1 .i.d2 i.d7
1 2 .2"i:acl i.a7 1 3.i.e l We7 1 4.l2Ja4 b6
I Game 40 I
Anatoly Karpov - Nigel Short
Linares 1995
l.d4 d5 2.c4 dxc4
Short rarely plays this opening, although he
had used it a few times in 1 994, as well as in
his 1992 match against Karpov.
3.tDf3
In the 1 992 match Karpov played 3.e4 three
times but drew each game, so this time he goes
for something different. Part of his reasoning
a
b
c
d
e
f
g
h
1 5 .2"i:xd7t! I wonder whether Black missed
this move completely, or misjudged its
consequences. 1 5 . . . mxd7 1 6.l2Jxb6t!! i.xb6
1 7.i.a4 WcB I B.l2Je5 White has excellent
compensation for the exchange and Black was
The Prime Years
314
unable to deal with his problems, Karpov Gulko, Oropesa del Mar 1 996.
7... cxd4 8.exd4 i.e7 9.tiJc3 b5
Short chooses an ambitious approach,
intending to put his bishop on b7. I find
it slightly surprising that no-one has tried
9 . . . b6!? here.
8
7
13 ... tiJbd7
Short elects to cover the c5-square while
keeping the diagonal open for his bishop. After
1 3 . . . ttJc6 1 4 .i.g5 g6 1 5 .h4! ? it is not easy to
ease White's grip.
It was worth considering simplifying the
position with 1 3 . . . ttJxe4!? 1 4.i.xe4 1xe4
1 5 .�xe4 ttJ d7! (If 1 5 . . . �d5 1 6.�xd5 exd5
1 7.ttJe5 White is better according to Karpov.)
1 6 .i.f4 ttJ f6 1 7.�d3 ttJ d5 and Black has a
comfortable position.
6
5
4
3
2
1
a
10.a4!
b
c
d
e
f
g
h
Playing for a kingside attack is hard
when . . . b4 could appear at any moment,
so instead Karpov decides to soften Black's
queens ide.
10 ... b4
Several years later Kasparov demonstrated
a good alternative: 1 O . . . bxa4!? 1 1 .�xa4t i.d7
1 2.�c2 ttJ c6 1 3.1"lxa6 ttJ b4 1 4.1"lxa8 �xa8
1 5 .�e2 ttJxd3 Black has enough compensation
to maintain the balance, Kramnik - Kasparov,
Moscow (blitz) 200 1 .
l 1 .tiJe4 i.b7 12.�e2
1 2.ttJc5!? is an interesting alternative which
has scored well for White.
14 ... a5
Black has the d5-square firmly under control,
and can choose between posting his queen,
bishop or either knight there. Short opts for
the second option; he could not do it at once
as the a-pawn would fall. There were several
alternatives available:
After 14 . . . ttJ d5 White can play 1 5 .ttJb3 or
1 5 .ttJ c4 with a small edge.
12 ... 0-0 13J'!dl
Karpov avoids the immediate knight retreat,
as 1 3.ttJed2 ttJ c6!? is reasonable for Black.
Another idea was 1 3.i.f4!?
1 4 . . . ttJ b6?! Putting this knight on d5 is less
than ideal as it gives up the e5- and c5-squares.
1 5 .a5 ttJ bd5 1 6.ttJb3 i.d6 1 7.ttJe5 �e7 1 8.1d2
1995
Anatoly
White controls more of the important squares
in the centre.
With 14 . . . WcS! ? Black intends to post his
bishop on bS, without relinquishing control
over the bS-square. l S .tLJc4 i.dS 1 6.tLJfeS Wb7
17.f3 The position is complex.
Karpov
-
Nigel Short
315
15 ..ib5!
Karpov immediately places his bishop on the
newly-created outpost, from which it will exert
a strong influence over the queenside.
15 ... tiJb6 16.tiJb3
From this square the knight eyes the as­
pawn and prepares to hop to cS later.
Black can even try to do without occupying the
d5-square at all: 1 4 . . . Wc7!? I S .tLJc4 ( 1 s . tLJ b3?!
id5 is comfortable for Black; I S .aS! ? could be
considered though.)
a
a
b
c
d
e
f
g
16 ....id5
h
1 5 ... i.e4!? Like in the Semi-Slav, it is useful
for Black to swap the light-squared bishops.
( l 5 ... tLJdS 1 6.g3 2"1acS looks playable as
well.) 1 6.i.xe4 tLJ xe4 1 7.dS exdS I S . tLJ e3
(or l S.2"1xdS 2"1acS) I S . . . tLJ df6 Black has no
problems.
b
c
d
e
f
g
h
Short sticks to his plan and puts his bishop
on the central outpost.
Putting either of the knights on dS would
give up squares for a white knight. For
example, on 1 6 . . . tLJ fdS 1 7.tLJcS is unpleasant,
or if 1 6 . . . tLJ bdS 1 7.tLJeS ( 1 7.tLJcS!?) 1 7 . . . Wb6
White's pieces are better placed after either
I S .i.d2 or I S.tLJc4.
On the other hand 16 ... WdS !? deserved
consideration; the queen and bishop generate a
lot of force on the diagonal and restrict the f3knight. 1 7.tLJcS i.c6 ( 1 7 . . . i.xcS? I S .dxcS WxcS
1 9 .i.e3) I S .i.xc6 ( 1 S .i.f4 2"1fcS 1 9.i.a6 2"1eS
Black is not worse.) I S . . . Wxc6 1 9.tLJeS WdS
20.i.e3 2"1fcS Black has a safe position. White is
unable to exert pressure on the queens ide with
2 1 .WbS? because of the cute 2 1 . . .tLJxa4!.
a
b
c
d
e
f
g
h
Th e Prime Years
316
Short probably wanted to stop the other
knight from arriving on the fifth rank as well.
If 17 . . . Wd6 1 8 .lLle5 White's pieces on the fifth
rank give him a firm grip on the position, and
the coming �f4 will create further problems
for the defence.
The regrouping plan of 1 7 . . . :ga7 1 8.lLle5 Was
can sometimes work well in positions of this
type, but in this case 1 9.f3 maintains a solid
plus for White.
18 ..ie3 :ga7?
This unfortunate move is too slow. Black
should have preferred:
1 8 . . . lLlxc5!
Exchanging the strong knight would have
relieved most of Black's troubles.
1 9.dxc5 Wc7
The position is deceptive. It looks as though
the passed c-pawn should give White some
advantage, but Black can block it securely.
Similar situations can occur in the Hungarian
Variation of the 5 .Wb3 Griinfeld.
Chess Informant annotations, evaluating the
position as clearly better for White, but I
find this assessment to be overoptimistic.
2o . . . lLlg4!
It is important to get the dark-squared
bishop.
2 1 .c6
2 1 .g3 lLlxe3 Black will have enough play on
the dark squares.
2 1 . . . �f6 22.g3 lLlxe3
Black's domination on the dark squares gives
him a good game.
8
7
6
5
4
3
2
1
a
19.1Lle5!
b
c
d
e
f
g
h
Karpov offers his opponent a choice of
exchanges, each of which comes with a serious
drawback.
19 .. ,tiJb8!
a
b
c
d
e
f
g
h
20.:gac l
In the event of 20.h3 Black can obtain
counterplay with 2o . . . lLlh5 or 20 . . . �b3 .
Another possibility i s 20.�d4 :gad8 2 1 .c6
�d6 22.lLle5 lLl e4 with a double-edged
position.
Karpov mentions the text move in his
This move looks ugly, but it seems to
be Black's best chance. Exchanging either
knight is problematic due to the unfortunate
placement of the a7 -rook, which gives White a
huge tempo in some lines.
1 9 . . . �xc5 ? 2o.lLlxd7 lLlxd7 2 1 .dxc5 :gc7 22.c6
wins.
1 9 . . . lLlxc5 20.dxc5 Wc7 2 1 .c6 :gaa8 (2 l ...id6
22.f4) 22.lLld7 (22.�d4 is also strong)
22 . . . lLlxd7 23.cxd7 The chances of Black
Anatoly Karpov - Nigel Short
1 995
surviving against such a mighty passed pawn
are almost nonexistent.
2o.if4 ttJe8?!
Playing an unfamiliar opening creates
problems even for great players. Short's move
prepares a bishop exchange as well as a possible
knight manoeuvre, but it was not the best.
2o ... id6 would have offered stiffer resistance:
2 1 .l"1ac 1 'iffie7 White is somewhat restricted by
the pressure on the c5-knight, so his advantage
is smaller than in the game.
317
24.'iffig4 ig5) 23 . . . ig5 (23 . . . lLl h4 24.:§:d l )
24.ixg5 'iffixg5 25.'iffid3 White i s better but
Black has fair chances to survive.
22.hgS V;¥xgS 23.:§:g3 V;¥e7 24.V;¥hS!
Karpov wastes no time bringing his heavy
artillery into the attack. After the inferior
24.id3? (with the transparent threat of taking
on h7) 24 . . . g6 25.'iffih 5 lLl c6 Black takes the
sting out of the attack.
24
•••
£5
24 . . . g6 25.'iffih6 'iffif6 26.lLl ed7 wins an
exchange. Alternatively after 24 . . . lLl f6 25.'iffih6
g6 26.:§:h3 Black is in trouble.
2S.gh3
25.f3 is also promising, for instance 25 . . . lLl f6
26.'iffih 4 :§:c7 27.:§:e l and White's pieces are
dominating.
2S ttJf6 26.V;¥h4 g6
•..
8
a
b
c
d
e
f
g
h
21.�d3!?
Karpov swings his rook into posltion to
attack Black's king. It was also possible to
play on the queenside: 2 1 .:§:ac 1 lLl d6 22.id3
'gc7 23.lLlc4 lLl c6 24.lLlb6 White keeps some
advantage thanks to his strong knights.
21. .igS?!
Exchanging minor pieces is normally helpful
to the side playing against an IQP, but in this
position it accelerates White's attack.
.
A better idea was 2 1 . . .lLl d6, transferring
the knight to f5 where it is useful for both
active and defensive duties. 22.:§:h3 (22.:§:e l
IiJf5) 22 . . . lLl f5 23.g4!? (Black comes close
to equalizing after 23.'iffig4 g6 or 23.:§:c 1 h6
7
6
5
4
3
2
1
a
27J�e3?!
b
c
d
e
f
g
h
This inaccurate move gives Black the
opportunity to ease his suffering by exchanging
pieces.
27.:§:e l ?!
This would have been a mistake for the same
reason.
The Prime Years
318
27 . . . tD e4!
This move does not equalize fully, but it gets
close.
28.tDxg6! ?
28.tDxe4 'lWxh4 29.2"lxh4 ixe4 30.2"lcl 2"ld8
3 1 .f3 id5 Black has good chances to hold
this slightly worse endgame, as the rook on
h4 is out of play.
28 . . . 'lWxh4 29.tDxh4 tD xc5 30.dxc5
8
7
6
5
4
3
2
1
������--���
a
27 J�d8?
b
c
d
f
e
h
g
..
a
b
c
d
e
f
g
h
30 . . . 2"lc8
Another possibility is: 30 . . . 2"lc7!? 3 1 .tD f3
2"lxc5 32.tDg5 (32.tDd4 e5 33.2"lxe5 2"lc l t
34.ifl ic4 3 5 . tD e2 2"lb 1 Black will have
a very dangerous passed pawn.) 32 . . . 2"lfc8
33.2"lhe3 2"lc2 34.tDxe6 2"lxb2 35.tDd4 f4
36.2"le7 White has good attacking chances,
but the position is sharp and thanks to the
passed b-pawn any result is possible.
3 1 .2"ld3 f4
3 1 . . .2"lxc5? 32.tDxf5! exf5 33.2"lxd5 2"lxd5
34.ic4 wins.
32.2"lxd5 exd5 33.tDf5 2"lxc5 34.2"le8t �f7
35.tDd6t �f6 36.2"lxb8 2"lc l t 37.ifl 2"le7
Black has strong counterplay.
It looks like Short changed his mind; I would
have thought that the purpose of his previous
move was to defend the h7-pawn along the
second rank in preparation for . . . tDe4. He
could have obtained better drawing chances by
means of a small modification of this idea:
27 . . . tD g4!
It is never nice to have a worse endgame
against Karpov, nevertheless this was his best
chance. Exchanges will ease his cramped
position, and the elimination ofthe e5-knight
will give the one on b8 some freedom.
27 . . . tD e4!? 28.'lWxe7 2"lxe7 29.tDxe4 ixe4
was also possible, although in that case the
knight on b8 will remain paralysed for the
foreseeable future.
28.'lWxe7 2"lxe7 29.tDxg4 fxg4 30.2"lc l
30.2"lae 1 2"lc8 (30 . . . 2"lf6 3 1 .2"lc l 2"lc7) 31.:1l:g3
h5 32.f3 gxf3 33.gxf3 'it>g7 The change in
the pawn structure favours Black, who has
excellent drawing chances by this point.
The correct way for White to maintain his
advantage was: 27.f3! Preventing the knight
exchange. 27 . . . tDh5 Black's position is cramped,
so exchanging queens should help him. 28.2"le 1
Despite the queen exchange and the slightly
misplaced rook on the h-file, White keeps a
big advantage.
a
b
c
d
e
f
g
h
1995
Anatoly
30 . . Jk8
30 .. J''1f4? is a mistake due to 3 1 .ttJd3!.
30 .. J'k7!? is reasonable though.
3 1 .:i:'lg3 ttJ c6 32.i.xc6 i.xc6
Black cannot afford to relax yet, but he has
reason to be cautiously optimistic about
holding this somewhat worse endgame.
33.:i:'lxg4
33.:i:'le 1 h5 defends.
33 ... i.xa4 34.:i:'la l i.c2 35.:i:'lxa5 i.f5
Black is worse, but very much alive.
28.f3!
a
b
c
d
e
f
g
319
Karpov - Nigel Short
is also clearly better, but Karpov was always
happy to obtain an advantageous endgame.
He won a great many of them with a much
smaller advantage than the one he possesses in
this game. . .
30. . J�xe7 3 1 .'i!lf2!?
Karpov decides to make a small improvement
on the kingside before doing anything else. It
was also possible to take direct action:
3 1 .ttJc4!?
Karpov mentioned this alternative, which
also gives excellent chances.
3 1 . . . i.xc4
3 1 . . . ttJ c6 32.ttJxe6 (32.i.xc6 i.xc4 33.b3
:i:'lxd4! Black gets counterplay.) 32 . . . :i:'lxe6
33.:i:'lxe6 i.xe6 34.:i:'lxe6 ttJxd4 35.:i:'la6 ttJ b3
36.ttJxa5 :i:'ld l t 37.i>f2 ttJxa5 38.:i:'lxa5 :i:'ld2t
39.�e3 :i:'lxb2 40.:i:'la7 White is winning
thanks to his superior bishop and his stronger
passed pawn.
32.i.xc4 :i:'lxd4 33.i.xe6t
h
Now White gets everything he could have
wished for: he prevents Black from easing his
position through a knight exchange, and his
rook is ideally placed in the centre.
28 .. J:k7 29J�ael!?
With this move Karpov virtually forces a
queen exchange, as otherwise the pressure on
the e6-pawn will be too much for Black to
bear.
Karpov mentions the alternative 29.:i:'lc l !?,
which is also strong, for instance: 29 ... ttJ e8
(or 29 . . . :i:'ld6 30.ttJc4) 30.'!Wxe7 :i:'lxe7 3 1 .ttJc4
Black is in trouble.
29 ... ttJh5 30J'Nxe7
The alternative was 30.'!Wf2 ttJ f4 when White
a
b
c
d
e
f
g
h
33 . . . �f8
33 . . . �g7 34.ttJb3 :i:'ld6 35 .i.c4! This subtle
move, which takes control over the a6-square,
is the most convincing path to victory.
(Karpov analyses 35.ttJxa5 ttJ f4? 36.i.c4 :i:'lxe3
37.:i:'lxe3 :i:'ld2 38.g3 when White should be
winning, but Black can regain his pawn and
stay in the game with 35 . . . :i:'la6!.) 35 . . . :i:'lxe3
36.:i:'lxe3 :i:'ld l t (36 . . . ttJc6 37.:i:'ld3 :i:'lxd3
38.i.xd3 The a5-pawn falls.) 37.�f2 :i:'lb l
38.:i:'le7t �h6 39.:i:'le2 White is winning.
The Prime Years
320
34.lt:Jb3 Ei:d6 35.ttJxa5
If 35 .i.c4 Ei:xe3 36.Ei:xe3 ttJc6 Black avoids
losing material.
35 . . . ttJ f4 36.i.c4! Ei:xe3 37.Ei:xe3 Ei:d2 38.g3
ttJh3t 39.�fl Ei:xb2
by a pawn on b4. See for instance Karpov's
classic win over Unzicker, Game 37 of the first
volume.
32 .. J�c7 33J:�cl ! g5?!
Short tries to do something active, but the
pawn on g5 will be a target.
If 33 . . . fx:g4 34.fx:g4 ttJ e8 35 .b3 ttJ d6 36.ttJc4
Black is in huge trouble on the queens ide.
a
b
c
d
e
f
g
Black should have waited patiently with
33 . . . Ei:f8 or 33 . . . Ei:dc8, although his position
would remain depressing. White does not have
to hurry, and should probably play a move like
34.h3 before deciding on a definite plan.
h
40.i.b5!
White nicely traps the b8-knight.
40 . . . �f7 4 1 .Ei:e8 ttJ a6 42.Ei:c8! Ei:xh2 43.i.xa6
Ei:f2t 44.'it>e 1 Ei:xf3
Black continues to fight, but White's extra
piece should be sufficient to win the ending.
3 1 ...tt:lg7
a
34.�eel!
b
c
d
e
f
g
h
With his last two moves Karpov forces his
opponent to worry about one more thing: the
invasion on the c-file.
34... h5?
32.g4!
a
b
c
d
e
f
g
h
Karpov exerts pressure from both flanks.
The situation on the kingside is reminiscent
of certain variations of the Closed Ruy Lopez,
where a black knight on b7 is dominated
Perhaps in time trouble, Short commits a
fatal error. Other moves were also gloomy for
him, for instance 34 . . . Ei:dc8 3 5 .h3 h6 36.l"lh l
Ei:f8 (36 . . . 'it>f8 37.h4) 37.ttJcd3 Ei:fc8 38.l"lc5
and White invades on the queenside.
35.h3
This
preserves
White's
overwhelming
32 1
Anatoly Karpov - Nigel Short
1 99 5
positional advantage, but there was an even
stronger continuation leading to an immediate
win: 35.tLle4! Eixc 1 (35 . . . Eidc8 36.tLlf6t �f8
37.Eixc7 Eixc7 38. gxh 5 White traps the g7knight.) 36.tLlf6t c;t>f8 37.Eixc1 White will
deliver checkmate in a few moves.
It is possible that Karpov was short of time
and took a pragmatic decision not to change
the position. Nevertheless I doubt that the
Karpov of the seventies and eighties would
have missed this opportunity.
tLlxe5 4 l .Eif8t �h7 42.tLlxe5 Black's king is
defenceless.
Perhaps Karpov was short oftime and needed
to get past move 40. The game continuation
certainly does not throw away White's
advantage, so there was nothing wrong with it
as a practical decision.
4o ...ixc6 41.Eif6 hxg4 42.hxg4 .id5
43.Eig6
Karpov continues his plan and rounds up
the g-pawn.
35 . . . fxg4
White's position is so dominant, virtually
any continuation would have helped him to
invade in one way or another.
3 6.fxg4 �U8t 37.�e3 �U4 38JUI
Karpov chooses to infiltrate on the kingside.
He could also have played 38.tLlcd3 and
invaded on the queenside.
43.tLlcd7! ?
White can also resume the king-hunting plan.
The attack is dangerous, but not necessarily
winning by force. On this occasion Karpov
was right to avoid it in favout of the simpler
winning plan.
43 . . . Eic8
38 ..Eixfl 39.:Sxfl tLlc6
.
39 . . . .ig2 is refuted by 40.Eif6 .ixh3 4 1 .tLlxe6!
tiJxe6 42.Eixe6 as pointed out by Karpov.
a
a
4o.ixc6
b
c
d
e
f
g
h
Karpov decides to cash in and win a pawn.
Unusually for Karpov, he missed a chance to
go after his opponent's king, which would
have ended the game more quickly: 40.tLlcd7!
b
c
d
e
f
g
h
44.Eif2
After 44.Eif7 �h7 45.tLlf6t c;t>h6 White is
winning but there is no checkmate.
44.tLlg6 tLl e8! (44 . . . Eie8 45.Eif2! c;t>h7
46.tLlge5 The other knight will arrive on
f6, with lethal consequences.) 45.tLle7t
(45.EifBt �g7) 45 . . . �g7 There is no mate,
so White will have to settle for a slow win.
44 . . . tLle8!
44 ... .ib3 45.tLlf6t c;t>f8 46.tLle4t c;t>g8
47.tLlg6 Eic2 48.tLle7t c;t>h7 49.Eif6 wins.
45.Eif7
The Prime Years
322
45.Ei:f8t \t>g7 46.Ei:f7t �g8 47.Ei:e7 ib3
Black stays in the game.
45 . . . ib3 46.Ei:e7
Here Black should play 46 . . . Ei:a8 or 46 . . . Ei:d8.
In both cases White remains with a winning
position, but the plan to catch the enemy king
has failed.
43 ... :Ek8 44J�xg5
Eleven moves after Short pushed the g-pawn,
Karpov collects it.
44...Ei:m
48.lLlc5 .idl 49.ctJe4
Karpov's use of the knights is exemplary.
8
7
6
5
4
3
2
1
a
49...Ei:m
b
c
d
e
f
g
h
49 . . . ixa4 50.Ei:xa5 wins.
50.ctJil .ixa4 5U�xa5 .ic6 52.ga7
1-0
Pinning the knight is the simplest win, and
it was enough to force Short's resignation.
a
45.llJcd3!
b
c
d
e
f
g
h
Karpov brings the knight to f4, where it will
block the enemy rook while creating powerful
threats on the kingside.
45 ... @h7
Black has no time to activate the rook:
45 . . . Ei:fl 46.ctJf4! \t>h7 47.ctJxd5 exd5
48.ctJc6+-
46.ctJf4 .ib3 47.lLld7!
This move is easy to find, but cute
nonetheless.
47 . . :EU7
In Round 6 Karpov faced Ivan Sokolov, who
opted for an unusual handling of the Torre
Attack involving long castling. He then
became even more adventurous and sacrificed
a piece, but it was not fully sound and Karpov
gradually neutralized his initiative and won
a nice game. In the next round Karpov got
a small edge in an English Opening against
Khalifman, and soon reached a position with
opposite-coloured bishops in which Black had
no serious weaknesses and appeared to be safe.
Nevertheless Karpov was able to maintain a
lingering plus, and the problems eventually
proved too great for his compatriot, who made
some mistakes before the time control and
allowed his king to be caught.
.
If 47 . . . Ei:c8? 48.ctJf6t leads to an immediate
checkmate after 48 . . . �h6 49.Ei:g6 or 48 . . . \t>h8
49.ctJg6.
Here is a brief extract from Karpov's Round
8 game, which began with a 4.Wi'c2 Nimzo­
Indian:
1 995
a
b
c
d
e
f
g
h
1 7 . . . bxc5! Karpov takes over the initiative
with this excellent exchange sacrifice. (White
was probably hoping for 1 7 .. J�c6 1 8.ib5 !'1:c8
1 9 .ia6 and an early handshake.) 1 8.ixc7
Wxc7 1 9.Wlc3 e5 2o.id3 exd4 2 1 .lLlxd4 !'1:e8t
22.mfl Wlb6 Lautier - Karpov, Linares 1 995.
White was never able to solve the problems
associated with his vulnerable king, and
Karpov went on to win a fine game, his fourth
in succession in the tournament.
In Round 9 Karpov faced Ivanchuk, who was
also playing superbly with a 'plus four' score.
The game featured a most unusual version of
the Dutch Defence, involving the opening
moves l .d4 f5 2.g3 lLl f6 3.ig2 d6 4.lLlc3!?
dS!? For a long while the game was roughly
balanced, but then Ivanchuk took over and
got some advantage. But Karpov managed to
stay in the game and found a way to sacrifice
an exchange for two pawns, after which the
Ukrainian grandmaster felt compelled to force
a perpetual.
In the next game Karpov played the Caro­
Kann against Topalov, who responded with the
Panov Variation. For a while Karpov seemed to
be under pressure, but he managed to stabilize
his position with the help of a pawn sacrifice.
Topalov seemed to have trouble adapting to
the new situation at the board, and his position
soon deteriorated, allowing Karpov to score
another valuable win.
323
In Round 1 1 Karpov transposed to the Panov
from the opposite side of the board, via a l .c4
c6 move order against Beliavsky. The resulting
IQP position was balanced, and remained so
until the players agreed a draw deep into the
endgame. In the next game Karpov played
the Caro-Kann against Ljubojevic, and made
an easy draw against the 2.lLlf3 d5 3.lLlc3
variation, which was incidentally a favourite
line of the young Bobby Fischer.
In the final round Karpov had to try and
beat Illescas for a chance of winning the
tournament. In a sharp Semi-Slav Karpov
sacrificed an exchange for two pawns, but it
was not fully sound and he soon found himself
in trouble. Fortunately for Karpov, the Spanish
grandmaster later misplayed the endgame and
allowed him to escape with a draw. Karpov
finished Linares with an impressive score of
9/ 1 3, with five wins and eight draws. It was a
good result, but sadly only enough for second
place, as Ivanchuk was in outstanding form
and finished a full point ahead.
***
Karpov's next tournament was Dos Hermanas.
His first opponent was Adams, who used
the Bogo-Indian. Karpov soon got the two­
bishop advantage and later won a pawn in
the endgame, but Adams fought hard and was
able to draw thanks to the reduced material.
In the second round Karpov got a tiny edge
on the black side of a Queen's Indian against
Piket, but was unable to do much with it and
the Dutchman held the draw without any
problems.
Karpov's next opponent was Salov, who
played a Queen's Indian. Karpov got a slight
plus, and a sharp tactical sequence resulted in
a rook and opposite-coloured bishop endgame
in which Karpov had a securely defended
pawn on g7. It took some time to win it, but
Karpov's excellent technique made the process
The Prime Years
324
look easy. In Round 4 Karpov had the black
pieces against Lautier, who played the Catalan.
A sharp tactical sequence resulted in a position
where Karpov had some initiative thanks to
a strong passed d-pawn, and he eventually
converted his advantage by catching the white
king.
In the next round Karpov played the English
Opening against Illescas, who responded with
1 . . . e5 followed by Karpov's own system. White
never got more than a tiny edge and the game
eventually ended in a draw. In Round 6 Shirov
unleashed a surprise in the form of the Centre
Game, but Karpov reacted solidly and exploited
some inaccuracies from his opponent. Shirov
soon found himself in a horrible position and
resigned on move 23. In Round 7 Karpov
played a safe line against Gelfand's Semi-Slav,
and the game was soon drawn.
In the penultimate round Karpov faced
Kamsky, who was trailing him by a point and a
halE Kamsky employed the Short System in the
Advance Caro-Kann, and launched a ferocious
attack involving a piece sacrifice, which Karpov
declined. White's initiative persisted, so Karpov
sacrificed an exchange in order to stabilize his
position and bring his king to safety, although
his position remained worse. He lost on time
on move 38, in a position that would have
offered good drawing chances.
In the final round Karpov had the white pieces
against Judit Polgar. He got a slight edge in a
Fianchetto King's Indian, but Judit played well
and took over the initiative in the middlegame.
At one point she had a winning position, but
she overestimated Karpov's counterplay and
simplified to a drawn rook and opposite­
coloured bishop endgame. Karpov's final score
of 5 Y2/9 was enough to tie for first place with
Kamsky and Adams; not a bad result overall,
although it must have been disappointing
for Karpov to finish the tournament so
weakly.
***
Karpov's next tournament was in Dortmund.
In the first round he faced Bareev, who played
the English Opening and got a nice position
with two bishops. For a while Karpov's position
was difficult, but he eventually equalized.
Bareev proceeded to make some mistakes, and
Karpov punished him. In the next two games
Karpov made quick draws with Kramnik and
Ivanchuk, lasting a combined total of thirty
five moves. The former game contained some
interesting moments but the latter looked
prearranged.
In Round 4 Karpov had the black pieces for
second consecutive game. His opponent was
Jeroen Piket, who got some advantage in the
Catalan. Karpov defended well and sacrificed
a pawn in order to reach a drawn endgame
with queen and opposite-coloured bishops,
which he held comfortably. In the next round
Karpov faced Beliavsky and got a slight plus in
an Orthodox Queen's Gambit. Later he gained
space and obtained a strong passed pawn on d6.
Following a long manoeuvring phase, Karpov
pushed the pawn to d7 and forced Black's
resignation. In Round 6 he scored another fine
victory, outplaying Lobron on the black side of
a Petros ian Queen's Indian.
In Round 7 Short chose the Chebanenko
Slav. Karpov was unable to achieve a significant
advantage in the roughly symmetrical position,
and a draw was agreed in the endgame. In
Round 8 Lautier equalized with the Semi-Slav,
but Karpov kept playing and eventually got
some advantage with the help of a minority
attack. Lautier made a serious mistake shortly
before the time control, and soon had to resign.
In the final round Karpov had the black pieces
against Peter Leko. He equalized in one of the
sharpest main lines of the 4 . . . ltJ d7 Caro-Kann,
and soon had the more comfortable side of a
somewhat simplified position. But Leko held
firm and eventually Karpov had to settle for
a draw.
Karpov finished with a good score of 6Y2/9,
1995
Anatoly Karpov - Curt Hansen
with four wins and five draws. Alas, it was
only good enough for second place behind
Kramnik, who won by half a point.
***
Karpov's last tournament of the year was in
Groningen. In the first round he got into
trouble on the black side of a Nimzo/Queen's
Indian hybrid against Ivan Sokolov, but
managed to simplify to a pawn-down endgame
with knight against bishop, which he duly
drew. In Round 2 Karpov played the English
Opening against Adams, but achieved no real
advantage and the game remained balanced
until the ending when a move repetition
occurred. In the next round Lautier tried the
Petros ian Variation against Karpov's Queen's
Indian. The Frenchman played energetically
but Karpov was equal to the challenge and a
draw was agreed j ust before the time control in
a level endgame.
In Round 4 Karpov got his first win of the
event against Piket, who opted for the sharp
3.e4 Nc6 variation in the Queen's Gambit
Accepted. He sacrificed a pawn but failed to
obtain any real compensation, and Karpov
pressed home his advantage in the queenless
middlegame in a most convincing fashion.
Part of the game can be found in the notes to
Game 48.
In the next round Karpov made a very quick
draw with the black pieces against Tiviakov;
it must have suited him well, as he knew he
would need all his energy in the next round,
where he was due to have the black pieces for
the second time for a row against Kamsky.
Karpov tried a different approach against the
Short System, and equalized comfortably. He
gradually took over the initiative, and later won
a pawn. But Kamsky defended stubbornly,
and after missing some opportunities Karpov
eventually had to settle for a draw after more
than a hundred moves.
325
Karpov's next opponent was Curt Hansen.
The Danish grandmaster has played a total
of twenty games against seven different world
champions, scoring nine draws and eleven
losses. Prior to this game Karpov had beaten
him twice and drawn once. This was their last
meeting over the board.
I Game 41 I
Anatoly Karpov
-
Curt Hansen
Groningen 1995
l.d4 ttlf6 2.c4 g6 3.lLlc3 i.g7 4.e4 d6 5.B
0-0 6.i.e3 ttl c6
Karpov faced this variation four times in
regular games, scoring three wins and a draw.
Hansen rarely played the King's Indian, but he
must have prepared this line.
7.ttlge2 a6 S.\Wdl gbS 9.ttlc1
Karpov had played one previous game
from this position, which continued: 9.h4 h5
1 O.ttJcl e5 l 1 .d5 ttJ d4 1 2.ttJ l e2 c5 1 3.dxc6
bxc6 1 4.ttJxd4 exd4 1 5 .ixd4 Ei:e8 1 6.ie2 d5
1 7.cxd5 cxd5 1 8 .e5 ttJ d7 1 9.f4 ih6 20.�e3
White obtained a small edge and went on to
win, Karpov - Kindermann, Baden-Baden
1 992.
9 ...e5 10.d5 lLld4 11.ttl le2
Later Karpov switched to another main line:
l 1 .ttJb3 ttJxb3 1 2.axb3 c5 1 3.g4 h5 1 4.h3 ttJ h7
1 5 .gxh5 �h4t 1 6.�f2 �xf2t 1 7.\t>xf2 gxh5
1 8.id3 h4 1 9.\t>e2 \t>h8 20.b4 cxb4 2 1 .ttJa4
Ei:g8 22.ib6 id7 23.ic7 White obtained a
nice positional advantage, Karpov - Xie Jun,
Guanzhou 2000.
1l ... ttlxe2
1 1 . . .c5 is the other main line.
12.i.xe2 ttlh5 13.0-0-0
The Prime Years
326
Castling long is normal in the Samisch
variation.
13 ... £5 14.c;f;bl
Karpov does not hurry with c5 , but moves
the king to a safer place.
A waiting move like 1 7 . . . Wfe7 was worth
considering, for instance: 1 S.exf5 �xf5t
( 1 S . . .gxf5 1 9.f4!) 1 9.<;h 1 �d7 White's
advantage is small.
8
7
6
5
4
3
2
1
a
14 ... lLlf4
b
c
d
e
f
g
h
This move had hardly ever been played
before, and has never been repeated. It is not so
bad, but does not achieve much as the knight
will soon be driven away.
The main line of 1 4 . . . �d7 has scored well
for Black.
a
18 ..ig5!
b
c
d
e
f
g
h
With this subtle move Karpov enables
himself to control the e4-square with his pieces
in the event that Black exchanges with . . . fxe4.
18 ... �e8
Black does not have enough time to prepare
. . . h6, for instance 1 S . . .1''U7 1 9.exf5 �xf5t
20. 'it> a 1 WffS 2 1 .h4 and White has the initiative
on the kingside.
15 ..ifl
Naturally Karpov keeps his light-squared
bishop.
15 ... b6?!
This defensive move betrays the fact that
Hansen was inexperienced in the King's Indian.
The more flexible 1 5 . . . �d7 was better.
16.g3 lLlh5 17..ie2 lLlf6
1 7 . . . fxe4 would have led to similar positions
as the game: 1 S.lLlxe4 lLl f6 1 9.�g5 �f5
( 1 9 . . . WfeS 20.�xf6!) 20.�d3 WfeS 2 1 .�xf6
�xf6 22.h4 According to Karpov White
stands clearly better, and indeed Black is rather
passive.
19..id3
Karpov puts pressure on the f5-pawn.
Playing for g4 with 1 9.h3?! allows Black to
obtain counterplay with 1 9 . . . lLlh5!.
1 9.1''1 dfl !? was a decent alternative. Karpov
mentioned this move, along with some lines
to show that it is tactically viable: 1 9 ... fxe4
(Black could also consider 1 9 . . . b5!? instead
of the simplification.) 20.�xf6 �xf6! (The
zwischenzug 20 . . . exf3?? loses to 2 1 .�xg7 fxe2
22j''1xfSt) 2 1 .lLlxe4 �h3 22.E1fg 1 �f5 23.h4
White is a bit better, regardless of whether or
not Black decides to go for a position with
opposite-coloured bishops.
Anatoly Karpov - Curt Hansen
1 99 5
327
19 fxe4
Hansen decides to simplifY. In doing so he
reduces Karpov's active play, but also limits his
own prospects for counterplay.
recapture. Black will be able to exchange the
blockading piece on e4, but in the resulting
simplified position he will still have some
problems, as we will see.
1 9 . . . b5!
This move is much more in the spirit of the
King's Indian. Black should be willing to
accept a few pawn weaknesses in order to
generate dynamic counterplay.
20.cxb5
20.h3 lZlh5 2 1 .lZle2 bxc4 22.i.xc4 fxe4
23.fxe4 �f3 and Black gets counterplay.
20 ... axb5
The c7 - and b5-pawns have the potential to
become weak, but Black has enough activity
to keep his opponent from attacking them
effectively.
20.i.xf6? was not an option as 20 . . . exd3! is
strong.
..•
a
b
c
d
e
f
g
If 20.fxe4 b5! is reasonable, as with the knight
on c3 it is not easy for White to exert pressute
against the c7 -pawn.
20 tlJxe4 21.i.xe4 i.f5
. •.
After 2 1 . . .b5 22.�c1 b4 23.g4! both black
bishops remain passive.
h
2 1 .'lWc2
2 1 .a3 fxe4 22.fxe4 b4 gives Black
counterplay.
2 1 .lZle2 fxe4 22.fxe4 lZl g4 23.�c1 'lWf7 This
time Black gets active on the kingside, and
his chances are not worse.
2 l . . .b4 22.lZle2 fxe4
22 . . . 'lWd7!? looks odd but may not be bad.
23.fxe4 b3!? 24.axb3 i.h3
Black has promising compensation for the
pawn.
20.tlJxe4!
Karpov chooses the strongest way to
22.g4?! i.xe4t 23.fxe4 �f3 (23 ... �f4? The
exchange sacrifice on f4 is often strong in the
King's Indian, but here after 24.i.xf4 exf4
25 .'lWxf4 the position is too simplified for
Black to make anything happen.) 24.'lWe2 'lWf7
25.�hf1 �f8 Black is not worse as he controls
the only open file.
22 ...i.xe4t?!
The Prime Years
328
Black releases the tension on the kingside.
It allows him to force simplifications on the
f-file, on the other hand it virtually eliminates
any remaining chances he had to obtain active
play.
28 ...i.xf8
22 . . . b5!? would have kept the possibility of
creating some play on the b-file.
2B . . . �xfB 29.Yfif3t 'it>eB 30 . .ih4! is
troublesome for Black, as pointed out by
Karpov.
23.Cxe4 Yfif7
Hansen drags a rook to the f-file in order to
exchange it.
his bishop with . . . .if6. Karpov takes the took
voluntarily as he wants to keep his bishop on
the c 1 -h6 diagonal.
8
7
6
24J3hfl Yfid7
5
8
4
6
2
3
7
1
5
4
c
d
f
e
g
h
.
Karpov feels that the bishop should help to
carry out the b4-advance.
2
25.g4!
b
29 .idl
3
1
a
a
b
c
d
e
f
g
h
Highlighting another advantage of White's
22nd move. Now the black queen will not be
able to go to h3.
29 ... ..t>g7
Hansen wants to improve his bishop.
With 29 . . . a5! ? Black anticipates his opponent's
queens ide pawn advance.
25 .. Jhfl 26J3xf1 gfB 27.h3
Karpov has no qualms about exchanging
rooks, so he calmly defends the g4-pawn.
27... h6
27 . . .:!'hfl t 28.Yfixfl is similar to the game.
28.gxf8t
Karpov pointed out that after 2B . .ih4 13f4
29 . .ig3 13xfl t 30.Yfixfl Yfif7 3 1 .Yfie2 a5 Black
will solve most of his problems by activating
a
b
c
d
e
f
g
h
The idea should be sound, although it
1 99 5
carries certain risks, for instance in the event
of a queen exchange the white king may be
able to penetrate using the light squares on
the queenside. It is hard to j udge the merits
of such a move under the pressure of a ticking
clock, even for a fine grandmaster like the
Dane. White has three main ideas:
) 30.h4 ie7 3 1 .h5 'it>g7 (3 1 . . .ig5 32.ixg5
hxg5 33.hxg6 We8 is a difficult queen ending,
although Black may be able to hold it.) 32.Wf3
ig5 (32 . . . We8 33.hxg6 Wxg6 34.Wf5 is
somewhat inconvenient for Black.) 33.ixg5
hxg5 Black should be able to hold this queen
ending.
a
b) 30.b3
Karpov evaluates this move as clearly better
for White, without any further analysis.
30 ...ie7! 3 1 .Wf2
31 .Wfl <j{g7 32.Wc l h5 Black gets
counterplay.
31 .ixh6 g5 Black traps the bishop in an
original way. 32.Wf3 'it>h7 (32 . . . We8 33.Wf5
Wff7) 33.if8 id8 White has no winning
chances with the buried bishop.
a
b
c
d
e
329
Anatoly Karpov - Curt Hansen
f
g
h
3 1 . ..ig5 ?
The simplification favours White.
31 ... 'it>g7! gives good drawing chances.
32.ixg5 hxg5 33.Wf6 Wfl
33 ... 'it>h7 34.<j{b2! White will soon be able
to take the g5-pawn.
34.Wxflt!
If34.Wxg5 ? Wfl t 35.<j{b2 Wf2t Black either
gets a perpetual or takes the e4-pawn.
34 . . . 'it>xfl
It is instructive to see how White can win
this endgame.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
35. 'it>b2
Going for the kingside achieves nothing:
35.'it>c2 'it>f6 36.<j{d2 <j{fl 37.<j{e3 'it>f6
38.'it>f3 'it>g7 39.<j{g3 <j{h7 40.h4 gxh4t
4 1 . 'it>xh4 'it> h6 Black draws easily.
It would be a mistake to advance the pawns
before the king: 3 5 .a3? 'it>e7 36.<j{b2 'it>d7
37.b4 axb4 38.axb4 c5! 39.<j{b3 (39.bxc5
bxc5) 39 . . . <j{c7 40.<j{a4 cxb4 4 1 .'it>xb4 'it>b8
Black blocks the invasion.
35 . . . 'it>e7 36.<j{a3 'it>d7 37.<j{a4 c6
The best chance, but it is still not good
enough.
37 . . . 'it>c8 38.'it>b5 'it>b7 39.a3 Black is in
zugzwang.
38.a3 'it>c7 39.dxc6 <j{xc6 40.b4 axb4 4 1 .axb4
b5t 42.cxb5t 'it>b6 43.<j{b3
White's outside passed pawn decides the
game.
c) 30.Wf3
White tries to stop . . . ie7 and control the
f-file.
30 . . . Wfl
30 . . . ie7 3 1 .ixh6 g5 32.h4 gxh4 33.g5 h3
34.g6 We8 35 .Wh5 if6 36.ie3 is unpleasant
for Black.
The Prime Years
330
30 . . . 'it>g7 3 1 .ixh6t 'it>xh6 32.Wxf8t 'it>g5
33.Wf2 Wa4 Black faces a difficult queen
endgame a pawn down, but his active king
gives him some drawing chances.
30 . . . We7 3 1 .'it>c2 ig7 32.b3 'it>h7 33.a3 if6
34.b4 axb4 35.axb4 ig5 Black holds.
3 1 .Wxf7t 'it>xf7
The bishop endgame is not much fun for
Black, but he should be able to hold it.
30 . . . b5? is feeble: 3 1 .c6 We8 (3 1 . ..'lWf7
32.ia5) 32.We3 (or 32.ia5) 32 . . . Wf7 (32 ... g5
33.'it>c2) 33.'it>c2 There is not much Black can
do to prevent the coming invasion.
30 . . . Wa4?! Activating the queen is insufficient,
although White will have to play accurately to
prove it. There are two main options.
a) 3 1 .c6?!
White fixes the c7-pawn and pushed his
pawn closer to the promotion square. On
the other hand White risks allowing the
position to become so closed as to enable
Black to build a fortress. Even though this
plan is ultimately incorrect, it is nonetheless
interesting to analyse.
3 1 . . .ie7
a
b
c
d
e
f
g
h
32.'it>c2 'it>e8 33.'it>b3 'it>d7 34.'it>a4 c6! 35 .ie3
'it>c7 36.dxc6 'it>xc6
I do not see a way for White to crack his
opponent's fortress.
8
7
a
6
4
3
2
30.c5!
c
d
e
f
g
h
From this position White has two interesting
options.
5
1
b
a
b
c
d
e
f
g
h
Karpov opens the position in order to
increase his chances to invade.
30 ... dxc5
a l l 32.b3
Forcing the queen to declare her intentions.
32 . . . Wd4!
This looks like the most convincing route to
a draw.
32 . . . Wa3 might just be sufficient as well:
33.Wc4 Wc5! (33 . . . b5? 34.Wc l \Wxc l t
35.'it>xc l wins a s the king walks t o b4.)
34.Wxc5 (After 34.Wxc5 ig5 Black's active
queen enables him to hold the balance.)
34 . . . dxc5 35.h4
Anatoly Karpov - Curt Hansen
1995
33 1
a2) 32.i.c3!
This move was given by Karpov in his
annotations. White takes the d4-square away
from the queen.
8
7
6
5
4
L" .
.
///////
3
2
a
b
c
d
e
f
g
h
35 . . . h5! (35 . . . i.xh4? 36.g5!+-; 35 . . . g5
36.hxg5 hxg5 37.i.e3! a5 This stops b4, but
opens the way for the white king. [37 . . . mf7?
38.b4! White breaks through; 37 . . . b5 38.b4!
cxb4 39.i.b6 i.d6 40.i.a5 White wins.]
38.mc2 i>f7 39.md3 i>e8 40.i>c4 mf7
[40 ... md8 4 1 .mb5 i>c8 42.d6+-] 4 1 .i.d2
mf6 42.i>b5 i>f7 43.i>a6 i>f6 44.i>b7
id6 45 .i>c8 i>g6 46.md7 i>f6 47.a4 mg6
48. me6 White wins the g5 pawn and the
game.) 36.gxh5 (36.g5 b5 37.i.e3 i>f8
Black's fortress withstands the pressure.)
36 ... gxh5 37.i.c3 i.d6 38.mc2 i>f6 I see no
way for White to penetrate.
33.�e3 1Mfxe3 34.i.xe3 i.d8
The fortress seems to hold, for instance:
a
b
c
d
e
f
g
h
35.mb2 g5 36.i>a3 mf8 37.i>a4 me8 38.b4
tf6 39.b5 a5
White can sacrifice on a5 but that is not
enough for a win, as the black king gets to b8
in time.
a
b
c
d
e
f
g
h
32 . . . i.g5 ?!
This natural move is an inaccuracy, but not
a fatal one.
There is a more convincing route to a draw:
32 . . . a5! 33.a3 1Mfb3 34.1Mfa6 1Mfd l t 3 5 .i>a2
1Mfc2! 36.1Mfc8 a4! Black secures a perpetual
check.
33.a3!
With the idea of i>a2 and b3.
33 . . . i.f4
33 . . . 1Mfb3? allows 34.1Mfxa6.
33 . . . a5 ? is refuted beautifully:
a
b
c
d
e
f
g
h
34.h4!! (34.i>a2 i.c l ! 35.h4 g5 Black narrow­
ly holds.) 34 . . . i.f4 (34 . . . i.xh4 3 5 .ma2!+-)
3 5 .ma2 i.c l 36.g5! h5 37.1MffI +34.i>a2!
The Prime Years
332
White sets up a wicked trap.
34 . . . �f6?
34 . . .�g5? 35 .�b4 �e7 36.b3 'MIb5 37.'MIxb5
axb5 3B.�d2 is a winning ending.
Correct is 34 . . . �f7! 35 .g5 (35 .b3 'MIb5)
35 . . . h5 36.h4 'MIb5 and it looks like Black
escapes.
37 . . . �f4
37 . . . �xh4 3B.'MIg4 'MIb5 39.'MIxh4t wins.
3B.'MIg4 'MIb5 39.'MIe6t �fB 40.'MIcBt c;i{gl
4 1 .'MIxclt mf6 42.'MIxd6t
White wins. However, I must re-emphasize
that this beautiful win only arose after some
inaccuracies on Black's side. Therefore if White
is to find a winning continuation he must
deviate at an earlier stage.
b) 3 1 .cxb6!
White requires a bit more open space if he is
to win the game. The winning method was
shown by Karpov in his annotations.
3 1 . . .cxb6
8
a
b
c
d
e
f
g
h
3 5 .g5t!!
With this lovely move White opens the path
for his queen to invade.
35 . . . �xg5
35 . . . hxg5 36.b3 'MIb5 37.'MIg4 'MIn 3B.'MIe6t
�g7 39.'MId7t mh6 40.'MIcB 'MIf2t 4 1 .�b2
g4 42.'MIhBt mg5 43.h4t wins.
36.�b4!
Blocking the queen's path to the c4-square,
and hinting at a possible sacrifice on d6.
36 . . . �e7
a
b
c
d
e
f
g
h
37.h4!!
37.'MIg4 'MIb5! Black gets enough counterplay
for a perpetual.
7
6
5
4
3
2
a
b
c
d
e
f
g
h
32.b3! 'MId4 33.'MIe3!
White traps the queen in order to exchange
it. The resulting same-coloured bishop
ending is winning for White, as Black will
be forced to weaken his queenside pawns.
33 . . . 'MIxe3
33 . . . �f6 34.'MIxd4 exd4 3 5 .�f4 wins.
34.�xe3 b5
Had Black been able to put his bishop on
dB, he would have been able to draw, but
now the queenside weakening proves fatal.
35 .�c2 �e7 36.mc3 �h4
After 36 . . . �dB 37.mb4 �c7 3B.a4 bxa4
39.mxa4 White invades.
37.�d2!
Preventing the check on e 1 .
37 . . . �f2
Anatoly Karpov - Curt Hansen
1 995
37 . . . i.d8 38.c;i{b4 h5 39.a4+38.c;i{b4 i.b6 39.a4 bxa4 40.�xa4!
This winning line was pointed out by
Karpov himself, although interestingly he only
evaluated it as "±" whereas in reality the final
position is just winning for White.
333
34 . . . �g8 3 5 .Wff6 dxe4 36.Wfxg6t i.g7
37.c;i{c 1
Black's situation looks desperate, but he has
a narrow path to survival.
8
7
31.Wlxa6
6
5
4
3
2
a
a
31...h5
b
c
d
e
f
g
h
Hansen decides to embark on kingside
counterplay.
3 1 . ..c6! ?
This move should also suffice for a draw,
although Black will have to find a few
difficult moves along the way.
32.i.c3
If 32.Wfd3 cxd5 33.exd5 Wfa4 Black holds.
32 ... cxd5 33.i.xe5t �f7!
33 . . . c;i{h7 34.Wfxb6 dxe4 35 .Wfc7 White's
passed a-pawn decides the game.
34.'!Wfl t
Here is another nice line: 34.'1Wxb6 '!We7!! In
his analysis Karpov misses this nice defensive
move. (If 34 . . . dxe4 35.'!Wf6t c;i{e8 36.'1Wxg6t
\tJd8 37.'1Wb6t c;i{e8 38.Wfb8t White exchanges
queens and wins.) 35 .'!Wc7 (Or 35 .'!Wb8 i.g7
36.i.xg7 Wfxe4t and Black is safe.) 35 . . . Wfxc7
36.i.xc7 dxe4 It is unlikely that White's
advantage will be enough to win the game.
b
c
d
e
f
g
h
37 . . . e3!
37 . . . Wff7? 38.Wfxf7t c;i{xf7 39.i.xg7 c;i{xg7
40.�d2 �f6 4 1 .h4! �e5 42.�e3 wins.
38.Wfxg7t
38.Wfd6 e2 �d2 Wfa4 draws.
38 . . . Wfxg7 39.i.xg7 �xg7 40.c;i{d 1 �f6 4 1 .h4
4 1 .c;i{e2 c;i{g5 42.�xe3 �h4 43.c;i{f4 �xh3 is
a simple draw.
4 1 . . .c;i{e5 42.�e2 c;i{f4 43.g5 hxg5 44.hxg5
�xg5 45.�xe3 �f5
Black's king gets back to the queens ide in
time.
a
b
c
d
e
f
32.Wle2 hxg4 33.Wlxg4 Wlxg4?!
g
h
The Danish grandmaster makes a poor
The Prime Years
334
practical decision, notwithstanding the fact
that the position remains a draw with perfect
play. Inferior same-coloured bishop endgames
are extremely difficult to hold over the board.
33 . . . Wb5! would have maintained a more
comfortable drawing margin. By keeping
the queens on the board, Black prevents his
opponent's king from moving freely. 34.Wf3 c6
35 .h4 (35 . .tg5 c;t>g8) 35 . . . cxd5 36.exd5 Wa4
White king is too open to win, for instance
37.h5 gxh5 38.Wxh5 We4t 39.c;t>al c;t>g8 and
the draw is in sight.
insufficient: 37 . . . c;t>d7 38.c;t>b3 c6! 39.c;t>c4
a4 40 . .te3 cxd5t 4 1 . \t>xd5 c4 42. c;t>xc4 c;t>c6
Black holds.
37 . . . \t>d7
It looks unpleasant to have the c5- and e5pawns on dark squares, but White is unable
to attack them both at the same time.
38.c;t>c2
34.hxg4 �f7 35.a4!
Karpov wants to exchange his a-pawn for
the opposing b-pawn before he plays b4. In
his subsequent annotations he showed why
this was necessary: 35 .c;t>c2 \t>e8 36.b4 cxb4
37.\t>b3 \t>d7 38 . .txb4 .txb4 39.\t>xb4 c6!
White cannot invade.
35 ... �e8 36.a5
8
7
6
5
4
3
2
1
a
36 ... �d7
b
c
d
e
f
g
h
36 . . . bxa5!?
Despite its ugly appearance, this was also a
valid defence.
37 . .txa5
37 . .tc3 .tg7 does not change much.
37.c;t>c2!? is interesting, but ultimately
a
b
c
d
e
f
g
h
From here Black must select the correct
drawing plan.
a) 38 . . . c6?
This tempting move is in fact a decisive
mistake.
39.c;t>d3!
39.dxc6t c;t>xc6 40.\t>d3 .th6 (40 ... c;t>b5
4 1 . .tc7 .tg7 42.g5 c;t>c6 43 . .tb8 c;t>d7 44. c;t>c4
\t>c6 45 .b3 .th8 46 ..ta7+-) 4 1 .\t>c4 .te l
42.b3 .te3 Black should be able to draw.
39 . . . .th6
39 . . . cxd5 40.exd5 \t>d6 4 1 . c;t>e4 .tg7 42.b3
.th6 43 ..tb6 .tf4 44 ..ta7 wins.
39 . . . c;t>d6 40.\t>c4 cxd5t 4 1 .exd5 Black has
no good defence against .tb6.
40.c;t>c4 .te3
40 . . . \t>d6 4 1 ..tb6 .te3 42.b4+4 1 ..tb6 .td4 42.b3 \t>d6 43 ..ta7!
Black is in zugzwang, for instance:
43 . . . cxd5t 44.exd5 e4 45 .c5+b) 38 . . . .th6!
By leaving his pawns in place, Black prevents
Anatoly Karpov - Curt Hansen
1995
the enemy bishop from utilizing the b6square, which should enable him to draw.
a
b
c
d
e
f
g
335
37.axb6 cxb6
h
39.md3 ic l 4o.ic3 md6 4 1 .mc4 ie3
From this position I have not been able to
find a winning plan for White, although he
can set a devious trap.
42.h3!
42.mb5 id4! draws without too many
problems.
a
38 ..tc3
b
c
d
e
f
g
h
The immediate 38.b4? is premature:
38 . . . c4 39.b5 ic5 40.mb2 id4t 4 1 .ma3 c3!
The position is equal as pointed out by Karpov.
If White takes the c3-pawn then the pawn
ending is an easy draw.
38 ....td6
It is not easy for Black to decide how to
defend the e5-pawn. There were two other
possibilities.
a
b
c
d
e
f
g
h
42 ... c6!
Black must act promptly as after 42 . . . if2?
43.mb5! White has a strong plan of invading
with his bishop on the kingside with the aim
of getting to the f8-square.
43.ia5
Trying to set up the same winning plan as
in line a) above, but here Black has enough
time to arrange a better defence.
43 ... ic l ! 44.ib6 ia3
White has no way through. (44 . . . cxd5t
45.exd5 ia3 is also fine.)
38 . . . md6?!
This move is playable in itself, but it does
not make much sense as in order for Black to
draw, he will need to use a defensive method
shown in the notes to the main game.
Attempts to use the king actively only lead
to disaster.
39.b4! c4
39 . . . ig7 transposes to 38 . . . ig7 as analysed
below.
40.b5 mc5?
Black is in no position to challenge his
opponent to a pawn-race.
Instead he should prefer 40 . . . ie7! 4 1 .ma2
if6 42.ib4t md7 when his fortress should
hold firm. Compare 39 . . . c4! ? in the notes to
the main game.
The Prime Years
336
to Black's 39th move in the main game.) 42.g5!
'kf1b5 43 . .tb2! White wins by dominating the
enemy bishop.
Instead Black should prefer 40 . . . .tf6!
followed by giving up a pawn on the queens ide.
This should enable him to draw using the
method shown in the note to Black's 52nd
move in the game.
a
b
c
d
e
f
g
h
4 1 . .txe5 'kf1xb5 42. <j;>c2 <j;>c5
42 . . . 'kf1a6 43 . .tf6 <j;>b7 44.e5 <j;>c7 45 .'kf1c3
b5 46.d6t <j;>d7 47.<j;>d4 <j;>e6 4B.g5!+43. <j;>c3 <j;>b5
43 . . . b5 44 . .td4t! <j;>d6 45.<j;>b4+44 . .tf6 .tb4t 45.'kf1c2 .tfB 46.e5 <j;>c5 47.d6
White's pawns are too strong.
3B . . . .tg7! ?
Like the game continuation, this leads to a
difficult endgame for Black, but one which
can be drawn with correct defence.
39.b4!
If39.<j;>c2 b5! Black draws more comfortably
as he is ready to block the queenside.
39 . . . <j;>d6
b) 40.g5
I would evaluate this move as slightly less
precise than the immediate capture on c5,
although it does set one particularly devious
trap.
40 . . ..tfB!
40 . . . b5 4 1 .bxc5t 'kf1xc5 reaches line b)
below.
Here is the banana skin which Black must
avoid: 40 . . . c4? 4 1 .b5! 'kf1c5 (4 1 . ..'kf1d7
42 . .tb4+-) 42.d6!! .tfB (42 . . . 'kf1xd6 43 . .tb4t
<j;>e6 44.<j;>c2 Black will soon succumb to
zugzwang.) 43 . .tb2! 'kf1xb5 44 . .ta3 White
prevails, as d7 will win the black bishop.
4 1 .bxc5t bxc5 42.<j;>c2 .te7
8
7
6
5
4
3
2
a
a
b
c
d
e
f
g
h
White has a few possibilities from this
position.
a) 40. <j;>c2!? sets a small trap: 40 . . . c4?! 4 1 .b5!
'kf1c5? (4 1 . . . .tf6! should still draw - see the note
b
c
d
e
f
g
h
The attack on the g5-pawn ties up White's
bishop. The first player can still press, b ut it
is doubtful that he can win.
43 . .td2 'kf1c7 44.'kf1b3 <j;>d7
44 . . . 'kf1b6 4 5 .<j;>c4 .tfB 46 . .te l .te7 47.ig3
.td6 4B . .th2 <j;>a5 ? 49 . .txe5 White wins.
45.<j;>c4 <j;>d6 46.<j;>b5 <j;>c7 47 . .te3 <4id7
4B.<j;>b6
Anatoly Karpov - Curt Hansen
1 9 95
White wins the c-pawn, but the ending
should still be a draw. A similar situation arises
in the main game, but with the white pawn
on g4 instead of the more exposed g5-square.
Black is drawing there too, so the exposed
pawn on g5 can only help him.
337
5 1 .@d4 @xg5 52.e5 @h6 53.@d5 White
wins.
44.@b3 i.h8
44 . . . @c5 45.@a4+45.d6 @c6 46.@xb4 @xd6
46 . . . i.g7 47.@c4 @xd6 48.i.a3h-
b) 40.bxc5t!
This looks like a better winning attempt.
Now Black must make a decision.
a
a
b
c
d
e
f
g
h
40 . . . <;t>xc5 ?
This allows White to win in a most instruc­
tive fashion.
40 . . . bxc5! should draw, as long as Black is
willing to sacrifice his c-pawn. If he tries to
preserve material equality for too long then
he loses, for instance 4 1 .i.d2 i.f6 42.@c2
1h4 43.@d3 i.f2 44.@c4 i.gl 45 .@b5 i.d4
46.i.h6 @e7 47.@c6+-. Instead he should
play . . . c4 at some point, reaching the same
type of endgame as in the note to Black's
52nd move in the main game.
4 1 .g5!
After 4 1 .@c2 i.f6 42.@b3 b5 White cannot
win as his bishop has to guard the d4square.
4 l ...b5 42.@c2 b4 43.i.b2 @b5
43 ... @b6 44.@b3 @b5 also loses, albeit by a
single tempo: 45.d6! i.f8 (45 . . . @c6 46.@xb4
mxd6 47.@c4 [or 47.i.a3l 47 . . . @c6
48.i.a3+-) 46.i.xe5 @c6 47.@xb4 i.xd6t
48.1xd6 @xd6 49.@c4 @e5 50.@d3 @f4
b
c
d
e
f
g
h
47.@b5
White wins thanks to the extreme passivity
of the enemy bishop.
47 . . . i.g7 48.i.a3t @d7 49.i.b4 i.h8
49 . . . @c7 50.@c4 @c6 5 1 .i.c5+50.i.f8 @e8 5 1 .i.h6 @d7 52.@c5 @c7 53.@d5
@d7 54.i.f8 @e8 5 5 .i.d6 i.g7 56.@e6
White invades and wins.
39.M!
a
b
c
d
e
f
g
h
White has to act quickly before Black gets a
chance to play . . . b5. Now the defender must
make a difficult decision.
338
The Prime Years
39 ... c:xb4
If Black is to draw the game then he must
build a fortress. The question is whether he
should try to do it with or without b-pawns
on the board. If he keeps them on the board,
then it will be harder for White to invade on
the queens ide as the position is closed. On
the other hand it will mean having another
potentially vulnerable pawn on a dark square.
Having analysed the position I believe Black
can draw with either approach, but taking on
b4 seems to be the slightly more straightforward
way.
39 . . . c4!? 40.b5
The c4-pawn is doomed, but the blocked
nature of the position makes it hard for the
white king to penetrate.
40 .. .';:ge7 4 1 .�c2 �f6 42.�d2 �c5 43.'it>c3
�d4t 44.'it>xc4 �c5
Crucially the pawn ending will be drawn
as long as the black king can reach the d6square in time.
45 . . . �d6
45 . . . g5 ? 46.d6 �e6 47.�d2 wins.
Quite incredibly Black seems to be drawing
this endgame; at least I have not found any
way to set up a zugzwang or invade.
46.�e l 'it>e7 47.�f2 �c7 48.�e3
48.�h4t �d7 49.�f6 �b8 50.�d3 'it>d6
defends. The attempt to invade on the king­
side leads nowhere: 5 1 . �e3 �c5 52. 'it>f3?!
�d6! (but not 52 . . . �xb5?? 53.�e7!+-) By
now it is White who must work for a draw.
48 . . . 'it>d7 49.�d3
49.d6 'it>xd6 50.�g5 �e6
49 . . . 'it>d6 50.�e2 'it>e7 5 1 .'it>f3 �f7 52.'it>g3
�d8
White has not made any progress.
40.,tdl q;c7 41.Q;c2 b5 42.Q;b3 Q;d7
43 ..bb4
Karpov collects the pawn. It is likely that
Hansen overlooked the zugzwang that follows,
nevertheless the result still hangs in the
balance.
43 ...,tc7 44.,td ,td6
a
b
c
d
e
f
g
h
45 .�c3
45 .�b4? �xb4 46.'it>xb4 �e7 is a draw.
45 .d6!? Sacrificing the pawn is a nice idea,
but it seems Black can defend here too.
45 . . . �e6 46.d7 (46.�h6 �xd6) 46 . . . �e7!
Black simply refuses to take the pawn.
47.�c l �d8 48.�e3 �d6 (48 . . . �c7?? loses
to 49.�g5 ; 48 . . . g5! ? 49.�d2 �e7 looks
playable though.) Black holds the endgame.
a
45.,tb2!!
b
c
d
After this superb
zugzwang.
45 ... b4!
e
f
g
h
move Black is in
Anatoly Karpov - Curt Hansen
1995
Hansen does not panic, but finds the best
defence and gives up the b-pawn in order to
play for a fortress, which happens relatively
often in same-coloured bishop endings.
45 ... WcB was not helping, and after 46 ..ta3
1c7 47 ..tfB! White wins the b5-pawn.
Unfortunately the pawn endings are hopeless
for Black. For instance: 45 . . . Wc7 46 . .ta3!
1xa3 47.'iftxa3 g5 4B.'iftb4 'iftb6 49.d6 Wc6
50.d7 Wxd7 5 1 . 'iftxb5 White wins trivially.
46.ic1 ic5 47.idl ig! 48.<i>xb4 ifl
49.<i>c4 ig!
339
Hansen commits a decisive error. Incredibly
it turns out that the black bishop does not have
enough squares available on the queenside.
52 . . . .th4! was necessary, as the bishop has
enough space on the kingside to avoid a
zugzwang scenario. When Hansen wrote
about this game in Skakbladet, a Danish chess
magazine, he mentioned that Karpov showed
him this move immediately after the game.
53 . .tc5 .tg5 54.d6 'iftc6 55 .d7 .tdB 56 . .ta7!
White must force his opponent to take the
d7-pawn. (56 . .tb4 .tc7!? holds) 56 . . . 'iftxd7
57.Wd5 .tf6 5B . .tbB .tg7 59 . .txe5 .th6
8
7
6
5
4
3
a
2
a
50.<i>d3!
b
c
d
e
f
g
h
b
c
d
e
f
g
h
Karpov evaluates this ending as slightly
better for White in his annotations. Indeed,
there does not seem to be any way for White
to win the g6-pawn.
White is clearly unable to drive his opponent's
king back, so he must work to restrict the
enemy bishop. Karpov starts by taking away
the a7-g1 diagonal.
Karpov fixes the black bishop to the
queenside, and soon manages to trap it.
50...ic5 51.ie3 i.e7 52.<i>c4
53 ... ie7 54.i.c3 id6
52.g5 With the bishop on e7, this method
of restriction is ineffective: 52 . . . Wc7 53. Wc4
<j;>d7 54 ..td2 .tdB 5 5 .'iftb5 'iftd6 56 . .te3 (or
56 ..tb4t Wd7 57 . .tc3 Wd6) 56 . . . 'iftd7 57.Wc5
1e7t Black succeeds in blocking the invasion
attempts.
52 ...ia3?
53.idl!
If 54 . . . .tf6 5 5 .g5! .tg7 56 . .tb4! White traps
the bishop and wins easily: 56 . . . Wc7 57.'iftb5
Wd7 5B.Wb6 .thB 59 . .tfB 'ifteB 60.'iftc7 WxfB
6 1 .d6+-
55.ib4! ib8
Alternatively 55 . . . .tc7 56.'iftb5 .tbB 57 . .tc5
.tc7 5B.d6 .tdB 59 ..ta3 .th4 60.'iftc5 wins.
The Prime Years
340
56.c.t>b5 i.a7
Allowing a pretty finish. Instead if 56 . . . �c7 57.�f8 �d8 58.d6 �g5 59.Wc5 wins.
8
7
6
5
4
3
2
57 ..ic5 .ib8 58.d6!
1-0
a
b
c
d
e
f
g
h
Karpov beautifully traps Black's bishop. This was a highly instructive game, from which one
can learn a lot about same-coloured bishop endgames. Here is a good piece of advice: if you
have a worse position and are faced with a choice of whether to go for a same-coloured bishop
endgame, then try to avoid it! Very often the inferior ending will turn out to be losing, and even
if a sophisticated drawing method does exist, it is always hard to withstand the constant pressure
over the board with a ticking clock.
In Round 8 equalized with the 4 . . . l2J d7 Caro-Kann against Almasi. Gradually he gained the
upper hand, and Almasi unnecessarily sacrificed a pawn. The Hungarian grandmaster defended
stubbornly, but Karpov eventually won the endgame. In the next game Karpov played super­
solidly against Svidler's GriinfeId and the game never looked like being anything other than a
draw.
In Round 1 0 Karpov played the Queen's Indian against Van WeIy, who responded with the
Petrosian System. Following an early queen exchange White had a slight edge thanks to his extra
space, but Karpov gradually organized his position and equalized. White sacrificed a pawn and
got decent compensation, but Karpov gradually neutralized his initiative and won with his extra
material. In the final round Karpov only needed a draw to secure outright first place, so with
the white pieces against Peter Leko, he played l .d4 and offered a draw which was accepted. He
finished with 7Y2/ 1 1 , half a point clear of Sokolov and Kamsky.
34 1
1995 Summary
Spanish Team Championship: 1 Y2/3 (+0 =3 -0)
EU Cup: 1 1 1 (+ 1 =0 -0)
Baden-Baden knockout ( l st place) : 7/8 (+6 =2 -0)
FIDE candidates semi-final match versus Gelfand, Sanghi Nagar: Won 6-3 (+4 =4 - 1 )
Linares (2nd place) : 9/ 1 3 ( + 5 =8 -0)
Dos Hermanas (2nd-3rd place) : 5 Y2/9 (+3 =5 - 1 )
Dortmund (2nd place) : 6Y2/9 (+4 =5 -0)
Groningen ( l st place) : 7Yz/ l l (+4 =7 -0)
Total 69.8% (+27 =34 -2)
i!!Il Wins • Draws • Losses
199 6
Rating 2770 (3 in the world)
Karpov began the year ranked third in the world, as he was narrowly overtaken by Kramnik.His
first event of the year was the Spanish Team Championship. He started with two solid draws,
with Black against Campora in a Caro-Kann, and White against Rogers in a Bogo-Indian. In
the third round he managed to grind out a win from an almost dead-looking position against
Cramling. In the fourth and final round Karpov defeated Gulko with the help of a nice exchange
sacrifice, as shown in the notes to Game 40.
Karpov played a short, four-game match against the Yugoslavian (now Montenegrin) grandmaster
Bozidar Ivanovic. The match took place in Podgorica, the capital of Montenegro. In the first
game Ivanovic played the Panov against the Caro-Kann. At one point he declined a move
repetition, and a complex position occurred with hanging pawns. Ivanovic sacrificed a pawn but
then immediately made a tactical oversight, which allowed Karpov to win a piece and the game.
Karpov won the second game. The third game was a quick draw in the Panov, and the fourth
game was also drawn. Games 2 and 4 are not shown in the database. The final score of 3- 1 was
a decent result for Karpov.
According to the database, in 1 996 Karpov played three games in the Yugoslav Team Championship,
two of which were against Beliavsky. In the first game Karpov was a bit worse on the black side of
a Catalan, but managed to hold the draw. In the second Beliavsky played the Tartakower System
in the QGD. He got close to equality and later sacrificed a pawn for some activity. For a while
he had sufficient compensation to maintain the balance, but he made a mistake in the endgame
and Karpov punished him. Karpov's other game was against Kiril Georgiev, who played a Benko
Gambit. Karpov held onto the extra pawn but was unable to make any headway in the major
piece endgame, and had to settle for a draw.
Karpov also took part in a tournament in the same region, in Belgrade. His first opponent was
Salov, who also tried the Benko Gambit. Once again Karpov held onto the extra pawn, but
eventually he had to return it in order to neutralize his opponent's queenside pressure. The
resulting position was equal and the players agreed a draw. In the second round Karpov had the
white pieces again. His opponent was the Serbian grandmaster Zlatko Hincic, who only played
this one game against a world champion.
The Prime Years
344
I Game 4l 1
Anatoly Karpov - Zlatko Ilincic
Bel grade 1996
l .tLla
<;f;>h7 although Black remains somewhat worse
here too. White can choose between the calm
1 6.a3 lOa6 1 7.Elab 1 and the more ambitious
1 6.c5!? cxd5 1 7.exd5, taking advantage of the
fact that 1 7 . . . dxc5 ?! I S .d6 is strong.
15 ..hf3 e5
Karpov probably chose this move order
because his opponent sometimes played the
Griinfeld.
1...tLlf6 2.e4 g6 3.tLl c3 �g7 4.e4 d6 5.d4 0-0
6.i.e2
Karpov plays the traditional main line for
the first time since his 1 9 9 1 match against
Kasparov. In 1 996 he played this variation four
times, scoring two wins and two draws.
6... e5 7.0-0 tLla6
Hincic had played this line before, so Karpov
probably anticipated it.
Later in the year Kamsky chose the main
line: 7 . . . lO c6 S.d5 lO e7 9.b4 lOh5 1 O.Ele l h6
1 1 .1Od2 1O f4 12 ..tfl a5 1 3.bxa5 Elxa5 1 4.lOb3
ElaS 1 5 .c5 f5 1 6.cxd6 cxd6 1 7.lOd2 g5 1 8 .Elb l
g4 1 9.\Wb3 fxe4 20.lO dxe4 The strong knight
gives White a positional advantage and he won
convincingly, Karpov - Kamsky, Elista (7)
1 996.
8J�el e6 9.�fl i.g4 1O.d5 tLlb4 11.i.e2 a5
12.�g5
Karpov provokes the following pawn move,
which weakens Black's kingside slightly.
12 ... h6 13.i.e3 \We7?!
This move was a novelty at the time and has
not been repeated. 1 3 . . . i.d7 and 1 3 . . . c5 are
the normal moves.
14.h3 .hf3?
Giving up the bishop results in a passive
position. More logical was 1 4 . . . .td7 1 5 .\Wd2
16...\We7 17.\Wd2
Having provoked the weakening . . . h6 earlier,
Karpov wins a tempo.
17 ... i>h7 18.l3ae1 tLla6 19.i.dl tLle7 20.g3
tLld7
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Anatoly Karpov - Zlatko Ilincic
1996
345
21.ic2
25 ... ttJ xc3
Karpov improves his bishop, which now
stands on the same diagonal as the enemy king.
Now he is ready or almost ready (he could still
improve his king slightly by placing it on g2 or
h2) to advance his f-pawn.
25 . . . ttJ b6 can be met by 26.Wd3 intending
f4.
21...tLJb6
Ilincic wants to keep Karpov busy on the
queenside.
8
7
6
5
4
22.id3 �d7 23.@g2 llJa4
Ilincic continues to play sensibly. Having less
space, it makes sense for him to exchange a pair
of pieces. By exchanging White's last remaining
knight, he also increases the potential value of
his own knight.
3
2
1
a
26.bxc3!
b
c
d
e
f
g
h
Karpov accepts a structural weakening, having
correctly j udged that the benefits associated
with the open b-file are far more significant.
Besides, after 26.Wxc3 b5! Black liberates his
position and obtains some counterplay.
26 ... a4
a
24J�bl!
b
c
d
e
f
g
h
Karpov reacts t o the new situation and
prepares to fight for the b-file. It is instructive
to observe how quickly and effectively Karpov
changes his focus from one flank to the other.
Avoiding exchanges with 24.ttJdl would have
allowed Black to obtain some counterplay with
24 ... b5!.
24.. J�fb8 25.ic2
Karpov forces his opponent to make a
decision regarding the a4-knight.
Black had to make a difficult decision
between several somewhat difficult defensive
options. With the text move Black prepares
. . . b5, but Karpov never gives him a chance to
play it.
The immediate 26 . . . b5? is impossible due to
27.cxb5 ttJxb5 28.ia4 with a lethal pin.
26 . . . b6!
This would have been the best chance. Black
prevents the rook invasion and prepares . . . b5
more slowly.
27.f4
27.a4 is quite a radical measure. White
prevents
his
opponent's
queens ide
counterplay, but also limits his own active
prospects on that flank. White still has
the advantage, as he will develop a lasting
The Prime Years
346
InItIative on the kingside. Nevertheless
Black's defensive task is simpler than in the
game; his b-pawn demands only minimal
attention, and most of his pieces will be able
to partake in the defence of the kingside.
27 . . . a4
27 . . . exf4 2B.gxf4 bS 29.cxbS tLlxbS 30.�a4
�xc3 3 1 .Wd3 �b4 Black is surviving for the
moment, but the pin is most awkward for
him.
on its starting square, where it may become
vulnerable. The rook move also makes f4 more
dangerous, as . . . exf4 might be met by ixf4
hitting the d6-pawn.
The immediate 27.f4? is a mistake due to
27 . . . exf4 2B.gxf4 bS! When Black has decent
counterplay.
27.. J�a7
Hincic prepares to push the rook away;
perhaps he still hoped to carry out . . . bS.
a
b
c
d
e
f
g
h
From the present position Black certainly
stands worse, but at least he will be able to
generate some activity with . . . bS. It would
have been interesting to see whether Karpov
would have chosen to gain space with 2B.fS or
look for play on the f-file after 2B.fxeS dxeS.
27 ... Ela6 deserved attention. 2B.Elxa6 (After
2B.Elfb 1 Elxb6 29 .Elxb6 White is still better, but
there is no doubt that the exchange of a pair of
rooks eases Black's defensive task.) 2B . . . bxa6
(2B . . . tLlxa6 29.Wd l a3 30.We I White wins a
pawn.) 29.Wd l a3 (29 . . . Elb2? 30.�xa4 Wie7
3 1 .�b3 White traps the rook.) 30.�a4 Wid8
3 1 .�e I tLl aB 32.�xa3 tLl b6 33.�b3 White has
an extra pawn but it is doubled. It is hard to
say whether this or the game continuation gave
better drawing chances.
28JUbl
White's pressure on the b-file will not win
the game by itself, but it is useful to tie up the
enemy pieces for a few moves.
28".lLla8 29J�6b5 b6
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Anatoly Karpov - Zlatko Hindc
1 996
30.f4!
Karpov gains space and switches side once
again. White now threatens to win a pawn by
taking on e5.
30 ... f6?!
The Serbian grandmaster sets up a wall of
pawns, but weakens his light squares severely.
30 . . .1''1: e S was a better defensive move. Black's
position remains difficult, but at least the
O-pawn helps to guard a few light squares. I
suspect that Karpov would have pushed with
3 1 .f5 all the same.
347
By limiting his opponent to an open rook's
file, Black tries to maximize his defensive
chances.
Another idea was 32 . . . g4 33.h5 2"1:gS, intending
to give up the g-pawn and build a fortress,
although the chances of this plan succeeding
appear remote to say the least.
33.�hl i>g8
8
7
6
5
4
3
a
2
31.£5!
34.�bbl!
a
b
c
d
e
f
g
h
Karpov gains more space and opens up the
light squares on the kingside.
31. .. g5
Other moves would have been equally
depressing for Black. After the text move, the
situation on the kingside strongly resembles
that from Karpov's classic win over Unzicker
from 1 974 - see Game 37 of the first volume.
32.h4
Karpov prepares to open a file in order to
invade.
b
c
d
e
f
g
h
Karpov exploits his extra space by preparing
to switch his attention to the kingside.
34 ... �a6?!
Hincic probably wanted to defend the b6pawn in order to free the as-knight and transfer
it to the kingside, but now the rook is severely
misplaced. Other moves would also have failed
to save the game, for instance:
34 . . . WeS 35.id l
White improves his light-squared bishop.
35 . . . 2"1:b7 36.Wc2 2"1:a7 37.if3 WcS
37 . . . Wfl 3S.We2 The bishop goes to h5
next.
3s.ih5 Wa6
3S . . . WfS 39.ig6 Wa6 40.Wd3 White will
soon execute a sacrificial breakthrough:
40 . . . 2"1:b7 (40 . . . LOc7 4 1 .hxg5 hxg5 42.ixc5!!
White cracks the enemy position from
The Prime Years
348
an unexpected angle. 42 . . . dxc5 43.d6+-)
4 1 .hxg5 hxg5 42.1'!xhSt i.xhS 43.1'!h l i.g7
44.i.xg5! fXg5 45.f6+8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
39.i.eS!
It is remarkable how White can use the
whole board including the back rank.
39 . . . lLl c7
39 . . . WExc4? 40.i.b5 White traps the queen in
a highly unusual way.
40.i.c6 'tt> fS 4 1 .WEe2 WEa5 42.hxg5
There is a remarkable alternative: 42.WEb2
1'!a6 43.WEb4!! White invades after this
spectacular queen exchange.
42 . . . hxg5
42 . . . fXg5 43.f6 i.xf6 44.WEf3+43.1'!xhSt i.xhS 44.WEh5 i.g7 45 .i.xg5 ! fXg5
46.f6
Black can resign.
35J:�bfl
Karpov keeps playing strategically and
prepares to sacrifice on g5 .
His chosen continuation is good enough
to win the game, but he could have broken
through more quickly with: 35.hxg5! hxg5
36.i.xg5!! The immediate bishop sacrifice
opens up Black's kingside with decisive effect.
36 . . . fXg5 37.WExg5 lLl c7 (37 . . . WEf7 3S.1'!xh8t
'tt> xhS 39.1'!h l t 'tt> gS 40.1'!h6 'tt> fS 4 1 .1'!xd6+-)
3S.1'!xhSt 'tt> xhS 39 .1'!h I t 'it>gS 40.f6 ltJe8
4 1 .fXg7 WExg7 (4 1 . . .lLlxg7 42.WEg6+-) 42.Wf5
1'!a7 43.i.xa4 White wins.
35 ...'?Nd8
It was essential to prevent the sacrifice on
g5 , as shown by the line 35 . . . lLl c7 36.hxg5
hxg5 37.1'!xhSt i.xhS 3S.i.xg5 and White
wins.
35 . . . 1'!a7 would not have changed much:
36.hxg5 hxg5 37.i.xg5! fXg5 3S.WExg5 1'!xh l
39.1'!xh l WEe7 40.WEg6 WEf6 4 1 .1'!h6 White gets
too many pawns for the piece.
a
36.'?Ndl!
b
c
d
e
f
g
h
Karpov prepares to use his queen on the
light squares, while incidentally attacking the
a4-pawn.
a
b
c
d
e
f
g
h
36 ... a3
Anatoly Karpov - Zlatko Hincic
1996
Defending the pawn with the queen would
have left the kingside too exposed: 36 .. :�e8
37.hxg5! hxg5 38.:gxh8t �xh8 39.�xg5!
White demolishes his opponent's defences.
39 . . . fxg5 40.f6! :ga7 4 1 :�g4 The attack is
devastating.
349
8
7
6
5
4
3
2
1
a
43.i.xg5!
b
c
d
e
f
g
h
A cute finish, although it was not hard to
find as two of Black's pieces are completely out
of play on the queenside.
a
37.�a4!
b
c
d
e
f
g
h
Karpov uses the whole board to convert his
advantage. The less instructive 37.hxg5 hxg5
38.:gxh8t �xh8 39:�h5 was also winning,
but improving the bishop first will make the
inevitable kings ide breakthrough even more
devastating.
37 ... tDc7 38.i.c6
From here the bishop more or less traps the
rook on a6 for the rest of the game.
38 ... i>f7 39.VNh5t i>g8 40J'�bl
Karpov sets up another threat of winning the
b6-pawn with �b7. It was not the only way to
win, nevertheless it is instructive to observe the
way Karpov stretches the defence.
40 ... VNb8 41 .VNg6
Having distracted the enemy queen, Karpov
invades with maximum force.
41. ..VNfS 42.hxg5 fxg5
Unsurprisingly there were other routes to
victory, including the equally attractive
43.�xc5! when 43 . . . dxc5 44.d6 and 43 . . . bxc5
44.:gb7 are both crushing.
43 ... hxg5 44J'hh8t i>xh8 45.:ghl t i>g8
46.VNh7t i>f7 47.f6 VNh8 48.VNf5
1-0
In the next round Karpov faced Bareev, who
used the Petrosian Variation against the
Queen's Indian. Karpov's king settled on e7
in the middlegame, but his position remained
playable. But later Karpov made a mistake
and got into trouble. Bareev won a pawn and
converted his advantage efficiently in the queen
endgame. In Round 4 Karpov faced Popovic,
who played a Griinfeld. Black equalized and
at one point had two extra pawns on the
queens ide, but Karpov had enough activity to
maintain the balance and a draw ensued. In
the final round Leko went for safety against
the Caro-Kann, and a draw was soon agreed.
Karpov's final score of 2Y2/5 was enough to
share third place behind Leko. Bareev took
sole first place, a point and a half in front.
The Prime Years
350
FIDE World Championship match
versus Kamsky
It had been three years since Karpov won the
'official' version of the World Championship,
following the decision of Kasparov and Short
to break away from FIDE, and now it was
time to defend his title against Gata Kamsky.
It was an important and prestigious event,
notwithstanding the fact that Kasparov was
undeniably the strongest player in the world.
Karpov was the favourite as he had a +5 =4
-2 record against the Russian-born American
grandmaster, but Kamsky had been improving
steadily, and had developed into a formidable
grandmaster, an especially stubborn fighter
with excellent nerves. Since their 1 992
encounter (Game 2B in the present book) ,
Karpov had scored two victories to Kamsky's
one, with three draws.
Any world title match is a historic event in
the chess world, but the present match was
especially notable for being the last such match
to feature adjournments. The match was held
in Elista, the capital of Kalmykia. Both players
enlisted American grandmasters as their chief
seconds; Karpov was assisted by Ron Henley
and Kamsky by John Fedorowicz.
In Game 1 Karpov surprised his opponent
with the 5 .Wb3 variation against the Griinfeld.
Kamsky came close to equalizing, but then went
wrong. It is worth looking at the part of the
game where Karpov increased his advantage.
8
I.
"'''=/..
6
5
L='.////,,/,
3
V='%
:
2
a
b
8
7
6
5
4
3
V="'" " , , ,
2
a
7
4
27.l"k2! E:xd5 ?!
The exchange of the d5-pawn for the one
on c7 favours White. Black should have
preferred one of the alternatives: 27 .. Jl,a7
2B.ttJb4 E:b6 29 . .td7; 27 . . . f5 2B . .te2 �a7
29.ttJc5; or 27 . . . ttJ eB 2B.ttJb4. In each case
White keeps a slight plus, but his advantage
is smaller than in the game.
2B.E:xc7 E:a5 29 . .tf3 .tg5 ?!
Kamsky decides to wait, but it soon transpires
that White can improve his position. I prefer
29 . . . .tfB!? or 29 . . . E:b3!? 30.E:d 1 .tfB.
30.E:d 1
Karpov and Henley mention that 30.�el
(intending .td 1) would be less effective in
view of 30 . . . E:b3!.
30 ... E:cB?!
Black's position was already unpleasant, but
now after the rook exchange Karpov will be
able to activate his king.
3 1 .E:xcBt ttJ xcB 32.h4 .tf6
32 . . . .th6 33.ttJb4 .tg7 34.E:c l ttJ d6 35 .Wfl
Black will have serious problems once the
king arrives in the centre.
c
d
e
f
g
h
b
c
d
e
f
g
h
33.E:c l !
Karpov characteristically improves all his
pieces.
33 . . . ttJ d6 34.Wf1 ! .te7 35 .�e2 WfB 36.�c7
.tf6 37.�d2 h5?
With this move Kamsky fixes his pawns on
the same-coloured squares as White's bishop,
making them vulnerable later in the game.
1996
The same comment does not really apply
to Karpov's pawns, as the black bishop is
restricted by the d4-pawn and his pieces are
generally rather passive.
38.<j;le2
Both players were in time trouble, so Karpov
just wants to pass the time control.
38...tLlf5?!
Kamsky makes a mistake in a bad position.
On 38 . . .id8 Karpov shows a nice win:
39.�c5! (Since both players were short
of time, it is possible that Karpov would
have simply maintained his advantage with
39.�d7.) 39 . . . �xc5 40.tLlxc5 tLl c4 4 1 .tLlxa4
if6 42.id5! (42.cj;Jd3 tLle5t 43.We4 d3)
42 ... d3t 43.Wxd3 lLlxb2t 44.lLlxb2 ixb2
45.a4 ie5 46.a5 ic7 47.a6 ib6 48.f4
White is winning, as his king threatens to
invade and if Black stops it then he loses his
kingside pawns. Now it is easy to see why
Kamsky's 37 . . . h5 was a mistake.
a
b
c
d
e
35 1
Anatoly Karpov - Gata Kamsky
f
g
h
39.�c4!
Even with little time on the clock, Karpov
immediately takes the opportunity to go
after the weak a4-pawn.
39 ... tLld6 40.�b4 �a6
40 . . . cj;Je7 4 1 .ic6 wins the a4-pawn.
41 .tLlc5!
Having remained in place for a long time,
the knight finally moves, and in doing so
seals the fate of the a4-pawn.
4 l . . .�a7 42. Wd3!?
The immediate 42.�xa4 was recommended
by Karpov as a simpler route to victory.
42 . . . �c7 43.lLlxa4 �cl
Kamsky has succeeded in activating his rook,
but this is not enough to compensate for
the missing pawn, and Karpov converts his
advantage smoothl�
a
b
c
d
e
f
g
h
44.lLlb6 ig7 45.a4 �a1 46.lLl d7t cj;Je8 47.lLlc5
cj;Je7 48.cj;Jc2 �fl 49.lLld3 �a1 50.Wb3 f5
5 1 .�b6 ih6 52.id5 g5 53.�a6 gxh4 54.gxh4
�dl 5 5 .ic4 �h l 56.a5 �xh4
1-0
Karpov - Kamsky, Elista ( 1 ) 1 996. The game
ends after Black's 56th move, which suggests
that Kamsky resigned after the adjournment
without resuming.
In Game 2 Kamsky chose the Panov variation
against the Caro-Kann. In the resulting IQP
position Karpov made a serious mistake and
allowed a thematic and powerful d4-d5 break.
Karpov sacrificed his queen for a rook and
knight, but Kamsky eventually broke through
and won. Interestingly it was the third time he
got into serious trouble against a d4-d5 pawn
sacrifice. Previously he had lost to Smyslov and
miraculously survived against Portisch from
what seemed like a hopeless situation.
In Game 3 Karpov got no advantage in a
4.g3 ia6 Queen's Indian, and although the
game lasted until the endgame it was always
roughly equal until the draw was agreed. Here
is the fourth game.
The Prime Years
3 52
I Game 43 I
12 ...�b6!
Gata Karnsky - Anatoly Karpov
Elista (4) 1996
l.e4 c6
Karpov repeats the Caro-Kann despite
suffering a loss in the previous game.
2.d4 d5 3.exd5 ad5 4.c4 tlJf6 5.tlJc3 e6
6.tlJf3 �b4 7.ad5 tlJxd5 8.i.d2 tlJc6 9.�d3
�e7 10.0-0 0-0 1 1 .�e2 tlJf6
Karpov repeats not only the variation but
the exact line as well. He must have analysed it
thoroughly before taking that decision. Against
Kasparov he usually abandoned a variation
after suffering a loss, even if theoretically his
position had been okay.
Black could have won a pawn with 1 1 . . .t2j db4
1 2.i.e4 ttJxd4 1 3.ttJxd4 �xd4 but White has
promising compensation. Kamsky would have
been ready for this, and it was obviously not to
Karpov's liking.
Two years later Van Wely introduced a novelty
in 1 1 . . . Wb6!? which was probably influenced
by Karpov's play in the present game.
12.tlJe4
This was Karpov's improvement, and a
novelty at the time. In the previous game he
preferred 1 2 . . . i.d7. That move had already
occurred in Mortensen - Danielsen, Ringsted
1 99 5 . In that game Black placed his queen on
b6 on the following move, so Karpov may have
got the idea from there. The queen move not
only attacks the b2-pawn but also vacates the
dB-square for a rook. Karpov's idea has stood
the test of time and has been used frequently
in the new millennium.
13.a3 i.d7
Sometimes 1 3 . . . 1"i:dB is played first, but
Karpov prefers to put the other rook on the
d-file.
14J�fdl
If White made the mistake of putting the
other rook on d 1 , Black would be able to take
the b2-pawn.
14.. J:�ad8
1 4 . . . 1"i:fdB is playable, but Karpov's choice
has been more popular. 1 5 .ttJxf6t ( 1 5.b4 is the
main move.) 1 5 . . . i.xf6 1 6.We4 g6 1 7.�f4 In
his annotations Karpov stops here, indicating
that White has pressure against the f7-square.
Nevertheless after 1 7 . . . i.g7 I B.i.e3 4Je7
1 9.ttJg5 f6 20.ttJe4 ttJ d5 Black achieved a lovely
position in Kunte - Sasikiran, Muzaffapur
1 99B.
15.tlJxf6t?
This move is a mistake, and has never been
repeated. Perhaps Kamsky did not expect
Karpov to repeat the same line and reacted
badly to the surprise. It is well known that
minor piece exchanges tend to favour the side
playing against the isolani. Kamsky probably
thought he would get attacking chances against
the black king, but if that was the case he was
being far too optimistic.
a
b
c
d
e
f
g
h
Gata Kamsky - Anatoly Karpov
1 996
In almost all subsequent games White has
preferred 1 5 .b4, safeguarding the b-pawn and
preparing to install the knight on c5. Let's have
a short look at how Karpov's novelty has fared
in subsequent years:
353
17 tlJe7!
•.•
Presumably Kamsky had been expecting
Karpov to move his queen, but this strong
move effectively refutes his entire concept. The
d5-advance is not a real threat, so Karpov takes
the opportunity to deploy both his knight and
light-squared bishop on more active squares.
18.tlJe5
After I B.d5? 'IMlxb2 Black is simply a pawn
up.
If I B.:B:d2!? Black has to play carefully as
I B . . . ic6? 1 9.d5! is strong. Instead Black can
keep the advantage with I B . . . ib5! 1 9.ib l
'IMlc7 intending . . . ic6.
a
b
c
d
e
f
g
h
) 15 . . . a6?! seems risky: 1 6.:B:ac l icB ( 1 6 . . . ltJ a7
1 7.ltJxf6t ixf6 I B.ig5 ixg5 1 9.ixh7t!
Wxh7 20.ltJxg5t 'kt>g6 2 1 .'IMlg4 f5 22.'IMlg3 f4?!
23.Wg4 White got a strong attack: Arizmendi
Martinez - K. Rasmussen, Linares 2002.)
1 7.ltJxf6t! ( 1 7.ic3 ltJd5 I B .ltJc5 if6 1 9.1tJe4
ie7 20.ltJ c5 if6 2 1 .ltJ e4 ie7 Yz-;6 Tsurtsumia
- Edzgveradze, Tbilisi 1 999) 1 7 . . . ixf6 I B.ie3
Black is under unpleasant pressure, Baron
Rodriguez - Svetushkin, La Roda 2005.
a
18 tiJf5!
.•.
They say one should blockade an isolated
pawn, but sometimes attacking it directly
is stronger. White is unable to advance the
d-pawn under favourable conditions, so the
direct approach works well here.
19.tlJc4
1 9 .d5? is tactically unsound, and after
1 9 . . . ltJxe3 20.ltJxd7 :B:xd7 2 1 .dxe6 ltJ xd l
22.exd7 ltJxf2 Black wins.
b) Black has a simpler and stronger way to
continue: 1 5 . . . ltJxe4! 1 6.'IMlxe4 f5 1 7.'IMle3 if6
Black has scored very well from this position.
15 ixf<i 16.¥ge4 g6 17.i.e3
...
a
19 ...¥9a6!
b
c
d
e
f
g
h
The queen steps out of the firing line of one
bishop, and into the sights of another.
a
b
c
d
e
f
g
h
The Prime Years
354
20.a4?!
This move weakens the queenside a bit.
20.ttJe5! would have minimized White's
disadvantage: 20 . . . �b5 2 1 .�c2 �c4 22.Wf4
(There is also 22.E!:ac l �d5 when White's
queenside is less vulnerable than in the game.)
22 ... �d5 23.�d2 �b3 (23 ... �g7 24.�b4)
24.�xf5 (24.�xb3? E!:xd4) 24 . . . exf5 25.E!:el
Compared with the game continuation,
White's queenside is more stable and his queen
will not be sidelined on the kingside.
20 ...i.c6 21 .Wff4 i.d5 22.llJe5 Wfb6
Having placed the bishop on its best square,
Karpov improves his queen and puts pressure
on the enemy queenside.
Karpov already a few moves earlier decided to
undertake the double pawn, but it is beneficial
for him as he gains control over the light
squares and his kingside gets strengthened.
24.E!:d2
8
7
6
5
4
3
2
1
a
24 i.g7!
b
c
d
e
f
g
h
•••
This is a much more subtle move than it
looks. Karpov removes the bishop from the
slightly exposed f6-square and prepares a
kingside advance with . . . h6 and . . . g5 at some
point in the future.
25.h4
a
23 ..ixf5
b
c
d
e
f
g
h
Kamsky gives up his light-squared bishop
in order to ease the pressure on the d4-pawn.
White cannot win a piece with 23.g4?? as
23 . . . g5! wins for Black.
Perhaps Kamsky anticipated that . . . h6 would
come.
25 .. J3fe8 26.Wfg3 �c8
Karpov mentions the possibility of
26 . . . h6 followed by . . . c;t>h7, but he finds it
more important to occupy the c-file.
27.llJd7
23.a5 is also unsatisfactory: 23 . . . Wxb2
(23 . . . Wb3!?) 24.E!:db l Wc3 25.E!:cl �xe5
26.E!:xc3 �xf4 27.�xf4 ttJxd4 White's bishop
pair does not compensate for his two-pawn
deficit.
23 ... exf5
This is not particularly attractive, but it is
hard to suggest a constructive alternative.
27...Wfc6
Karpov mentions 27 . . . Wd8!? which
indeed looks slightly stronger than the game
continuation as Black attacks both the d4- and
1 996
Gata Kamsky - Anatoly Karpov
h4-pawns. 2B.ltJc5 b 6 29.ltJd3 �e4 30.ltJ f4
ib7 Black will play . . . �c4 next, and White
will lose material without any compensation.
355
d4-pawn, and seeks to prevent the knight from
defending it via e2.
33J�ac1
28.ltJc5 b6 29.ltJ d3 VNd7
From this square the queen keeps an eye on
two weak pawns.
30.a5
The knight is unable to reach the c5-square:
33.ltJd3 i.xd4 34.ltJ f4 �dB 35 .Wf3 (3 5 .�d2
�xe3!) 35 . . . Wd6 36.g3 We5 Black is winning.
8
7
6
5
4
3
2
1
a
30 �e4!
b
c
d
e
f
g
h
•..
Karpov creates the powerful threat of . . . :ag4.
and virtually forces the following knight
move.
31.liJf4
Defending against the threat, but now all
three of White's kingside pieces are virtually
paralysed.
31...b5
The knight is only fWO moves away from the
c5-square, yet it is already too far.
a
33 h6!?
b
c
d
e
f
g
h
..•
This is partially a waiting move, although
it also makes White pay attention to the
possibility of . . . g5 . White's position is so
difficult that he has problems making a move
at all.
34J�c3 b4 35.:ac2 :ac6?!
Karpov gets ready to exchange a pair of
rooks, which improves his chances to obtain
something on the queenside. However, this
plan gives White time to improve his position.
Perhaps this slight inaccuracy crept in as a
result of time pressure.
32 ....ic4!
35 . . . i.b5!? was one improvement, for instance
36.�xcBt WxcB 37.�c l WdB 3B.�c5 (3B.b3
Wxa5) and here 3B . . . i.eB! is the most effective
way for Black to maintain his advantage.
Karpov shows flexibility in his thinking. His
bishop was superb on the long diagonal, but it
has done its duty there, and now he moves it
away. The reason is that he wants to take the
Even more convincing is: 35 . . . i.a6! 36.�xcBt
WxcB 37.�c l WdB! Black manages to attack
the d4- and a5-pawns at the same time.
32.�ddl
White had virtually no other piece to move.
The Prime Years
3 56
After 36.d5 l"&c8 37.l"&dc 1 b3 38.l"&d2 i.e5
White is dangerously tied up.
sacrifices. Perhaps Karpov's time shortage
even helped him by making him look for a
noncommittal move.
36 ...i.b5 37.@h2?
38.gxc6
Kamsky decides to wait and see how Karpov
intends to crack his position. He will find out
soon enough . . .
38 .l"&c5 does not help now because of
38 . . . i.xd4! .
36.EMcl
38 ...,ixc6 39.gc4
Much better was:
37.l"&c5!
This strong move does not equalize, but it
reduces White's disadvantage considerably.
The idea of this rook move is to prepare the
d5-push.
37 . . . l"&xc5
37 . . . i.xd4? 38.l"&xc6 i.xc6 39.lDxg6! White
gets counterplay.
Black cannot just improve his position
patiently with a move like 37 .. .'j;lh7 in view
of the strong 38.lDd3!.
38.l"&xc5 i.c6 39.d5 i.a8 40.lDh5 i.xb2
40 . . . i.h8 is unconvincing due to 4 1 .l"&c7.
4 1 .Wfb8t l"&e8 42.Wfxb4 i.e5 43.lDf4
White is only somewhat worse.
8
39 ... i.f8 40.tiJd3 VNe6
Karpov gets past the time control and
preserves his decisive advantage.
40 . . . Wfd5! was an even more effective route to
victory, as 4 1 .lDe5 l"&xe5! 42.l"&xc6 l"&xe3 is the
end for White.
However, 40 . . . l"&g4? would have been a serious
mistake due to 4 1 .lDe5!.
41 .d5
Kamsky postpones the inevitable by
exchanging rooks, but it fails to put up serious
resistance. If 4 1 .l"&c 1 i.d6 wins.
8
7
7
6
6
5
5
4
4
3
3
2
1
39.d5 i.a8! 40.l"&c5 i.e5 is also hopeless.
2
a
37 ... @h7!
b
c
d
e
f
g
h
Karpov finds a strong prophylactic move
which is typical of his style. The king move
anticipates White's potential counterplay
on the eighth rank and prevents any lDxg6
1
a
41. ..,ixd5
b
c
d
e
f
g
h
Black has more than one route to victory,
although he is not yet at the stage where he
can do whatever he wants.
Gata Kamsky - Anatoly Karpov
1 996
The attempt to be clever with 41 ... i.d6?? ends
in disaster after 42.dxe6 i.xg3t 43.fxg3 Elxc4
44.exf7 <;!;>g7 45 .i.xh6t! <;!;>xf7 46.tt:le5t when
White wins, as pointed out by Karpov.
4 l . . .'lWxd5 should be good enough: 42.Elxe4 fxe4
43.lLle5 i.e8 (It is not too late for Black to throw
the win away: 43 . . . i.d6?! 44.i.f4 i.xe5? 45 .i.xe5
�xa5 46.'lWf4! f5 The only move! 47.i.d4 Despite
the two-pawn deficit, White is in no danger
of losing. However, 43 . . . i.b5 is good enough
to win, for instance: 44.tt:lg4 ig7 45 .i.xa7 f5
46.lLle3 'lWe5-+) 44.tt:lg4 i.g7 45 .i.xh6 i.xb2
White will not last much longer.
42.Elxe4 i.xe4
43.i.xa7
This is losing, but it is probably no worse
than any other move, for instance 43.tt:lf4
�a6 and 43.tt:le5 a6! 44.i.d4 'lWd5! both win
comfortably for Black.
43.tt:lc5 is slightly more resilient, but still
leads to the same result: 43 . . . 'lWd5 44.lLlxe4 fxe4
45 .'lWh3 (45 .i.xa7? i.d6-+) 45 . . . i.g7 46.i.xa7
1xb2 47.a6 'lWd6t 48.g3 'lWxa6 49.'lWd7 'lWe6
Black wins easily.
8
7
6
5
4
3
2
b
43 ...i.d6!
Karpov drags the knight into a three-piece
pin, allowing no resistance. Kamsky was
probably hoping for 43 . . . 'lWa6? or 43 . . . 'lWd7?,
both of which allow White to continue the
fight with 44.i.c5! .
44.�f4
44.f4 'lWd7! 45 .i.d4 i.xd3 46.'lWxd3 i.xf4t
47.<;!;>h3 i.e5 wins, as Karpov pointed out.
44 .. JWe5 45.�h3
45.i.e3 'lWxb2 46.a6 b3 47.a7 'lWc3 wins, as
the b-pawn is unstoppable.
45 �e7
0-1
..•
42 . . . fxe4 was also good enough, but Karpov
prefers not to block the view of his light­
squared bishop. Black is winning comfortably;
aside from his extra pawn, the b2-pawn is
weak, and White's kingside is also vulnerable.
a
357
c
d
e
f
g
h
White is losing at least a piece, so Kamsky
resigned.
In the fifth game Karpov again played the
5 .'lWb3 variation against the Griinfeld. He came
up with a novelty but Kamsky responded well
with an exchange sacrifice. The game became
extremely sharp, but soon ended in a draw
in a repetition on move 23, as neither player
wanted to risk losing.
In Game 6 Karpov switched to the Petroff,
and by the end of the opening he had placed
his king on d8 in response to a rook check
on e l . His position looked dangerous, but he
kept it under control and gradually arranged a
kind of artificial castling by means of . . . Ele8e7 followed by a king transfer to f8. Kamsky
committed a few inaccuracies followed by a
losing blunder, and Karpov doubled his lead.
In the next game Kamsky tried the King's
Indian, but Karpov obtained the upper hand
after securing over control of the e4-square,
and he eventually converted his advantage to
take a 5-2 lead. An extract from the game can
be found in the notes to Game 42.
In Game 8 Karpov returned to the Caro­
Kann, and Kamsky switched to the main line.
358
The Prime Years
Simplifications resulted in a position in which
Karpov had an IQP. He had to be careful,
but was never in any real danger and held
the draw without any problems. In the next
game Kamsky returned to the Griinfeld and
the players repeated the same sharp variation.
Karpov deviated and Kamsky decided to
sacrifice his queen for a rook and a bishop. His
compensation proved to be insufficient, and
Karpov's accurate technique brought home the
win.
Trailing by four whole points, Kamsky
desperately needed a win, and he managed
to get it in Game 1 0. He switched to l .d4
and chose the Petrosian Variation against the
Queen's Indian. Both players held back castling
and Kamsky stirred up complications by
advancing his h-pawn. Karpov's king remained
in the centre. He made some mistakes and
Kamsky was able to obtain a strong attack,
which resulted in the win of a piece and victory
in the endgame.
In Game 1 1 Karpov played safely against the
Semi-Slav. He always had a symbolic edge, but
Kamsky was never in serious danger and the
players agreed a draw in the endgame. In the
next game Kamsky tried once again to crack
the Caro-Kann. After castling on opposite
sides he sacrificed a pawn to obtain play on
the g-file. He was able to regain his pawn, but
did not get any advantage and it was Karpov
who had a modest advantage before a draw
was agreed in the endgame. In the next game
Kamsky switched to the Queen's Indian, and
a complex endgame ensued. Karpov sacrificed
a pawn to create a passed pawn, but the idea
proved too optimistic. He then lost a second
pawn, but his strong bishop pair j ust enabled
him to hold a draw.
In Game 1 4 Kamsky decided to take on
the Nimzo-Indian, employing the Rubinstein
System with �d3 and lLl e2. In the ensuing IQP
position Karpov took control. He exchanged
queens, built his position in the endgame, and
won a pawn which he converted smoothly.
In the next game Kamsky tried the Modern
Benoni in attempt to win with the black
pieces, but Karpov played solidly and the
game remained roughly equal until a draw was
agreed in a queen endgame.
In Game 16 the players followed a long
theoretical line in the 4.g3 �a6 Queen's Indian.
On this occasion Karpov did not manage to
equalize, and a subsequent blunder of a pawn
led to a winning position for White, which
Kamsky converted smoothly. An extract from
the game can be found on page 400 of the first
volume. Karpov still kept a commanding lead
in the match, and in the next game he opted
for safety with the English Opening. Kamsky
tried hard and indeed managed to obtain some
advantage in the endgame, but eventually he
had to settle for a draw. In Game 1 8 Karpov
repeated the same line of the Queen's Indian.
Kamsky deviated and obtained a microscopic
advantage, but never got anything serious
and Karpov successfully held the draw in the
endgame.
With two games of the scheduled twenty
remaInIng, the match was halted as Karpov
had an unassailable lead of l OY2-7Yz. This
convincing victory over a tough challenger
enables Karpov to retain his title of FIDE
World Champion. It also elevated his rating to
one of its highest points of2775, although this
was partially due to rating inflation; Karpov
was still a great player, but he was not the same
force that he had been during the late seventies
and throughout the eighties.
***
Karpov's next event was the Biel tournament.
He started with a quick draw with Black
against Lautier, then outplayed and beat Zoltan
Almasi on the white side of a Queen's Indian.
1 996
Anatoly Karpov - Oleg Romanishin
In Round 3 Karpov took a very quick draw
with Andersson, followed by a slightly longer
draw after failing to get any advantage against
Glek's King's Indian.
In Round 5 Karpov got a nice position
on the black side of a Nimzo-Indian against
Onischuk, but then blundered badly. The
American grandmaster missed his opportunity
and blundered in return. Karpov was not so
generous and punished his opponent to score
his second win of the competition. In the
next round Karpov got no advantage against
Tukmakov's Semi-Slav, but kept playing and
almost got serious winning chances. Eventually
he had to settle for a draw after accurate defence
from his opponent.
In Round 7 Karpov took a day off by means
of a quick draw with Ehlvest. His next game
also ended quickly, but with a different result.
Facing Lajos Portisch, Karpov equalized against
Petrosian Variation of the Queen's Indian,
when the Hungarian grandmaster sent his
queen on a risky pawn-grabbing expedition.
He quickly got into trouble, and a subsequent
blunder meant he had to resign on move 20.
In Round 9 Karpov had the white pieces
against Milov, who played the King's Indian
with 7 . . . ltJ a6. Karpov obtained a small edge,
but the Russian-born Israeli-Swiss grandmaster
defended stubbornly and held on for a draw.
In the next game Karpov drew quickly against
Miles.
Karpov's opponent in the tenth and final
round was Oleg Romanishin. The grandmaster
from Lvov played all the world champions
from Smyslov to Anand, with the exceptions
of Fischer and Topalov. Out of sixty eight
games he scored twelve wins, thirty six draws
and twenty losses. Karpov was one of his most
troublesome opponents; previously he had
defeated Romanishin six times, with seven
draws and not a single defeat. This was the last
time they met over the board.
359
I Game 44 I
Anatoly Karpov - Oleg Romanishin
Biel 1996
l.d4 ttJf6 2 . c4 e6 3.tlJa b6 4.a3
Karpov rarely employed the Petros ian
Variation, but he obviously prepared it for this
game.
4 ..ib7
..
Romanishin scored well with 4 . . . �a6, but
on this occasion he decides to use his other pet
line.
5.ttJc3 g6
According to the database this move was first
used by Smyslov in 1 966. It was a favourite of
Romanishin, so Karpov m ust have been ready
for it.
6.VNd3
This move was introduced by Cvitan. White
prepares to seize the centre with e4, while
also preventing Black from inflicting a pawn
weakness by exchanging on f3.
6.. ..ig7?!
Black does better to prevent the central
occupation by means of 6 . . . d5, against which
White has so far been unable to demonstrate
any advantage.
7.e4
Karpov is glad to occupy the centre while
also restricting the b7-bishop.
7... d6
Earlier that same year Romanishin played
7 . . . d5 and drew against Cebalo. This time he
deviates, perhaps fearing Karpov's preparation.
8 ..ie2 0-0 9.0-0
360
The Prime Years
Long castling would be dangerous as Black
can still play . . . d5 .
9 ... �bd7
A year later Romanishin deviated with
9 . . . d5!? 1 O.cxd5 exd5 1 1 .e5 ttJ e4, and after the
further 1 2.�c2 c5 he had good counterplay
and eventually won the game Kalinitschew Romanishin, Muenster 1 997.
Karpov may have played differently, for
instance with 1 2.if4 or 1 2.�d l .
I B.ttJh4 ttJh5 1 9.�ab l �c7 20.ttJa4 �bB By
now Black had a reasonable position in all
- Romanishin, Yerevan (01) 1 996, but White
could have retained a small edge if he had kept
his dark-squared bishop on the board.
10.�c2
Karpov probably played this move in
anticipation of the moves . . . e5 and d5, after
which the knight j ump to c5 will no longer
attack the queen. The present position has
been reached numerous times via the 6.�c2
move order, with White having an extra
tempo. This does not mean that Karpov's move
order was incorrect, as it had the benefit of
discouraging the bishop from exchanging on
f3, and White's position remains slightly better
notwithstanding the loss of time.
With this great prophylactic move Karpov
prevents the black knight from occupying
the f4-square while also preparing to take
countermeasures against . . .f5 .
10 ... c5
12 ... ttJh5
12.g3!
a
b
c
d
e
f
g
h
A decade later Romanishin played 1 0 . . . �e7,
and after 1 1 .�e l e5 1 2.d5 a5 the players agreed
a draw in Arsovic - Romanishin, Belgrade
2006, although I prefer White's position and
it is safe to assume that Karpov would have
played on.
Romanishin rearranges his knights in order
to make room for his light-squared bishop
to come to the cB-h3 diagonal. If his bishop
remains on b7 then the attempt to obtain
counterplay with . . . f5 would weaken his
position too much.
1 1 .d5 e5?
12 . . . ttJ eB 1 3.�b l h6 ( l 3 . . . f5 ? 14.ttJg5) 14.b4
f5 ? 1 5 .ttJh4! �f6 1 6.exf5 gxf5 1 7.ib2 Black
cannot keep his kings ide together.
After this move Black's position remains solid
but both of his bishops are almost buried. From
a practical perspective, blocking the centre was
an especially bad decision against Karpov who
is a master of dosed positions.
Later the same year Romanishin improved his
play, although he still fell short of equality:
1 1 . . .�e7 1 2J:'lel a6 1 3.ifl ttJ eB 1 4.ig5
if6 1 5 .ixf6?! ttJ exf6 1 6.�d2 e5 1 7.g3 �acB
1 2 . . . h6 1 3.�b l ttJ h7 1 4 .ie3 f5 1 5 .exf5 gxf5
1 6.ttJh4! e4 1 7.ttJxf5! �xf5 I B.ttJxe4 White has
tremendous play for the sacrificed piece.
13.ttJel!
This move has numerous functions: White
hints at the possibility of taking the knight
1 99 6
361
Anatoly Karpov - Oleg Romanishin
on hS, while re-routing his knight to a better
position and also preparing f3 to defend the
king, or perhaps f4 to seize the initiative on
the kingside.
16.bxc5 bxc5
After 1 6 . . . dxcS 1 7.a4 Black is in serious
trouble on the queens ide.
17.2"i:bl �h3
13 ... .!Lldf6
Romanishin prefers not to allow the
weakening of his kingside pawns, but now his
f-pawn will be blocked.
8
7
6
5
4
3
2
1
a
18 . .!Llb5
a
14.tlJg2
b
c
d
e
f
g
h
Trying to gain space with 14.f4 was
tempting, but not altogether convincing:
14 . . . exf4! Black should sacrifice a piece
rather than allow White to increase his
space advantage. I S .g4 ( 1 S .gxf4 2"i:e8 1 6.tLlf3
'iWd7 Black exerts strong pressure on White's
centre.) I S . . . tLlxg4! Black gets two pawns and a
strong pawn chain for the piece. 1 6 ..ixg4 tLl f6
I 7 . .id l gS 1 8JWg2 h6 1 9.h4 tLl h7 Objectively
White may have a slight edge, but the position
is complex and Black will certainly have more
fun here than in the game.
14 ..ic8 15.b4
.
Karpov starts his play on the queens ide.
15 ...�d7
After I S . . . .ih3 1 6.2"i:b l Wd7 1 7.f3 2"i:ab8
I 8 . .id Black's kingside counterplay is lagging
well behind White's queenside initiative.
b
c
d
e
f
g
h
Karpov creates a direct threat on the
queens ide, although the text move also keeps
one eye on the opposite flank.
1 8.f3!?
Bolstering the kingside was also a reasonable
idea.
1 8 . . . tLle8
18 ... 2"i:d8?! is too passive: 1 9.tLldl Wd7 20.g4
tLl f4 2 1 .tLlxf4 exf4 22 . .ixf4 hS 23.h3 We7
24 . .igS Black has nothing for the pawn.
1 9.94!?
Winning a pawn, although Black gets some
counterplay. 1 9. tLl d l Wd7 is better for
White, but Black is still in the game.
1 9 . . . tLl f4
1 9 . . . tLl hf6?? 20.tLl d l traps the queen.
20.tLlxf4 exf4 2 1 ..ixf4 .ixg4 22.tLlbS
Also after 22.fxg4!? .ixc3 23.Wd3 .id4t
24.@h l White has some advantage.
22 . . . .id7 23.tLlxd6 .id4t 24.�h l tLl f6 2S.tLlb7
tLlhS 26 . .id6 2"i:ac8
Objectively White stands better, although
the position remains complicated.
The Prime Years
362
18 ...VNd7
21...tiJg4 22.hg4 hg4 23.8 ttJf6?!
1 8 . . . lLl e8 runs into a strong reply: 1 9.1"lb3!
ig4 ( I 9 . . .Wd7 20.ixh5 gxh5 2 1 .lLlh4 if6
22.lLlf5±) 20.f3 id7 2 1 .g4 lLl f4 22.lLlxf4
exf4 23.ixf4 Black loses a pawn for no
compensation.
This fancy move does more harm than good,
as it helps White to invade on the queenside.
19.i.d2!?
Karpov wants to avoid weakening his king,
so he resists the temptation to win material.
Instead he connects his rooks and threatens
ia5 and lLlc7.
The critical alternative was 1 9.94 lLlxg4!
( I 9 . . . a6 20.gxh5 axb5 2 1 .h6 ih8 22.cxb5
White remains a pawn up.) 20.ixg4 Wxg4
2 1 .lLlc7 id7 Though Black is objectively
worse, he has some compensation for the
exchange and remains in the game.
19 ... a6
Romanishin does not wait for ia5 but
instead sends back the knight. The problem is
that now the b6-square is weak.
20.tiJc3 VNh3
a
21.ttJa4!?
b
c
d
e
f
g
23 . . . id7
Retreating with the bishop would pose an
interesting question.
24.l"lb7
24.l"lb6?! f5! offers Black good counterplay:
25.lLlc3 (25 .l"lxd6? ixa4 26.Wxa4 fxe4)
25 . . . f4 26.g4 ixg4 The position is unclear.
24.lLl b6 l"lab8 25.lLlxd7 (25.a4 ie8 26.a5
f5) 25 . . . Wxd7 26.l"lb3 lLl f6 27.l"lfbl
l"lxb3 28.Wxb3 lLl e8 Black is only a bit
worse.
24 . . . l"lab8
24 . . . l"lad8 25 .ia5 (25.lLlb6 l"lb8! 26.lLlxd7
l"lxb7 27.lLlxf8 ixf8 28.l"lb l Wd7 White's
advantage is very small.) 25 . . . l"lb8 26.l"lfb l
l"lxb7 27.l"lxb7 ixa4 28.Wxa4 lLl f4 29.'lWc2
White has a clear advantage.
25.l"la7
If 25 .l"lfh l l"lxb7 26.l"lxb7 ic8 27.l"lb8
f5! 28.lLlb6 fxe4! Black gets back into the
game.
25 . . . l"la8 26.l"lxa8 l"lxa8 27.l"lb l
White maintains an edge, but Black has
good chances to defend.
h
Karpov proceeds with his plan and does
not mind giving up his light-squared bishop.
2 1 .l"lb6 also deserved consideration.
a
24.i.g5!
b
c
d
e
Karpov expels the bishop.
f
g
h
24 . id7 25J3b6
2S.tLlb6 Ei:abS 26.Ei:b3 also gives White a
nice advantage.
.
363
Anatoly Karpov - Oleg Romanishin
1 996
.
8
7
6
5
4
2S . . . �eS 29.tLle3 �h6 30.tLl d l +2S . . . hS 29.tLle3 �h6 30.tLl d l +29.tLle3 tLlhS 30.Ei:b2!
Now after a rook exchange, the queen will
be able to recapture without abandoning her
defensive duties along the second rank.
30 . . .f6
30 . . . tLl f4 3 1 .tLlxd7 does not help Black.
3 1 .tLl d l tLl f4 32.tLlxd7 Ei:xd7 33.Ei:bSt 'it>f7
34.Ei:aS
White wins the a-pawn.
3
2
a
b
c
d
f
e
h
g
25 .. l3ab8
25 ... Ei:fbS!?
Moving the other rook changes the position
somewhat.
26.Ei:fb l Ei:xb6 27.tLlxb6
27.Ei:xb6 tLl eS 2S.tLlh4 �fS Black is only
slightly worse.
27 ... Ei:a7
.
a
26.l3fbl
b
c
d
e
f
g
h
Of course Karpov is not tempted by 26.Ei:xd6?
�xa4 27.1lNxa4 Ei:b2 2S.Ei:f2 Ei:fbS when Black
obtains promising counterplay.
26 ... l3xb6 27.tiJxb6
This move preserves White's advantage, and
in some lines the powerful knight can hurt
Black.
a
b
c
d
e
f
g
h
28.�h4!
Surprisingly White can play to trap the
queen. This is the strongest continuation,
although White can also obtain a superior
endgame with 2S.tLlxd7 tLlxd7 29.tLlh4 f6
(29 . . . �fS 30.1lNg2) 30.�e3 �fS 3 1 .1lNg2.
28 ... Ei:b7
Nevertheless the alternative deserved serious
attention: 27.Ei:xb6!? 1lNhS 2S.�xf6 �xa4
29.1lNxa4 �xf6 30.1lNb3! Ei:dS 3 1 .Ei:xa6 1lNh6
32.tLle3 Black has no compensation for the
missing pawn.
27 ... l3b8
27 . . . h6 was a reasonable alternative: 2S.tLlxd7
(2S.�cl tLl h7 29.tLlel �eS 30.tLld3 1lNhS
The Prime Years
364
3 1 .We2 f5 Black has sufficient counterplay on
the kingside.) 28 . . . Wxd7 29 . .ie3 lD h7 30.�b6
�a8 3 1 .Wb3 White clearly has the upper hand,
but Black is not without chances.
8
7
6
29.lDe3
29.lDe l .ie8 The bishop move appears risky,
but Black can get away with it. 3o.ih4
(30.lDd3 Wh5; 30.Wc3 lDh5) 30 ... ttJh5!
3 1 .lDd3 lD f4 White can forget about
trapping the queen.
29 . . . .ie8
29 . . . lDh5 30.lDxd7 Wxd7 3 1 .lDg4 maintains
some pressure.
5
4
3
2
1
a
b
2B.gb3!
c
d
f
e
h
g
With this move Karpov sets up various
tactical motifs. From now on if Black takes
the rook it will not happen with a check.
Defending the f3-pawn is also useful.
2B ... tlJeB
Saving the bishop at once loses to a simple
tactical shot: 28 . . . .ie8 ?? 29 . .ixf6! .ixf6
3o.lDd7+28 . . . �b7!?
Black has time to play this rook move to save
the bishop.
a
b
c
d
e
f
g
h
c
d
e
f
g
h
30 . .ih4
White tries to trap the queen, but Black has
enough resources.
30 . . . lDh5! 3 1 .lDg4
3 1 . .id8 .id7 defends.
3 1 . . . lD f4!
Black is holding his own in the complica­
tions, for instance:
32.�b2 h6 33 . .id8
a
a
b
b
c
d
e
f
g
h
33 . . . f5! 34.exf5 e4! 35.f6 exf3 36.We4 'lMfxg4
Black is not worse.
Anatoly Karpov - Oleg Romanishin
1996
29.tlJel!
This is a tricky move to face, especially
with time trouble approaching. The knight is
heading for d3, from where it can influence
events in all areas of the board.
365
30 ... E1b7 3 1 . i2J d3 �f7 32.i2Jf2 WEh5 33.i2Jxd7
E1xd7 34.g4! WEh4 35.f4! exf4 36.�xf4 White
threatens to trap the queen, and Black has to
sacrifice a pawn to save her.
29 ... E:b7?
After this move Black's troubles soon worsen.
His position was already undesirable though.
29 . . . �b5?!
Romanishin is a gifted tactician. I doubt that
he missed this witty tactical shot, but rather
suspect that he rejected it after spotting the
refutation.
30.cxb5 E1xb6 3 1 .bxa6!
White wins thanks to a surprising feature of
the position, which was hard to anticipate
just a few moves earlier.
After 3 1 .a4 WEc8 32.WEc4 WEb8 White's
advantage is not so large.
3 1 .. .E1xa6 32.E1b8 WEd7
32 . . . f6 33.WEc4 E1a7 34.E1xe8t �f7 35 .E1d8
fxg5 36.WEb5 wins.
a
b
c
d
e
f
g
h
3 1 .cxb5
3 1 .a4!? is also promising: 3 1 . . .E1xb6 32.axb5
axb5 33.E1xb5 E1a6 34.E1b8 White continues
to press, but Black is still alive.
3 1 . . .E1xb6 32.a4!
32.bxa6 E1xa6 33.E1b8 WEd7 demonstrates
the value of Black's 29th move, as E1d8 is no
longer possible.
32 . . . WEc8 33.WEc4
White keeps a clear advantage, but Black has
better drawing chances than in the game.
30.tlJd3
a
b
c
d
e
f
g
h
33.E1d8! WEb5 34.WEc4!!
White wins thanks to the weak back rank.
29 .. .f6!
This ugly-looking move was the best chance.
Black makes room for his queen and sends
the troublesome bishop away.
30.�d2 �b5!
a
30 ...�h5
b
c
d
e
f
g
h
30 . . . f6? 3 1 .�h4 g5 32.E1b2 (32.i2Jf2 �a4
The Prime Years
366
33.tZlxa4 also wins} 32 . . . l"i:xb6 33.tZlf2 White
wins the queen.
3S.tZlc6 'Wg5 39.f4 exf4 40.tZlxf4 'Wf6 4 1 .e5!
dxe5 42.tZld3 .id 43.tZlcxe5
White is close to winning.
30 . . . .ih6?! 3 1 .tZl f2 'Wh5 32 . .ixh6 'Wxh6
33.tZlxd7 l"i:xd7 34.l"i:bS 'WfS 35 .'Wa4 White
wins a pawn and gets a decisive advantage.
3 1 .tLlxd7!
Karpov finds a strong, probably winning
method of simplification.
3 1 ..J:hd7?
Overlooking White's deadly reply.
3 1 . . .l"i:xb3!
Only this move would have enabled Black
to fight on.
32.'Wxb3 'Wxg5 33.'WbS 'We7 34.'WcS! a5
3 5 . <j;lg2 .ih6
8
7
6
b
c
d
e
f
g
h
Karpov once again proves that he is not only
a great strategic player, but that he also spots
tactical chances exceptionally well. Here he
traps Romanishin's queen.
32 .. JWxf3 33.llJfl
1-0
5
4
3
32.h4!
a
V-="-""",F- "'''-,-mm
2
a
b
c
d
e
f
g
h
36.a4
After 36.tZl 3xe5 dxe5 37.d6 'Wxd6 3S.'WxeSt
.ifS 39.h4 f6 40.tZlxfS 'WxfS 4 1 .'We6t the
queen ending is miserable for Black, yet he
still has chances to survive.
36 . . . <j;lg7
36 . . . .id2? 37.tZl 3xe5 dxe5 3S.d6+36 . . . .ig5 37.tZlbS 'WdS 3S.'WxdS .ixdS
39.tZlc6 .ib6 40.tZlcl White wins a pawn
and surely the game.
37.tZlbS mfS
37 . . . tZl f6 3S.tZlc6+-; 37 . . . 'Wg5 3S.'WxeS
'Wd2t 39.mh3 'Wxd3 40.tZld7+-
This win gave Karpov a total of7Yz/ 1 1 , which
was enough to share first place with Milov.
In 20 1 0 my good fortune enabled me to
conduct an interview with Romanishin. I did
not ask him about Karpov, but at one point
he mentioned that people nowadays do not
fully understand just how strong a player the
twelfth World Champion really was .
***
Karpov's next tournament was in Vienna. In
the first round he faced Ehlvest, who played
the QGD. Karpov pressed forward with a
minority attack but the Estonian grandmaster
defended well and the game was drawn. In the
next round Karpov once again had the white
pieces. Playing against Kramnik, he played a
novelty in the Semi-Slav and obtained a nice
1996
plus in a simplified middlegame. Kramnik was
unable to cope with his problems and Karpov
won a nice endgame.
In Round 3 Karpov suffered a setback
against Gelfand, who produced a strategic
masterpiece of which Karpov himself would
have been most proud. The game is referenced
briefly in the note to White's 1 1 th move in
Game 47 of the first volume (page 24 1 ) . In
Round 4 Karpov faced his old rival Korchnoi,
and chose the Petrosian Variation against the
Queen's Indian. Karpov got some advantage
by playing against Black's hanging pawns, and
won a nice game.
In Round 5 Judit Polgar played the Panov
against the Caro-Kann. The position quickly
simplified and a draw was agreed on move
1 9 . In the next game Karpov opened with
l .e4 against Yusupov, who played a Classical
Sicilian. Karpov agreed a draw two moves
earlier than in the previous game, although this
time the position was rich and full of pieces. In
Round 7 Karpov equalized on the black side of
an English Opening against Topalov, and the
players agreed a draw in an equal endgame.
In Round 8 Leko was able to equalize in the
Hungarian Variation of the Wb3 Griinfeld.
The game soon became a dead draw and the
players accepted the inevitable.
Karpov's ninth and final game was much
more exciting. Shirov chose the Short System
in the Advance Caro-Kann, and sacrificed two
pawns to get some initiative in a queen less
middlegame. For a while he stood better, but
he made some mistakes and allowed Karpov
to take over. Karpov won a nice endgame to
finish on a score of 5 Yz/9, which was enough
to share first place with Gelfand and Topalov.
***
Karpov's next event was Tilburg. In the first
round against Zoltan Almasi he played a
questionable variation in the main line of
367
the 4 . . . tLl d7 Caro-Kann. Almasi reacted
strongly and Karpov sacrificed an exchange
for insufficient compensation. He fought
hard but eventually had to resign. However,
he immediately recovered in the next round,
outplaying Lautier on the white side of a
Vienna which soon transposed to a Semi­
Slav. Karpov won a pawn and converted his
advantage smoothly.
In Round 3 Karpov equalized easily with
the 4 . . . . tLl d7 Caro-Kann against Sutovsky,
and the players agreed a quick draw. In the
next game he had the black pieces again,
and used the same opening to get a pleasant
position against Svidler. The players agreed a
draw in the middlegame, but perhaps Karpov
should have played on as his position was
more comfortable. In Round 5 Karpov got
some advantage against Shirov's Chebanenko
Slav. He collected a pawn but Black had some
compensation and Shirov was able to hold the
draw.
In Round 6 Karpov played the 4.g3 �a6
Queen's Indian against Van Wely. The
Dutchman had the two bishops and obtained
a space advantage with a well-timed d5. He
increased his advantage and won a fine game.
In the next round Karpov used the Fianchetto
System against Gelfand's King's Indian. The
position remained balanced throughout the
game until a draw was agreed in the ending. In
Round 8 Judit Polgar used the Panov Variation
and sacrificed a pawn for the two bishops and
some activity. Then on move 1 9 the players
agreed a draw in a roughly level position.
In Round 9 Karpov tried to repeat the
Fianchetto King's Indian, but Piket opted for
a Griinfeld with . . . d5, without a preliminary
. . . c6. Karpov was unable to obtain any
advantage, and following some simplifications
the players agreed a draw. In the next game
Adams tried an early h4 in the Advance Caro­
Kann. He sacrificed his c4-pawn and got
decent compensation in the complications.
The Prime Years
368
Following some simplifications White had a
small advantage in the endgame, but Karpov
was able to hold it. In the eleventh and final
round Karpov faced Leko. The Hungarian
grandmaster played a novelty in the 'lWb3
Griinfeld, but Karpov reacted strongly. He
castled on the queenside, launched a powerful
attack, and sealed the victory with some precise
moves.
Despite the excellent last round Karpov only
finished on fifty percent, with 5 Yz/ 1 1 , which
was enough to share sixth place with Adams. It
was a clear indication that his days as number
two in the world ranking list were finally at
an end.
***
Before Karpov's next classical event, he took part
in the European Rapidplay Championship. It
is not possible to work out from the database
the format in which the tournament was
played, but Loek van Wely was kind enough
to inform me that the tournament began with
preliminary qualifying groups, after which the
top eight players advanced to the knockout
stage.
One of Karpov's best wins from the
event came against Ilya Smirin. The Israeli
grandmaster played fifteen classical games
against the world champions, scoring three
wins, six draws and six losses.
I Game 4� I
Anatoly Karpov - lIya Smirin
6 ... liJc6 7.liJc3 a6
According to the database Smirin has a
varied repertoire and plays several variations
against the Fianchetto System.
8.l':'!el
This move is not one of the most heavily­
analysed main lines, although it was still
quite well-known. A few weeks earlier Karpov
played 8.b3 against Gelfand but obtained no
advantage at all. He would go on to score three
wins and one draw with the text move.
8 ... l':'!b8 9.l':'!bl ttJa5
Karpov's next game in this variation took
place the following year and continued as
follows: 9 . . . b5 1 O.cxb5 axb5 1 l .b4 e6 1 2.e4
0,e7 1 3.0,d2 c5 1 4.bxc5 dxc5 1 5 .dxc5 0,d7
1 6.0,xb5 0,xc5 1 7.0,c4
a
b
c
d
e
f
g
h
1 7 . . . 0, d3? (Better was: 1 7 . . . ia6! 1 8.0,cd6
[ 1 8.a4?! 'lWxd 1 1 9.1"i:xd 1 0,xa4 Black is fine.]
1 8 . . . ixb5 1 9.0,xb5 0, d3 20.1"i:e2 0,xcl
2 l .'lWxc 1 'lWa5 22.'lWc5 This position occurred
in Borges Mateos - Leyva, Cuba 2000,
and here Black should have tried 22 . . 1"i:fc8!
23 .'lWxe7 1"i:xb5 with decent drawing chances,
as the pawn deficit should not have a great
role.) 1 8.1"i:e2 ia6 1 9.a4 0,xc 1 20.'lWxc1 1"i:c8
2 l .1"i:c2 ixb5 22.axb5 'lWd4 23.b6 White was
already winning and Black soon had to resign,
Karpov - Shirov, Dos Hermanas 1 997.
.
European Rapid Championship, Cap d'Agde 1996
l.ttJa ttJf6 2.c4 g6 3.g3 .ig7 4 ..ig2 0-0
5.0-0 d6 6.d4
Just like Boris Gelfand, Ilya Smirin is
originally from Belarus and is a pupil of Albert
Kapengut. He is also an expert on the King's
Indian Defence.
10.'lWa4 c5?
Anatoly Karpov - IIya Smirin
1 996
It is early for Black to embark on a direct
confrontation.
A year after his crushing defeat, Shirov came
up with a different plan: 1 O . . . b6 I l .c5 i.d7
12.Wa3 tt:l c4!? 1 3.Wxa6 b5 This time Black
had reasonable compensation and went on to
draw in Karpov - Shirov, Wijk aan Zee 1 99B.
369
is no reason to risk the speculative gambit
1 5 .1"1bd l ?! i.xe5 1 6.i.xe5 tt:lxc4, even if White
does have some compensation.) 1 5 . . . Wxd7
1 6.1"1bd l We6 1 7.i.d5 Wh3 I B.i.e3 tt:l f6
1 9.i.g2 White keeps the bishop pair and a
long-lasting initiative on the light squares, but
Black has reasonable chances to defend.
15.13bdl
Just as in many positions with a 'Catalan'
bishop on g2, the black queen has problems
finding a good square.
8
7
6
5
4
3
2
a
1 l.dxc5!
b
c
d
e
f
g
h
By opening the position Karpov gives
himself the opportunity to win valuable
time and obtain a considerable advantage in
development.
15 ...Wfe8
1 1. ..i.d7 12.Wfc2 dxc5 13.i.f4! 13c8 14.lLle5
ie6?
The queen would have failed to find a safe
refuge on the queenside: 1 5 . . . Wb6 1 6.tt:la4!
Wa7
a
b
c
d
f
e
h
g
Black fights for equality but only encounters
bigger problems. Instead he should have settled
for a somewhat worse position with chances to
defend.
There was one subsequent deviation from this
game, but it failed to get close to equalizing:
14 . . . i.f5 1 5 .e4 i.e6 1 6.13bd l tt:l d7 1 7.tt:lxd7
ixd7 I B .e5 WeB 1 9.tt:ld5 White has a large
advantage, Stohl - Vrana, Martin 2003.
14 ... tt:lh5! With this move Black acknowledges
that he has lost the opening battle and fights
to minimize his disadvantage. 1 5 .tt:lxd7 (There
a
b
c
d
e
f
g
h
1 7.i.d2! White can profit from the misplaced
knight. 1 7 . . . tt:lxc4!? ( l 7 . . . tt:lc6 I B.tt:lxc6 bxc6
The Prime Years
370
1 9.�e3 tLl d7 20.tLlxc5! White wins a pawn.)
1 8.tLlxc4 �xc4 ( l 8 ... b5 1 9.tLlcb6 does not
change much) 1 9.iWxc4 b5 20.iWh4 bxa4
2 1 .iWxa4 Black maintains the material balance
(for the time being) , but his weak queens ide
pawns will give him serious problems.
White could also consider making a slight
improvement with 1 9.E1d2!? when he keeps
some advantage.
a
a
16.lLld5!
b
c
d
e
f
g
h
Karpov tempts his opponent into exchanging
the knight, which will enable him to increase
his space advantage.
16 ...hd5?!
Changing the pawn structure rather helps
White.
1 6 . . . tLlh5!?
This looks like Black's best chance. It turns
out that he can tolerate the strong knight,
and can ease the pressure in the centre
without altering the pawn structure.
1 7.�d2 �xe5 1 8 .�xa5 tLl f6
White keeps some advantage, but it is not
easy to do something with it.
1 9.tLlb6!?
This seems to be the most ambitious
approach.
1 9.�c3 �xc3 20.iWxc3 tLlxd5 2 1 .cxd5 �d7
22.d6 (22.e4 e5) 22 . . . �c6 23.d7 �xd7
24.�xb7 E1b8 25.�xa6 �e6 Black has some
compensation for the pawn.
b
c
d
e
f
g
h
1 9 . . . E1b8
19 . . . E1d8 20.iWb3 E1xd l 2 1 .E1xd l 'Wb8
22.iWa3! �d6 23.f4! E1e8 24.e4 Black faces
problems in the centre.
20.�c3
Another line is: 20.f4 �c7 2 1 .iWc3 E1d8
22.tLld5 �xa5 23.iWxa5 �xd5 24.cxd5 'Wb5
Black is worse but still alive.
20 . . . �xc3
20 . . . �c7 2 1 .tLld5 is strong.
2 1 .iWxc3 tLld7 22.tLlxd7 �xd7 23.iWe5 b6
Black is living dangerously, but it is not clear
if White can invade.
17.cxd5 liJh5?!
It is too early for Black to simplify the
position. White's pieces are already more
actively placed, and Karpov soon exploits this
latest inaccuracy to achieve an even greater
level of domination.
17 . . . iWb5! Black's top pnonty should have
been to find a way to bring his queen and rook
into play. 1 8.�f3 (The evaluation is similar
after 1 8.�g5 E1fd8 1 9.e4 c4) 1 8 . . .E1fd8 1 9.e4
c4 White is clearly better thanks to his strong
centre and bishop pair, but Black is still in the
game.
1 996
Anatoly Karpov - IIya Smirin
37 1
22 ... �h8
8
When calculating this position in advance,
Smirin may have missed that he would not be
able to expel the rook with 22 . . . if8 because
of 23.ttJd7!.
7
6
5
23.'lWe4 c!lJ c6
4
A desperate attempt, but Black has no
reasonable option as after 23 . . .if8 24.l"ld7 b6
25.l"la7 White has a winning advantage.
3
2
a
18.ih3!
Karpov
position.
b
c
forces
d
e
f
a weakness
24.tlJxc6 bxc6 25.ixe6
h
g
in
Karpov's accurate play pays dividends and he
collects a key pawn in the centre.
Black's
18 ... c!lJxf4 19.9xf4 e6
A sad necessity, but other moves would have
lost material, for instance: 1 9 . . . 1"k7? 20.b4+-
25 ... l"lf8 26.b3 gbS 27.e3 g5 2S.f5!?
Naturally Karpov could have taken the
second pawn, but he prefers to avoid giving his
opponent even the slightest random chance.
2S ... gb4 29.'lWf3 ie5
20.dxe6 fxe6 2U�d6
Karpov wastes no time in attacking the
weakness.
21 ..JU6 22J�edl
White is already winning. Smirin was clearly
having an off day, as a grandmaster at his level
rarely has problems of this scale so early. Of
course we should not forget that it was a rapid
game.
8
7
6
5
4
3
2
a
30.gdS
b
c
d
e
f
g
h
Simplifying to an easily winning endgame.
30 ... 'lWxdS 31.gxdS gxdS 32.f6! gf8 33.f7
White has a solidly-defended pawn one
square away from promotion, while Black has
numerous weak pawns. The remaining moves
can pass without comment.
a
b
c
d
e
f
g
h
The Prime Years
372
33 .. J�h4 34.h3 �g7 35.�xc6 h5 36.�xc5
.if6 37.c.f;>g2 g4 38.hxg4 hxg4 39.b4 �fh8
40.�c7 c.f;>f8 41.a4 g3 42.�d6t
1-0
According to the database Karpov achieved
plus scores against Glek, Hracek, Magem
Badals, Smagin and Smirin. He faced Van
Wely in the final and beat him in a play-off.
***
Karpov's final tournament of the year was Las
Palmas, a double-round-robin event involving
all the top players of that time, with the
exception of Kamsky. In the first round Karpov
faced Ivanchuk and played an unusual line in
the Fianchetto King's Indian. Ivanchuk not
only equalized but obtained some advantage
and Karpov had to fight hard to draw. In the
next game he had the white pieces again, but
obtained no more than a symbolic edge against
Anand in a 4.g3 i,a6 Queen's Indian, and had
to settle for a draw.
In Round 3 Karpov surprised Topalov with
the Queen's Gambit Accepted and drew fairly
easily. In the next game he equalized with
the l .c4 e5 English against Kramnik, and
another draw ensued. In Round 5 he faced
Kasparov, who used the Vienna System against
the Queen's Gambit. Karpov reacted timidly
and soon reached a slightly uncomfortable
queen less middlegame. Later Kasparov missed
a win in the pawn endgame and Karpov
managed to survive. In the next game Karpov
easily equalized and scored his sixth successive
draw, after surprising Ivanchuk with the
Rubinstein Variation of the French Defence.
In Round 7 Karpov's drawing streak came to
an end. With the black pieces against Anand, he
repeated the QGA but soon found himself in
a difficult position. At one point Anand could
have captured a pawn for no compensation,
but instead he sacrificed a bishop and won
in fine style. In the next round he tried to get
back on track against Topalov, who equalized
in the Modern Benoni but later made a few
inaccuracies. Karpov got some advantage but
the Bulgarian grandmaster defended well and
Karpov had to settle for a draw.
In Round 9 Kasparov played the 4.�c2
Nimzo-Indian. Karpov got a pleasant position
but later committed a serious inaccuracy and
Kasparov found a great tactical idea. Karpov
survived but had to enter a hopeless endgame
which the world number one converted
easily. In the final round Kramnik played an
unreasonably risky novelty in the Semi-Slav.
Karpov in his heyday would probably have
punished him for it, but he allowed Black to
obtain counterplay. Later he missed a win in
a sharp position and Kramnik escaped with a
perpetual check.
Karpov finished on a disappointing 4/ 1 0,
sharing last place with Ivanchuk without
winning a game.
***
In the early part of 1 996 Karpov played
excellently, but towards the end of the year
his play deteriorated. Perhaps he played in
too many tournaments again, especially with
a World Championship match taking place in
the middle of the year.
From 1 973 until 1 99 5 Karpov had been rated
either the number one or number two player
in the world. The Las Palmas tournament
marked the passing of the torch, by which
time the new breed of super-grandmasters
had begun to surpass him. There are precious
few players in the whole of chess history who
stayed within the top two places in the world
rankings for as long as Karpov. And even
though his level was gradually dropping, the
twelfth World Champion of course remained
a world-class player who would go on to
produce many more wonderful games in the
coming years . . .
373
1996 Summary
Spanish Team Championship: 3/4 (+2 =2 -0)
Yugoslav Team Championship 2/3 (+ 1 =2 -0)
Belgrade (3rd-4th place) : 2Y2/5 (+ 1 =3 - 1 )
FIDE World Championship match versus Kamsky, Elista: Won 1 0Y2-7Yz (+6 =9 -3)
Biel ( I st-2nd place) : 7Y2/ 1 1 (+4 =7 -0)
Vienna ( I st-3rd place) : 5Y2/9 (+3 =5 - 1 )
Tilburg (6th-7th place) : 5 Yz/ 1 1 (+2 = 7 -2)
Las Paimas (5th-6th place) : 4/ 1 0 (+0 =3 -2)
Totai 57.6% (+ 1 9 =38 -9)
EJ Wins • Draws • Losses
1997
Rating 2760 (3 in the world)
Karpov began the new year with various rapid events and simultaneous exhibitions. His first
classical tournament was Dos Hermanas, where he started with the black pieces against Illescas.
He equalized with the 4 . . . ctJ d7 Caro-Kann and the players agreed a draw in a level endgame. In
the second round Karpov played the Catalan against Judit Polgar, who equalized. Karpov got
some advantage in the middlegame, then he erred and allowed Judit to take over the initiative,
but it was not too serious and the game ended in a draw.
In Round 3 Karpov had the black pieces against Kramnik, and obtained a reasonable position
in a line of the 4.Wc2 Nimzo-Indian, which occurred via an English move order. But later he
fell into trouble and Kramnik was able to make his bishop pair count. In the next round Karpov
bounced back with a convincing victory over Shirov, which is referenced in the notes to Game 45
at the end of the previous chapter.
In Round 5 Karpov successfully defended the 4 . . . ctJd7 Caro-Kann and drew without too many
problems against Topalov. In the next game Anand played the Queen's Gambit Accepted. Karpov
played a novelty and a complex middlegame occurred. Karpov could have won had he found
the right path through the complications, but in the end he had to settle for a draw. In Round 7
Karpov entered a sharp queen less middlegame on the black side of an English Opening against
Salov. The tense position soon petered out to a level endgame and the players agreed a draw.
In the next round Karpov produced a great game against Nigel Short. Since their Linares 1 995
encounter (Game 40 in the present book) the players met just once and drew.
I Game 46 I
Nigel Short - Anatoly Karpov
Dos Hermanas 1997
l.e4 c6 2.d4 d5 3.e5 .if5 4.llJf3
4.ctJc3 e6 5.g4 �g6 6.ctJge2
Earlier in the year Karpov won a rapid match against Kotronias by the score of 2'12-1 '12, in
which this ultra-sharp line occurred twice.
6 . . . ctJ e7 7.ctJf4 c5 8.h4 cxd4!
This is Karpov's novelty, which he first introduced against Ljubojevic in the rapid section of the
1 994 Amber tournament.
9.ctJb5 ctJ ec6 1 O.h5 �e4 1 l .f3
The Prime Years
376
a
b
c
d
e
f
g
h
1 1 . . . �xf3!?
Karpov sacrifices a piece for three connected
passed pawns.
Black scores even better with 1 1 . . . a6!, which
occurred in several games in 1 999. The
database does not give the precise dates of all
the games, bur based on the ordering of the
games, it looks like this important move was
introduced by a player named Papa with a
rating of just I B70!
1 2.Wxf3 lLlxe5 1 3 .Wg3
Ljubojevic preferred 1 3 .'We2 and eventually
f6 I B .�d2 Wb6 1 9.a4 a6 20.a5 WdB 2 1 .g5?
axb5 22.�xb5 �f7 23.g6t hxg6 24.hxg6t
�e6 Black had a clear advantage and went
on to win, Kotronias - Karpov, Athens
(rapid - 3) 1 997.
A few years later Karpov deviated with 14 .. .f6
bur was unsuccessful. Interestingly Kotronias
also encountered this move in a couple of
subsequent games. 1 5 .lLlf4 �d7 ( l 5 ... �f7
led to an eventual draw in Kotronias Bouaziz, Las Vegas 1 999.) 1 6.g5 (White
was successful with 1 6.c3 in Kotronias Adianto, Buenos Aires 1 997.) 1 6 . . .f5 1 7.c3
dxc3 I B.bxc3 Wa5 1 9.�d2 Wa4 20.ie2
�c5 ? (20 . . . a6 or 20 . . . lLl c4 should have been
tried.) 2 1 .Ei:h4! lLlg4 22.lLld3 �e7 23.iMfc7t
�eB 24.Wxb7 1-0 Anand - Karpov, Monte
Carlo (blindfold) 200 1 .
1 5 .�xd3 e5 1 6.0-0 �c5
won, but Black's p lay could have been
im p roved and at this stage he is doing
well.
1 3 . . . lLl bc6 14.lLld3
8
��
�.�
��
' ' ' /'� 'i�.i
6 �r�
" %�1�'�%
'%
m
�
t!i� l.�� �
�
� ��" ' !�
�� 1t5� ��
!�!��.''
' { ' %�7� m;Z 0( ��' %�
7
5
4
3
2
�
a
b
�
c
d
�j,� :
e
f
g
h
1 4 . . . lLlxd3t
Karpov played this move in the first
Kotronias game, and returned to it later in
the same year against Laurier.
1 4 . . . Wa5t 1 5 .�dl lLlxd3 1 6.�xd3 e5 1 7.Ei:e l
a
b
c
d
e
f
g
h
1 7.Ei:f5
1 7.h6 g6 I B.Wf2 'Wd7 1 9.�h l a6 20.lLlc3
Ei:fB 2 1 .lLla4 �a7 22.b3 Wxg4 23.�a3 iMfh5t
24.�gl Wg5t 25 .Wg2 Wxg2t 26.�xg2 e4
27.�xfB �xfB 2B .�e2 f5 Black obtained four
pawns for the rook and went on to convert
his advantage to a win, Lautier - Karpov,
Monte Carlo (rapid) 1 997.
17 ... 0-0 I B .h6 g6 1 9.Ei:xe5 a6 20.lLla3 Wic7
2 1 .�f4 lLlxe5 22.�xe5 Wb6 23.Ei:fl Ei:ae8
24.g5 Ei:e6
Black has a playable position and bur lost
after subsequent errors, Kotronias - Karpov,
Athens (rapid - 1 ) 1 997.
1 997
4..
.
Nigel Short - Anatoly Karpov
e6 5.a3
Karpov probably expected this move as Short
had used it to beat Leko in 1 996.
5
.
. �e7
.
Karpov allows his opponent to exchange his
light-squared bishop; he never did it again.
In his 2005 rapid match with Hamdouchi,
Karpov lost one game with 5 . . . h6 and scored a
win and a loss with 5 . . . ltJ d7.
6.�bd2 tlJd7 7.tLlh4!
Eliminating the bishop is critical. Black
must search for fluent piece play otherwise he
will simply be worse.
7. c5 S.d a6
Karpov makes a useful move while waiting
for his opponent to exchange on f5 , which will
help his development. Perhaps S . . .l'kS would
have been a slightly better way of achieving the
same goal.
..
377
Karpov wants to clarify the situation in the
centre. He is willing to undertake doubled
pawns and an isolated one, hoping that the
availability of the e6-square and the slight
weakness of the e5-pawn will make up for his
structural defect.
1 1 . . .�e7 and 1 1 . . .g6!? are both reasonable
alternatives.
12 ..ixfS exfS 13.tLlxd4
Short wins a tempo as Black has to defend the
f5-pawn; on the other hand he subsequently
moves the knight back in order to clear the
d-file.
When the same position occurred again White
took back with the queen:
1 3.Wxd4 ltJc5
9 .�xfS
Sooner or later White has to take. Short
wants his bishop on d3 so he does not delay
it.
9.
..
�xfS 10.tLla :B:cS 1 l ..id3
1 4.�g5
1 4.0-0 Wd7 1 5 .:B:b 1 �e7 ( 1 5 . . . ltJ e4!?)
1 6.:B:d 1 :B:dS The position is equal.
14.�e3 ltJ e6 ( 1 4 . . . �e7 1 5 .:B:d l ) 1 5 .Wa4t
( 1 5 .Wd3 f4 1 6.�d4 �e7 Black is not worse.)
15 ... Wd7 1 6.Wxd7t It>xd7 1 7.:B:d1 f4
1 S.�c l It>c6 White's advantage is tiny, if it
exists at all.
The text move was played in Kuta - Trs,
e-mail 2004, and here the soundest response
would have been:
14 . . . �e7 1 5 .�xe7 Wxe7 1 6.:B:d1 0-0 1 7.0-0
1 7.Wxd5? :B:fdS leads to trouble.
1 7 . . . :B:fdS
The Prime Years
378
Black should have no problem coping with
his pawn weaknesses.
13 ... g6 14.0-0
1 4.e6 eliminates the e-pawn as a future
target, but leads to no advantage as Black is
well-developed: 1 4 .. .t2l c5 1 5 .0-0 �g7 1 6.1'l:e l
( 1 6.exf7t r;t>xf7 Black's active pieces give him
comfortable play.) 16 . . . iWd6 1 7.e7 tLJ e6 Black
is in the game.
A more dynamic and challenging approach
would have been: 20.a4! Opening the a-file
and creating a weak pawn in Black's position.
20 .. .f4 (20 . . . iWd7 2 1 .axb5 axbS 22.Ela7
White is active enough.) 2 1 .�xf4!? (2 1 .�d4 is
simpler, and keeps White's position together
in the centre.) 2 1 . . .1'l:c4 22.�e3 iWaS 23.id4
1'l:xa4 24.1'l:xa4 bxa4 25 .iWa2 tLJ d6 The position
is roughly equal.
20 l3c4!
•. •
Karpov defends the d5-pawn by blocking
the queen, and prevents a4 in the process.
14 ... ttJc5 15 ..ie3
I S .tLJb3 can be met by I S . . . tLJ e6.
15 ....ig7
21 ..id4
The fianchetto-bishop forces White to pay
attention to his vulnerable e-pawn. Defending
it with f4 would be undesirable due to the
weakening of the e4-square and the worsening
of White's bishop.
It was possible to sacrifice the eS-pawn
as well: 2 1 .iWc2!? �xe5 22.tLJxeS Elxe5
23.�d4 1'l:e6 24.f3 White's strong bishop and
better pawn structure offer him sufficient
compensation.
21...iWb8!
8
7
Karpov attacks the e5-pawn with another
piece, while also preparing a possible minority
attack with . . . a5 and . . . b4.
5
22.l3d3
6
Another sensible continuation was 22.g3
f4!? 23.1'l:ee l iWc8 24.iWc2 iWfS 25 .iWd3 ih6
26.�b6 gS with a double-edged position.
4
3
2
1
a
16. ttJf.3
b
c
d
e
f
g
h
Short defends the pawn and clears the
d-file.
16 ... 0-0 17J3el l3e8 18.l3e2 ttJe4
Karpov installs his knight on a fine square
while somewhat isolating the eS-pawn from its
camp.
19.iWb3 b5 20.l3dl
Nigel Short - Anatoly Karpov
1 997
22 ... f4!
Karpov takes away the e3-square from the
bishop. Later he can advance his other kingside
pawns, perhaps to attack, or simply to gain
space and choke his opponent.
23.�dl gc6!
The rook has done its duty on c4, and in
the present position it will stand better on the
e-file.
379
after the d5-pawn. But it soon transpires that
the knight stands worse on c2 than on f3.
It was not too late to correct the error with
25.ctJf3!? g5 26.h3. Black can try to push his
kingside pawns or transfer his queen to c4,
but White should have enough resources to
hold the position as the d5-pawn will remain
weak.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
a
2S ...�b7!
b
c
d
e
f
g
h
Even though Black seems to have done more
to improve his position over the last ten moves,
White was still not objectively worse. But
starting with this rather passive moves, things
start to go awry for him.
This is a typical Karpov move: not the most
dazzling, but extremely strong nonetheless.
The point is to anticipate White's counterplay
by defending the d5-pawn in advance. Since
the white knight is no longer defending the e5pawn, the black queen can afford to find a new
occupation.
The black rook may have left the fourth rank,
but 24.a4?! was still not a good idea due to the
simple 24 . . . b4.
26.f3?!
24.lLlel?!
White would have done better to challenge the
knight with 24.ctJd2! after which 24 . . . ctJ g5 can
be met by the calm 25J�:e l . In that case the
position would have remained equal in view of
the mutual pawn weaknesses.
24... gce6 2S.llJc2?
Short wants to kick away the strong knight
with f3, and hopes to use his own knight to go
Short probably failed to appreciate how
strong Karpov's knight would become.
A better chance to survive would have been:
26.EI:e 1 !?
Black's pawn structure is imperfect to say the
least, so giving up the e5-pawn would have
given White some drawing chances.
26 . . . .ixe5 27 . .ixe5 EI:xe5
Here White can choose between solid
defence and attempting a counterattack.
The Prime Years
380
8
7
�{�'f'��
6
5
4
3
2
a
b
c
d
e
f
g
h
28J::!: d4
28.ttJb4 �a7! With this precise move Black
ensures that a capture on d5 will not come
with gain of tempo. (2B . . . �b6 29.l"i:d4
ttJxf2?? 30.l"i:xe5+-; 2B . . . a5 29.ttJxd5 'kt>g7
30.l"i:fl f3 3 1 .g3 White is living dangerously
but he seems to be okay.) 29.l"i:d4 ttJxf2!
30.l"i:xe5 ttJxd l 3 1 .l"i:xeBt 'kt>g7 32.l"i:e2
(Unlike the analogous position with the
queen on b6, 32.ttJxd5 does not work here
and after 32 . . . ttJxb2 White is in trouble.)
32 ... ttJ e3 White is struggling.
2B . . . a5! ?
2B . . . f6 gives back the pawn i n order t o draw
the enemy rook to an unfavourable square.
29.l"i:fl a5 30.f3 ttJ d6 3 1 .l"i:xf4 'it>g7 32.�d3
ttJ f5 The rook is misplaced on f4, but White
should be able to live with it.
29.f3 ttJ c5 30.l"i:xe5 l"i:xe5 3 1 .l"i:xf4 ttJ e6 32.l"i:g4
h5 33.l"i:h4 �b6t 34.'it>fl
34.'kt>h l ? �f2 wins.
34 . . . b4! 35.axb4 g5! 36.l"i:xh5 �a6t 37.'kt>gl
ttJ f4
White loses an exchange.
26 ... lL'ld6 27.b3
Another possibility was: 27.ttJb4 ttJ c4
2B.b3 a5 (2B . . . .txe5 29 . .tf2 ttJ b6 30.l"i:c2 �e7
3 1 .l"i:cl [but not 3 1 ..txb6? .td4t-+] White is
worse, but he can continue to resist.) 29.ttJxd5
�xd5 30 . .tf2 �c6 3 1 .bxc4 �xc4 32.l"i:e4 �c6
33.l"i:xf4 .txe5 34.l"i:e4 .tf6 White's position is
difficult but not hopeless.
a
27... lL'lf5!
b
c
d
e
f
g
h
The e-pawn is not going anywhere, so
Karpov postpones taking it and improves his
position first.
27 . . . .txe5 2B . .txe5 (2B . .tf2 .tf6 29.l"i:e l �c8
30.l"i:xd5 l"i:xe l t 3 1 ..txe l ttJ f5 White's king
is dangerously open.) 2B . . . l"i:xe5 29.l"1xe5
l"i:xe5 30.�d2 White fails to regain his pawn,
nevertheless he obtains some counterchances:
30 . . . �e7 3 1 .'kt>fl ttJ f5 (3 1 . . .�g5 32.ttJb4 tDf5
33.ttJxd5) 32.l"i:xd5 (32.�xf4? l"i:e2) 32 ... l"1xd5
33.�xd5 �h4 34.c4 White has realistic hopes
to survive thanks to his passed pawn on the
queenside.
28.'lWdl h5!
Karpov defends the f4-pawn in an original
way. This move also has a prophylactic element
as it opens the back rank, just in case.
29.gel a5!
Another strong move with a rook's pawn!
This time Karpov defends the d5-pawn
indirectly by taking away the b4-square from
the enemy knight.
30.�f1
Short may well have been short of time, but
anyway there is little he can do but sit and
wait.
1 997
38 1
Nigel Short - Anatoly Karp ov
30 .. .'IWc7 3 1 .<i>gl
8
7
6
5
4
3
a
2
1
a
31...<i>h7!?
b
c
d
e
f
g
h
Karpov makes his trademark prophylactic
king move, making every possible improvement
before collecting the doomed pawn.
Nevertheless 3 1 . . . �xe5 32.�xe5 l"lxe5
33.l"lxd5 Wb6t was winning as well.
b
c
d
e
f
g
h
37 . . . l"le5!
37 ... Wc4 allows 38.l"lxf5 ! gxf5 39.ttJd4 mh7
40.Wxf4 when White is very much alive.
38 .Wxe2
38.l"ld3 l"le7 39.l"ld5 Wc4! With the rook
on e7 the sacrifice is harmless: 40.l"lxf5 gxf5
4 1 .ttJd4 mh7-+
38 . . . l"lxe2 39.ttJd4 ttJ e3 40.ttJxe2 ttJxd5 4 1 . mf2
mf6
Despite the inferior pawn structure, Black
has excellent winning chances.
32.b4
Short weakens the c4-square, which hastens
the end slightly. An alternative finish might
have been 32. mf2 �xe5 33 .�xe5 l"lxe5 34.l"lxe5
:1'1xe5 35.l"lxd5 (35 . ttJ d4 ttJ e3) 35 . . . l"lxd5
36.'lWxd5 Wxc3 37.Wxf7t ttJ g7 38.ttJe1 Wb2t
and Black wins.
32 ... a4 33.<i>f1?!
33.ic5!
Stopping the queen invasion would have
resisted harder:
33 . . . l"lxe5 34.l"lxe5 Wxe5 35 .�d4
35.l"lxd5 Wxc3 36.Wxc3 �xc3 37.mf2 �f6
38.g3 fxg3t 39.hxg3 h4 40.gxh4 (40.g4 h3)
40 . . . �xh4t 4 1 .mfl ttJ g3t wins.
35 . . . We2
After 35 . . . ttJxd4 36.ttJxd4 �f6 37.mfl
(37.ttJxb5? �h4-+) 37 . . . �h4 38.ttJe2
'lWf5 39.Wd 1 Black is better but the game
continues.
36.�xg7 mxg7 37.l"lxd5
8
7
6
5
4
3
2
1
a
33 ...Wc4!
b
c
d
e
f
g
h
Before taking the pawn Karpov improves his
queen with decisive effect.
34. <i>gl he5
Finally Karpov devours the e5-pawn, under
conditions which leave no chances for his
opponent to survive.
The Prime Years
382
I Game 41 I
35.i.xe5 gxe5 36.gxe5 gxe5 37.llJd4
After 37.'it>f2 h4 38.Wd l (38.h3 lLlg3)
38 . . . h3 39.gxh3 Wa2 Black wins.
-
Dortmund 1997
8
7
l.d4 llJf6 2.c4 e6 3.llJa d5 4.tLlc3 ie7
5.�g5 h6 6.�h4 0-0 7.e3 llJe4
6
5
8
4
7
3
6
2
1
Anatoly Karpov Artur Yusupov
5
a
37 ... ge3!
b
c
d
e
f
g
h
By exchanging the rooks Karpov wins a
second pawn.
38.gxe3 fxe3 39.�dl �xc3
0-1
In the final round Karpov faced Gelfand, who
equalized with the Semi-Slav. The queenside
pawns disappeared in the endgame and the
game looked to be heading for a certain draw,
but the players kept fighting and Karpov
incredibly got a dead lost position, but
somehow Gelfand spoiled it and the game
ended in a draw after all. Karpov's final score
of 5/9 was enough to share third place with
Salov and Topalov, a point behind Anand and
Kramnik who tied for first.
Karpov's next slow-play event was Dortmund,
where he started with the black pieces against
Kramnik. Karpov was unable to equalize in a
mixture of the English and Queen's Indian,
and Kramnik beat him convincingly. Karpov's
opponent in the second round was Artur
Yusupov.
4
3
2
1
a
b
c
d
e
f
g
h
The Lasker Defence had been a mainstay of
Yusupov's repertoire since the late eighties, and
used in his matches with Karpov.
8.i.xe7 �xe7 9.gc1 c6 10.�d3
I I .gxc3 dxc4
lLlxc3
The usual choice, although three years prior
to this game Yusupov tried 1 1 . ..lLld7 and drew
against Nikolic.
12.i.xc4 llJd7 13.0-0
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
383
Anatoly Karpov - Artur Yusupov
1 997
13 b6
18 a5!?
Here is one of the crucial games from the
1 989 candidates semi-final: 1 3 . . . e5 1 4 .i.b3
exd4 1 5 .exd4 lU f6 1 6.1'l:e 1 Wfd6 1 7.lUe5 lU d5
1 8.1'l:g3 i.f5 1 9.Wfh5 i.h7 20.Wfg4 g5 2 1 .h4 f6
22.hxg5 hxg5 23.f4 1'l:ae8
Yusupov commits his queenside pawns
to a rigid formation in order to restrain the
enemy b-pawn. After 1 8 . . . 1'l:c8 1 9.1'l:c1 1'l:bc7
20.b4! Black was under pressure in Kramnik
- Kasparov, Las Palmas 1 996, although he
managed to hold a draw.
..•
•••
8
19.a3!?
7
6
5
Karpov returns the ball to Black's court and
discourages him from placing a rook on the
c-file.
bn�,J" ,'/n,,/;
V�/U'/'''nnJ�
4
19 ... 1'l:e8 20.1'l:dl 1'l:bb8 21 .h3
3
2
a
b
c
d
e
f
g
h
24.fxg5 fxe5 25.g6 i.xg6 26.dxe5 Wfe6
27.i.xd5 cxd5 28.Wfxg6t Wfxg6 29.1'l:xg6t <;t>h7
30.1'l:d6 White's energetic attacking play led to
a winning endgame which he converted with
ease, Karpov - Yusupov, London (7) 1 989.
14.id3 c5 15.ie4 1'l:b8 16JWa4
Karpov had used the alternative 1 6.Wfc2!? to
defeat Yusupov twice in 1 995, but now he feels
it is time for a change.
a
21.. 1'l:bd8
b
c
d
e
f
g
h
••
16 ... ib7 17.ixb7 1'l:xb7 18.Wfc2
It was worth considering the immediate
2 1 . . .e5!?, for instance: 22.dxe5 lUxe5 23.lUxe5
Wfxe5 24.1'l:cd3 Wfe4 Black certainly stands
worse here, but with no knights on the board he
should be able to exchange his three queens ide
pawns for White's two, leading to a probable
draw in the rook endgame.
8
7
6
5
4
22.1'l:cd3 1'l:c8
3
2
a
b
c
d
e
f
g
h
Yusupov decide not to force matters. The
following direct approach would not have
solved his problems: 22 . . . cxd4 23.1'l:xd4 (Also
after 23.lUxd4 lUe5 24.1'l:c3 1'l:d5 White is a
bit better.) 23 . . . lUc5 24.b4 axb4 25.axb4 lUa6
26.Wfc4! Karpov evaluates this position as
The Prime Years
384
clearly better for White; indeed, the a6-knight
will have a hard time finding a decent and
stable square.
23.d5!
Karpov poses a dilemma for his opponent:
should he try to live with a mighty passed
pawn or should he allow White to dominate
the d-file?
However after 27 . . . tLJ e4! Black's pieces have a
reasonable grip in the centre, and he should
be able to hold the position. 2S.tLJd4 Wfe5
29.tLJb5 1;Wf6 it is not easy for White to make
progress.
28.gxd8t �xd8 29.tlJe5 �d5
23 ... exd5
Yusupov chooses the lesser of the two evils.
Indeed, 23 . . . e5?! 24.d6 is depressing for
Black.
24Jhd5 tlJf6
8
7
a
6
30.tlJc4
5
c
d
e
f
g
h
Karpov fixes at least one enemy piece to the
defence of the b6-pawn. Both sides have a
pawn majority, but Yusupov's is static whereas
Karpov's is mobile and flexible.
4
3
2
1
b
30 ... tlJd7 3 1.h3 5?!
a
25J:�e5!
b
c
d
e
f
g
h
Karpov exchanges, as with fewer pieces on
the board he has better chances to invade.
Keep the rook on the d-file is not very effective:
25 .l"1:d6 l"1:bS 26.a4 tLJ e4 27.l"1:6d5 1;We6 Black
defends firmly.
25 ... 1;Wc7 26J3xe8t gxe8 27.a4 gd8?!
This is not a losing mistake, but it is a
step in the wrong direction for Black, as his
pieces will become tied to the defence of the
b6-pawn.
If 27 . . . tLJ d7 28.tLJd4! is strong.
Yusupov recommends 3 1 . . .h5 instead.
Another idea was 3 1 . . .f6 intending to advance
the king.
32.<'!lfi !
Karpov brings his king towards the centre. It
would be premature to gain space with 32.g4?
due to: 32 . . . fxg4 (32 . . . g6 is also possible, and
after 33.gxf5 gxf5 it will be harder for either side
to activate their king.) 33.hxg4 1;Wf3 (Black can
also ease the pressure with 33 . . . tLJ e5) 34.Wff5
Karpov evaluated this position as clearly better
for White, perhaps overlooking that Black can
safely pick up the b-pawn with 34 . . . 1;Wdl t
35 .�g2 1;Wxb3, when White has no more than
a draw.
385
Anatoly Karpov - Actur Yusupov
1 997
32 <i!;>f7 33.8
•••
Karpov enables his king to advance and gets
ready to create a passed pawn.
However, the immediate 36.g4!? was a good
alternative.
36 <i!;>d7 37.g4!
•••
33 <i!;>e7
33 . . . h5 would have prevented g4 but
committed Black's kingside pawns to a rigid
formation, and after 34.h4! Wie6 35.e4 fxe4
36.Wixe4 Black is struggling.
Karpov finally executes the intended pawn
break, and forces his opponent to make a
difficult decision just a few moves before the
time control. The timing was surely not by
accident . . .
34.<i!;>e2 'iNe6 3S.'iNc3 �f6!
37 <i!;>c6?
Yusupov blocks the diagonal with the knight
and tries to get his king to the queens ide. This
is a good idea, which might enable him to
exchange queens in certain variations.
The fact that Yusupov made a mistake
like this suggests that he was indeed short of
time. The king should have approached the
queens ide more carefully: 37 . . . 'it>c7! 3 8.Wic2
fxg4 (38 . . . g6 is arguably a better defensive try,
although White has a few unpleasant ideas
such as 39.'it>g3 and 39.Wib2.) 39.hxg4 tLJ d5
40.Wih7 Wif6 4 1 .Wie4 tLJ e7 42.c;t>g3 Black's
position is loose, but he has chances to keep it
together after 42 . . . tLJ c8 .
•••
•••
The immediate 35 . . . Wif6? was inadvisable:
36.Wixf6t 'it>xf6 37.g4 'it>e6 38.c;t>n White
improves his king and has excellent winning
chances. The attempt to obtain counterplay
with 38 . . . fxg4 39.hxg4 c;t>d5 fails after 40.f4
b5 4 1 .axb5 a4 42.'it>e2 and White wins.
8
7
6
5
4
3
2
1
a
36.�fl
b
c
d
e
f
g
h
White should not rush to exchange queens
with 36.Wie5?! as this involves a loss of time.
36 ... Wixe5 37.tLlxe5 c;t>d6 38.tLJc4t 'it>c6 The
black king reaches the queenside quickly,
allowing him to start exchanging pawns. 39.g4
fxg4 40.hxg4 tLJ d7 4 1 .e4 b5 Black is probably
drawing this position.
a
38.'iNeS!
b
c
d
e
f
g
h
Karpov ruthlessly exploits the mistake and
wins a pawn.
38 'iNxeS 39.lLJxeSt
•••
Now it is clear why c6 was such an
unfortunate square for the black king.
39 <i!;>dS 40.�c4 fxg4 41.�xb6t <i!;>c6
•••
The Prime Years
386
42.tLlc4 gxf3 43.�xf3
52.h4!
The extra pawn gives White a comfortably
winning endgame. Care is still required, but
for a technician of Karpov's ability the task is
not too difficult.
Karpov beautifully opens the way for his
king to invade.
52 tLlc6 53.a5 tLlb4 54.tLld2!
..•
The knight helps to open the path of
penetration.
54 tLlc6 55.a6 gxh4t
•. •
5 5 . . . lLl a7
winning.
56.hxg5t
'it>xg5
57.lLlf3t is
56.�xh4 �e6 57.�xh5 �d7
The king makes a beeline for the a-pawn, but
Karpov's two remaining pawns are enough to
seal the victory.
a
b
c
d
e
f
g
h
43 ... �d5 44.tLlxa5 g5 45.tLlc4 h5 46.tLld2
�e5 47.e4 tLle8
Trying to go after the b3-pawn was also
insufficient: 47 . . . Wd4 48.a5 lLl d7 49.a6
lLle5t 50.�g3 lLl c6 5 1 .lLl f3t 'it>c3 (5 1 . . .Wxe4
52.lLlxg5t Wd4 53.'it>h4 c4 54.bxc4 Wxc4
5 5 . Wxh5 White wins.) 52.e5 Wxb3 53.e6
White promotes a pawn.
48.�e3 tLlc7 49.tLlc4t �f6 50.�fl llJa6
5 1 .�g3 tLlb4
8
58.�g6 �c7 59.tLlc4 �b8 60.�f6 rll a7
61 .e5 �xa6
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
7
62.e6 �b5 63.e7
1-0
5
In the third round Karpov was somewhat
worse against Hubner in the Short Variation of
the Caro-Kann, but the German grandmaster
let his advantage slip away and the game ended
in a draw. In the next two rounds Karpov drew
very quickly with Judit Polgar and Gelfand. In
Round 6 Karpov faced Topalov in a Fianchetto
Grunfeld with a symmetrical pawn structure.
6
4
3
2
1
a
b
c
d
e
f
g
h
1 997
Anatoly Karpov - Vadim Milov
Simplifications ensued, and the players agreed
a draw in a level endgame.
In the next round Karpov faced Short, who
surprised him with l .d4 followed by the 4."\Mfc2
0-0 5 .e4!? variation against the Nimzo-Indian.
Karpov reacted with a suspicious line and Short
built up a promising attack. Karpov missed a
few opportunities to keep himself in the game
and paid the price. In Round 8 Karpov had
the black pieces again, and held a slightly
worse endgame against Ivanchuk after a Caro­
Kann main line. In the final round Karpov
faced Anand, who defended with the Queen's
Gambit Accepted, j ust as he had done in Dos
Hermanas earlier in the year. Karpov played a
novelty bur Anand reacted well, and a complex
middlegame petered out to an equal endgame
and the players agreed a draw. Karpov finished
in equal sixth place with a score of 4/9.
387
playing he eventually had to settle for a draw
in the endgame.
Karpov started the second half of the
tournament with the white pieces against
Pelletier. He got a comfortable edge against
the Semi-Slav, and soon won a couple of
pawns and converted his advantage smoothly.
He drew the next game quickly with Gelfand,
but then suffered a setback in the Caro-Kann
against Lautier, who caught his king in the
centre.
In the ninth round Karpov faced Vadim
Milov. The Russian-born Israeli-Swiss
grandmaster played six games against the world
champions, scoring one win, three draws and
two losses, both of which came against Karpov.
He drew two other games with the twelfth
champion.
I Game 48 I
***
Anatoly Karpov - Vadim Milov
Karpov's next event was a double-round-robin
tournament in Biel. In the first round he faced
Pelletier, who tried the Petrosian Variation
against the Queen's Indian. Karpov equalized
and later got a slight edge after some skilful
play. Pelletier became too optimistic and
snatched a pawn, but Karpov was able to catch
his king. In the next round Karpov had the
black pieces again, and was able to draw with
Gelfand in less than twenty moves. In Round
3 Laurier played the sharp Vienna Variation
against the Queen's Gambit. Karpov gave up
a pawn but later regained it while keeping his
initiative. Laurier had chances to equalize bur
was unable to cope with the problems and
Karpov beat him nicely.
In Round 4 Karpov got no advantage against
Anand's Semi-Slav, and the game was agreed
drawn in an equal endgame. In the next game
Karpov played the Nimzo-Indian against
Milov. The IQP middlegame was always
roughly level, and although Karpov kept
Biel 1 997
l.d4 d5 2.c4 dxc4 3.e4 lbc6 4.lba
Karpov deviates from the following game
which he won two years earlier: 4.i.e3 tLl f6 5.f3
e5 6.d5 tLl d4 7.i.xd4 exd4 8.Wxd4 c6 9.tLlc3
i.b4 1 O.We5t We7 1 1 .Wxe7t <;t>xe7 1 2.dxc6
bxc6 1 3.i.xc4 tLl d7 1 4 .0-0-0 tLl e5 1 5 .i.e2 g5
1 6.h4 gxh 4 1 7.l'!xh4 i.e6
a
b
c
d
e
f
g
h
388
The Prime Years
1 8.Elh5! Karpov mobilizes his central pawns
in a most unusual way, by using his rook on
the edge of the board. 1 8 . . . lLlg6 1 9.93! Karpov
returns his extra pawn in order to put his
central pawns in motion. 1 9 . . . .ixc3 20.bxc3
.ixa2 2 1 .Eld2 .ie6 22.f4 .id7 23. lLl f3 a5 24.5
White obtained excellent play in the centre
and pressed home his advantage convincingly,
Karpov - Piket, Groningen 1 99 5 .
Technically this posmon was new at the
time, although Azmaiparashvili had already
carried out the same plan before castling. The
whole variation was relatively fresh at the time.
The knight manoeuvre is useful in supporting
White's queenside play.
12 ....id7
Milov keeps his bishop in order to have
better chances for a successful kingside attack.
4 ....ig4 5.d5
This ambitious space-gaining move is the
main line.
5 ... tlJe5 6..if4 tlJg6 7..ie3
7 . .ig3 is occasionally seen, but after 7 . . . e5
the bishop is less than ideally placed.
7... tlJf6 8.l£lc3 e5
8 . . . e6 is playable, although 9.Wa4t! Wd7
1 0.Wxd7t 'it>xd7 1 1 .lLlg5 is a critical test.
9.,bc4 a6 10.0-0
More players have carried out the
forthcoming minor piece manoeuvre before
castling, but it is unlikely to come to any more
than a move order difference.
13Jkl V!fe7 14.a3
In his Chess Informant analysis Karpav
mentions 1 4 .g3!?, but in the game he probably
felt that it was better to press ahead on the
queens ide without delay.
10 ....id6 1 1 ..ie2 0-0
8
7
14... h5!?
Before launching his attack, Milov extends
the life-expectancy of his important dark­
squared bishop.
6
5
4
Black in unable to exchange the d5-pawn
under favourable conditions with 1 4 . . . c6?! in
view of 1 5 .lLl c4.
3
2
1
Karpov's long-time helper Mikhail Podgaets
reached the same position later the same year;
they may well have worked on this variation
together. 1 2 . . . .ixe2?! 1 3.Wxe2 Over the next
few moves Podgaets purposefully and skilfully
develops pressure on the queenside. 1 3 ... lLld7
1 4.lLlb3 We7 1 5 .Elac l Elfc8 1 6.Elc2 lLlf4
1 7.Wd2 a5 1 8.g3 lLlh5 1 9.1Llb5 lLlhf6 20.f3
a4 2 1 .lLl c l .ib4 22.Wf2 lLl e8 23.lLld3 White
is close to breaking through on the queenside,
and he went on to win a nice game, Podgaets
- Raetsky, Biel 1 997. Podgaets showed that he
also possessed a fine positional touch. Sadly he
passed away in 2009.
a
12.tlJd2
b
c
d
e
f
g
h
If 1 4 ... lLl f4 15 . .if3 'it>h8 1 6.lLlc4 White
will eliminate the bishop and keep some
advantage.
1 997
Anatoly Karpov
15.tlJb3 tlJf4
Exchanging the backward pawn remains
inadvisable for Black: 1 5 . . . c6?! 1 6.dxc6 ixc6
1 7.lt:Ja5! White gets the advantage of the
two bishops, as after 1 7 . . . ixe4 ( l 7 . . . id7?
1 8.ltJ b7+-) 1 8.ltJxe4 ltJxe4 1 9.if3 Black loses
material.
16.if3
Keeping the bishop is a good idea, especially
as it could prove useful as a defender of the
kingside.
-
Vadim
389
Milov
1 7.g3 was also a decent move though:
1 7 . . . g5 ?! Black sacrifices a piece for the attack,
but the first player remains on top. ( l 7 . . . ltJ h3t
1 8.i>g2 h6 is sounder, although White stands
better here too.) 1 8.gxf4 ( l 8.ltJa5!?) 1 8 . . . gxf4
1 9.id2 ltJ g4 ( l 9 . . J%g8t 20.c;th l ih3 2 1 .We2
White holds the attack.) 20.ixg4 ixg4 2 1 .f3!?
ih3 22.�h l Ei:g8 23.Ei:gl Ei:xgl t 24.Wxgl
Ei:g8 25.Wf2 White is better, as pointed out by
Karpov. Black can win the queen but White
will have more than enough pieces in return.
17... g5
16... c;th8?!
Milov prepares the kingside onslaught, but
his move is a little slow.
Black can cut off the a2-knight with 1 7 . . . a5
but after 1 8 .ltJ c5 ic8 1 9.Ei:e 1 White remains
slightly better.
1 6 ... g5! ? is more to the point: 1 7.ltJa5
( 1 7.g3!?) 1 7 . . . g4 1 8 .ie2 ltJxe2t ( l 8 . . . b4
1 9.axb4 ixb4 20.ltJ b3) 1 9.Wxe2 ltJh5 20.ltJc6
White eventually lost although at this stage
he still has a slight edge, Lugovoi - Kharlov,
Rethymnon 2003.
18.tlJc5
Karpov brings his knights closer to Black's
camp. The fact that Black has to pay attention
to them takes some of the power away from
his attack.
18 .. J�g8 19.tLlb4 gg6
Another interesting idea is 16 ... a5! ? with the
idea of preventing ltJa5 as well as ltJ a2-b4.
8
7
8
6
7
5
6
4
5
3
4
2
3
1
2
a
17.tlJal
b
c
d
e
f
g
h
Karpov decides not to move a pawn on the
kingside.
a
2o.Wfc2!
b
c
d
e
f
g
h
Karpov follows his plan and focuses on
the centre, without being distracted by the
a6-pawn. In his annotations he mentions an
interesting alternative in 20.g3!? g4 2 1 .ie2,
but the game continuation seems stronger.
The Prime Years
390
20 ... g4 21 .i.e2 �ag8
It was too late to attempt to neutralize
White's queens ide play: 2 1 . . .a5 ?! 22.lDc6
lDxe2t 23.1&xe2 �xc6 24.dxc6 �xc5 25 .�xc5
1&e6 26.Ei:fd l Black is in trouble.
22.�fdl!
Karpov creates an escape square for his
bishop, and more importantly his king, which
may need to run to the west if Black's attack
gathers speed.
22.lDcxa6? would have been a mistake due
to 22 . . . lDxe4!.
23 . . . g3!!
Black has time to blast through on the
kingside. 23 . . . 1&xd7? is wrong because of
24.g3 or 24.�f1 .
24.fxg3
24.�xh5 gxf2t 25.mf1 lDxh5 White is in
trouble.
24 . . . lDxg3 25 .�xf4!
White has to eliminate this important piece.
Karpov stops here evaluating the position
as clearly better for White, but the attack is
not yet over. 25.�f3 lDh3t! 26.gxh3 lD f5t
27.�g5 Ei:xg5t 28.<J?hl lDe3 gives Black
excellent attacking chances.
22 ... tiJ6h5
Black has a lot of pieces on the kingside,
but he must pick the right time to strike.
The hasty 22 . . . lDxg2? is no good: 23.mxg2
lDh5 24.lDxd7!? g3 (24 . . . 1&xd7 25.mf1 wins.)
25 .�xh5 gxh2t 26.�xg6 1&h4 27.mf3 Ei:xg6
28.<J?e2 And White wins.
8
7
6
'nn,J"'-""
5
4
L""J',7='",,,,,/,,nn,
3
2
a
b
c
d
e
f
g
h
25 . . . lDxe2t! 26.1&xe2 exf4
26 . . . Ei:xg2t 27.1&xg2 Ei:xg2t 28.mxg2 exf4 is
also interesting; White has a lot of material
for the queen, but his king lacks shelter.
27.e5 �xb4 28.axb4 1&xd7
Black will take on g2, and he is at least not
worse as White's king is rather exposed.
23.g3!
a
b
c
d
e
f
g
h
Karpov defends with special skill, and
appreciates the need to prevent the advance of
the enemy g-pawn, which would have brought
Black an extremely dangerous attack.
23.lDxd7?
Karpov gives an exclamation mark to this
move but he is not right.
23.�f1 ?
White tries to bolster his kingside, but the
same idea works well again:
23 . . . g3! 24.hxg3?
This natural move is refuted beautifully.
The correct defence is 24.fxg3! lDxg3 2 5 . hxg3
with two options for Black:
a) 25 . . . lD h3t 26.mhl (26.gxh3? Ei:xg3t
27.mf2 1&f6t 28.me2 �xh3-+) 26 . . . Ei:xg3
27.�g l lDxgl 28.<J?xg l �g4 29.Ei:d3 �f3
1 997
30.:B:xf3 The position is balanced.
b) 2S . . . ttJxg2 26.i.xg2 (26.ttJc6 i.xc6
27.i.xg2 :B:xg3 2B.dxc6 Wh4 White must
be careful to avoid troubles.) 26 . . . :B:xg3
27.ttJxd7 Wxd7 2B.<j;lfl (2B.:B:d2 :B:xe3
Black has decent attacking chances for the
two pawn deficit.) 2B . . . :B:xg2 (2B . . . :B:xe3?
29.:B:d3!) 29.Wxg2 :B:xg2 30.<j;lxg2 Wg4t
3 1 . <j;lf2 i.e7 The position is roughly equal as
White's king is exposed.
24...Wg5 25.J.fl!
2S.ttJd3! ? was also strong.
25 .. J3h6
a
26.Wc3?
a
b
c
d
e
39 1
Anatoly Karp ov - Vadim Milov
f
24 ... ttJh3t!!
This brilliant sacrifice opens up White's king
decisively.
24 . . . ttJxg3?! 2S.fxg3 transposes to 24.fxg3!
above.
25.gxh3
2S.Wh2 ttJxg3 26.fxg3 :B:xg3 27.i.f2 Wh4-+
25 ... Wh4! 26.i.g2
26.Whl :B:xg3 27.fxg3 ttJxg3t 2B .Wh2
tDxfl t-+
26.ttJxd7 ttJxg3 27.fxg3 Wxg3t 2B .i.g2
Wi'xe3t 29.Wfl :B:xg2 30.Wxg2 :B:xg2 3 1 .Wxg2
Wi'xe4t-+
26 ... tDxg3 27.fxg3 Wxg3 2B.:B:d3
28.tDxd7 Wxe3t-+
28 ... i.xh3 29.Wf2 Wg4 30.<j;lfl i.xg2t 3 1 .We 1
ixe4
White has no satisfactory defence.
23 ...i.c8 24.lLlc6!
Karpov is not interested in taking the a-pawn,
as he has spotted a much juicier target on eS.
b
c
d
e
f
g
h
Up to this point Karpov has played a great
strategic game, but his last move is an error.
Correct was: 26.ttJd3! f6 (26 . . . ttJ f6 27.ttJdxeS
WhS 2B.h4! [Strongest, although 2B.f3
is also good.] 2B ... ttJ h3t 29.i.xh3 gxh3
30.i.xh6+-) 27.i.g2 as (27 . . . Wg7 2B .Wd2)
2B .Wd2 Black's attack grinds to a halt, and
he is almost in zugzwang. 2B . . . a4 29.:B:c3 :B:g7
30.:B:dc 1 i.b7 3 1 .ttJa7 White wins on the
queenside.
26 ... lLlf6!
Milov finds the only move which keeps his
attack alive by freeing the h-file for the queen.
27.lLld3!
Karpov defends skilfully. Taking the piece
would have been a big mistake: 27.gxf4? WhS
2B.i.g2 Wxh2t 29.Wfl exf4 30.i.d4 :B:gg6
3 1 .ttJeS WgB Black's attack is dangerous, and
possibly winning.
27...Wh5
Milov keeps attacking.
The Prime Years
392
27 . . . 1"i:xh2? is premature: 2B.i.xf4 (Simplest,
although 2B.tt'lxf4 should also win.) 2B . . . exf4
29.Wxh2 fxg3t 30.fxg3 Wh4t 3 1 .Wgl i.xg3
32.i.g2 Black's attack is insufficient.
Black has another playable move.
27 . . . tt'lxe4!?
This leads to a very sharp tactical fight,
although White keeps an edge provided he
finds a series of accurate moves.
28.h4 gxh3??
Milov makes a fatal blunder; he may have
been short of time.
2B . . . tt'lxe4!
This move does not equalize fully, but it
would have kept Black very much in the
game.
29.tt'lxf4 Wf5
8
7
6
5
4
3
V'= �" " "�"
2
a
a
b
c
d
e
f
g
h
2B.i.xf4!
2B.tt'ldxe5 tt'lxc3 29.tt'lxf7t <;!{ g7 30.tt'lxg5
tt'l ce2t 3 1 . <;!{ h l tt'lxc l 32.i.xf4 tt'l b3 33.1"i:d3
tt'l c5 34.1"i:e3 The position is roughly equal.
2B . . . Wh5 29.i.xe5t f6 30.i.xf6t tt'lxf6 3 1 .i.g2
Wxh2t 32.Wf1 1"i:fB
32 . . . i.f5 33.Wd4 defends.
33.1"i:e l 1"i:h5 34.1"i:e3
White keeps the advantage, as Black's king is
the more exposed.
b
c
d
e
f
g
h
30.Wc2
30.Wxe5t!? i.xe5 3 1 .tt'lxe5 1"i:eB 32.ttJfd3
1"i:d6 33.1"i:xc7 i.d7 White has decent
compensation for the queen, but the game
goes on.
30.tt'lg6t!? 1"i:hxg6 3 1 .Wc2 tt'l f6 32.1�%3
White has a slight edge as he will be able
to attack on the queenside with a4 at some
point.
30 . . . exf4 3 1 .i.d4t tt'l f6 32.Wc3
The evaluation is similar after 32.Wd2 l"le8
33.1"i:e l .
3 2. . . Wg7 33.1"i:e l
8
7
6
5
4
3
2
1
a
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
Anatoly Karpov - Utut Adianto
1 997
White is somewhat better, although the
position remains complicated.
8
7
6
5
4
3
2
1
,, , ,
��r� �
� � y:�
��
�%l �
�%,.,
�����,
' �CZJm � �B
'��."" ' :� �
r
�
�.
.�
�:W "' .8''IJ
��,�"",: ,�
���CZJ�
_ "" '� �f{j
%•
""
� �
�""%��r.��
_..t_
'Y;
" " '
a
%
29.tiJdxeS!
" " '
b
%
c
d
e
f
g
h
Karpov takes the weak pawn which he
identified several moves ago. In one fell swoop
White undermines the f4-knight, covers the f3square and opens the diagonal leading towards
the black king.
29 .. J:�g7 30.i.xf4
Black has absolutely no compensation for
the piece.
played at a rapid time limit. Your author does
not consider blind chess worth organizing, nor
the results worth mentioning. Next were two
classical games. The first was a 4.'lWc2 Nimzo­
Indian, which had all the ingredients of a
classic Karpov win. White got an edge with
the two bishops, manoeuvred skilfully, found
a tactical breakthrough at the right moment,
and then skilfully converted his advantage in
a rook endgame. I had started to annotate the
game for this book when I noticed that the
winner of the game was in fact Adianto rather
than Karpov. . .
I n the return game Adianto played a Modern
Benoni and sacrificed a piece in return for
some pawns and an initiative. The position
simplified to a level endgame and a draw was
agreed. The final two games were played at a
rapid time limit. In the first Karpov equalized
against the Exchange Ruy Lopez and a draw
was agreed in a blocked middlegame position.
We will look at the second rapid game, which
was also the last of the match.
Adianto played nine classical games against
the world champions, scoring one win, four
draws and four losses.
30 ... tiJxe4 3 1 .Wfe3 Wff5
\ Game 49 \
Milov sacrifices more material, so Karpov
j ust takes it and wins. At least one of the players
must have been very short of time.
32.hh6 h2t 33.�xh2 liJill 34 ..bgn
i>xg7 3SJ3d4
1-0
Karpov finished the tournament with a ten­
move draw against Anand. His score of 6Y2/ 1 0
was enough to take second place behind the
Indian grandmaster.
Karpov's final event of 1 997 was a mixed-format
match against Utur Adianto, the Indonesian
number one. First there were two blind games
393
Anatoly Karpov - Utut Adianto
Jakarta (rapid) 1997
l.e4
In 1 997 Karpov did not use this opening
move in a single rated game, so it must have
come as a surprise to Adianto.
1 ... cS
The Indonesian grandmaster's main opening
against l .e4 is the Caro-Kann, but he prefers
not to see what Karpov has prepared against
his own favourite defence.
The Prime Years
394
2.ttla d6 3.d4 cxd4 4.ttlxd4 lLlf6 5.ttlc3 a6
13.ttlxa5 �xa5 14.�d2
A risky decision. It is true that Karpov
stopped playing l .e4 because he was unable
to find a convincing answer to Kasparov's
Najdorf/Scheveningen. On the other hand,
when he was not facing the best player of all
time, Karpov scored well from White's side.
He must also have done a huge amount of
work on this opening during the Kasparov
matches.
Karpov sets up the threat of tLJd5. In the
Kasparov - Anand world championship
match of 1 995, the Indian grandmaster chose
the d3-square for his queen, but in that game
the moves <;t>gl -h l and . . . Ei:f8-e8 moves had
been inserted. It is worth taking a quick look
to admire Anand's masterful handling of the
position.
6.i.e2 e6 7.£4 �c7 8.a4 i.e7 9.0-0 0-0
10.i.e3 ttlc6 H .i.a!
Adianto was not a Scheveningen specialist, so
Karpov deviates from the main line of 1 1 .<;t>hl
in order to avoid his preparation.
1 1 ... i.d7?!
a
b
c
d
e
f
g
h
1 5 .�d3 Ei:ad8 1 6.Ei:fd l ! i.c6 1 7.b4 �c7 1 8.b5
i.d7 1 9.Ei:ab l ! axb5 20.tLJxb5! i.xb5 2 1 .�xb5
Ei:a8 22.c4 e5 23.i.b6! �c8 24.fxe5 dxe5 25 .a5
i.f8 26.h3 �e6
a
12.ttlb3!
b
c
d
e
f
g
h
Korchnoi was the player who introduced the
early . . . i.d7 in this type of position. Black's
idea is to take on d4 and put the bishop on c6,
so Karpov prevents it.
12 ... ttla5
Exchanging the knight makes it hard for
White to carry out a pawn storm with g4, on
the other hand it encourages White to go for a
positional handling, which is hardly a problem
for Karpov. The other main line is 1 2 . . . b6.
a
b
c
d
e
f
g
h
27.Ei:d5! tLJxd5 ? 28.exd5 White obtained a
winning position and converted his advantage,
Anand - Kasparov, New York (9) 1 99 5 .
14...�c7
If 1 4 . . . <;t>h8 then 1 5 .�f2 is a good answer.
1 997
Anatoly
Karpov - Utut Adianto
395
Karpov improves his knight. 1 9.axb6 would
have transposed to 1 8 . . . b6 in the previous
note.
1 9 ... l'!d8!?
This witty move works well in the game,
but objectively Black should have preferred:
1 9 . . . Wb7! 20.Wd lLl c5 2 1 .l'!e l b4! 22.c3 f5
Black is not at all worse, as he has enough play
against the e4-pawn.
20 . .!Llb4 .ib7
15.a5!
a
b
c
d
e
f
g
h
Karpov gains space and fixes the b 7pawn. Giving up space is nothing new in
many Sicilians, especially the Scheveningen.
However, in the present position it is not so
easy for Black to counterattack as none of
his potential pawns breaks are particularly
attractive: . . . d5 would lose a pawn, . . . e5 would
weaken the d5-square, and with the white
g-pawn remaining on g2 there is little chance
of . . . f5 working well. If Black wishes to obtain
counterplay then the one remaining option is
to prepare . . . b5.
8
7
6
5
4
3
2
a
21 ..id4?
15 ....ic6 16 ..ib6 'lWb8 17JUdi
Improves the rook and prevents . . . .id8.
17 ... .!Ll d7 18 ..ifl b5
Black can also force the exchange of the
b-pawn: 1 8 . . . b6 1 9.axb6 (there is also 1 9.b4
b5 20.We3 .ib7 2 1 .l'!d3 Wc7 22.lLl e2;!;)
1 9 . . . lLlxb6 20.We2 (or 20.Wd4 lLl d7 2 1 .l'!a2;!;)
20 . . . Wb7 2 1 ..id4 lLl d7 22.l'!a2 White is a bit
better.
In the event of 1 8 . . . l'!c8 White has several
promising ideas, including 1 9.b4, 1 9.Wd3
and 1 9.1Lla4, each of which maintain some
advantage.
19 . .!Lla2!?
b
c
d
e
f
g
h
It looks like Karpov overlooked his oppo­
nent's next move. Instead after the superior
2 1 .c3! White stabilizes his well-placed knight
and keeps a slight plus.
21.. . .!Lle5! 22 ..ie2?
If one plays over this game quickly, one
might get the impression that it was almost
exclusively a strategic battle, but the reality is
that tactics played a significant role; or at least
they might have done, had Black seized the
tactical opportunity that was available to him
on the next move.
Objectively Karpov should have allowed his
bishop to be exchanged, after which he has
no advantage but should not be in trouble
either.
The Prime Years
396
22 ... tLle6?
Sadly for Adianto and the local fans, he
misses a chance to take over the initiative. Of
course we should remember that the game was
played under a rapid time limit.
22 . . . 4J c4 was one improvement over the game
continuation, and after 23.'\We l dS 24.b3 4J d6
2S.exdS 4J fS 26.�b6 'lWxf4 the position is
rather complicated.
22 . . .�xe4!
This capture would have brought Black the
advantage, as White is unable to exploit the
apparent instability of Black's minor pieces.
23.'lWe3 4J c6! 24.4Jxc6
24.'lWxe4? 4Jxb4 2S.fS (2S .�b6 dS-+)
2S . . . dS 26.'lWg4 eS The attack gets nowhere
and Black's material advantage decides the
game.
24 . . . �xc6
It is not easy for Black to exploit his extra
pawn, but with precise play he should be able
to take over eventually.
23.liJxe6 Le6
8
7
6
5
4
3
2
1
a
24 ..id3
b
c
d
e
f
g
h
Defends the e4-pawn while pointing the
bishop ominously towards the black kingside.
24...�b7 25.�e3
The queen prepares to switch to g3.
25 .. J:�ae8 26.c3 �e8
It is hard to suggest a constructive plan for
Black; there is little he can do except wait.
27.�g3 .ifS 28.�el
a
b
c
d
e
f
g
h
2S .�d3! ?
I f White tries t o d o something tactical
instantly, then Black's position withstands
the pressure.
2S.fS eS 26.�b6 2'l:c8 Black is fine.
2S.�xg7 <;t>xg7 26.'lWc3t �f6 27.'lWxc6 and
now 27 . . . 1"k8, 27 . . . dS and 27 . . . �xb2 all
leave White struggling.
2S .'lWc3 2'l:c8 26.�xg7 eS! 27.'lWg3 'lWa7t
28.<;t>hl exf4 Black is winning.
2S . . . 'lWb7 26.'lWg3 �f8 27.2'l:e l
Karpov makes room for the other rook to
come to d l .
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 997
397
Anatoly Karpov - Utut Adianto
28 ... b4
Adianto decides to change the pawn structure
on the queenside. In doing so he creates an
outpost on c3 for his bishop, but runs the risk
of the b-pawn becoming weak later.
Sitting and waiting was not ideal: 2s .. JkdS
29.f5 (White can also consider 29.E:ad l
Wfe7 30.e5 [or 30.f5 e5 3 1 .i.b6] 30 . . . i.d5
3 1 .i.c2 E:d7 White has the better prospects,
and can choose between the immediate
32.5 and further preparation with 32.E:e2.)
29 ... e5 30.i.b6 E:d7 3 1 .E:e3 (3 1 .f6!? g6 32.E:e2
The advanced f-pawn makes Black's life
unpleasant.) 3 1 . . .i.e7 32.b3 h6 33.c4 Black's
position remains passive and unpleasant.
2S . . . e5!? Perhaps this was the best practical
chance; Black tries to sacrifice a pawn to free
his position.
White could also have considered direct action:
29.f5!? e5
a
b
c
d
e
f
g
h
30.i.b6 bxc3 3 1 .bxc3 d5 (3 1 . . .i.e7 32.E:ab l
\'waS 33.c4 Black's pieces are passive, and with
\,Wg4 coming soon White's attack looks nice.)
32.exd5 i.xd5 33.E:xe5 E:xe5 34.\'wxe5 i.xg2
35 .i.xa6 \'wxa6 36.'it>xg2 White's extra pawn
gives him excellent winning chances despite
his exposed king.
29 ... E:ed8
a
b
c
d
e
f
g
h
29.i.b6!? White can decline the offer and
go for a positional edge. (After 29.fxe5 dxe5
30.i.xe5 [There is also 30 . . . E:e6 when Black's
pressure against the e4-pawn gives him chances
to hold.] 30 . . . i.c5t 3 1 . Whl g6 Black has some
compensation.) 29 . . . i.e7 (29 . . . exf4 30.\'wxf4
Black is worse here too.) 30.f5 i.f6 White
keeps a slight plus although Black is solid.
29.c4
Karpov gains space and leaves the b-pawn
on the board, thus giving himself an additional
target for later in the game.
a
30 ..ib6
b
c
d
e
f
g
h
Karpov decides to improve his pieces rather
than start an immediate attack.
30.f5 was also promising: 30 . . . e5 3 1 .i.b6
E:eS From here White has several good moves
including 32.E:e3, 32.f6 and 32.\,Wg4, all of
which give him some advantage.
The Prime Years
398
30 .. J�e8 31 .b3 g6 32J�adl!
In hedgehog positions Black rarely succumbs
to a direct attack on the d6-pawn, nevertheless
White can sometimes use the weakness to
distract the defending pieces from other
duties.
32 ....ig7
8
7
6
5
4
3
2
1
a
33.�e2!
b
c
d
e
f
g
h
Karpov makes another small improvement;
now Black must worry about a doubling of
rooks on the d-file.
33 ...Wfe7 34 ..ibl
35.�ed2
3 5 .e5!? Forcing matters in the centre gives
White an advantage, but not necessarily a
decisive one. 35 . . . d5 (35 . . . dxe5 36.Exe5 f6
37.�ed2 .ic6 38 . .id3 Exe5 39.c5;!;) 36 . .id3
f6 37.exf6 Wfxf6 3S.cxd5 (3S.�de 1 There are
a few targets in Black's position, which gives
nice play for White.) 3S . . . exd5 39.Wfe l �xe2
40.Wfxe2 Black has some problems, as his
weaknesses are more serious than White's.
3 5 . . . d5!
This looks to be Black's best chance.
35 . . . �c6 36.e5 d5 37.cxd5 exd5 3S.�xd5
�xb6 39.axb6 White's initiative may become
strong if he manages to transfer his bishop
to c4.
35 . . ..ifS 36.e5!? (The simple 36.�xd6 is also
strong: 36 . . . Wfxd6 37.�xd6 .ixd6 3S.e5 .ie?
Black's rooks are passive and after 39.h4
or 39 . .id3 his position is not much fun.)
36 . . . d5 37.�d4 (37 . .ie4 dxc4 3S . .ixg6 Exg6
39.�d7 c3! Black has counterchances thanks
to the strong c3-pawn; 37.h4!? is promising
though.) 37 . . . f6 3S.f5 ! (3S.c5 Wfg7 39.l"lxb4
l"lxc5! keeps Black in the game; instead after
39 . . . Exe5 ? 40.Exe5 �xc5 4 1 .�g4 his king is
caught.) 3S . . . exf5 39 ..ixf5 Exe5 40.l"lxd5
White has a tremendous attack.
Karpov continues to manoeuvre, but direct
action was also strong: 34.e5! ? dxe5 35.Exe5 f5
(3 5 . . . .ib7 36 . .ie4 White will invade on the
d-file.) 36.exf6 (36.'1W f4!?; 36.c5 .id5 37.Wfe3
�a8 38.Wff4 .ixb3 39.�b l .id5 40.�xb4
White has the upper hand on the queenside.)
36 ... Wfxf6 37.Wfg4 Black has problems with
the e6-pawn.
34....ic3
Adianto tries to stop White from lining up
his forces against the d6-pawn. The alternative
was:
34 . . . .ib7
Black could play this move with a view to
defending the d6-pawn, or sacrificing it.
a
b
c
d
e
f
g
h
36.Wff2!?
After 36.Wfe3 .ic3 37.�d3 Wfh4! 3S.cxd5
exd5 39.e5 f6 the play becomes double­
edged.
Anatoly Karpov - Utut Adianto
1 997
36.e5 f6 37 ..id3 is interesting, although it
is doubtful that White has much advantage,
if any.
Taking the pawn is more challenging:
36.cxd5 exd5 37.exd5 �c3 (37 . . . Wd6
38.�d3) 38.d6! Wd7 39.Wh4 �xb3 40.f5 !
Black i s living dangerously.
36 . . . .ic3 37.�d3 f5 38.exf5 exf5 39.�xd5!
This exchange sacrifice gives White excellent
chances.
39 . . ..ixd5 40.cxd5 We l t 4 1 .�xe l �xe l t
42.Wxe l .ixe l 4 3 . .id3 �a8 44.d6
White is in the driver's seat, despite his slight
material disadvantage.
399
a pawn.) 38.�e3 e4 39.�g3t (39.Wxd6 .ie5)
39 . . . �h8 40 . .id4t �e5 4 1 .We3 White will
soon develop play against the central pawns.
36.hd8
Taking the rook will certainly be sufficient to
win in the long run.
36 ... gxd8 37.V;Ye3 e5
a
38.V;Yb6!
b
c
d
e
f
g
h
From this square the queen ties up Black's
pieces.
a
b
c
d
e
f
g
h
38 ...i.b7 39.5
Adianto probably hoped to set up a fortress,
but there is little hope of it succeeding.
Karpov gains space and over the next
few moves he improves his position on the
kingside.
35 . . . Wb7 36.We3 (36 . .id4) 36 . . . d5 37.cxd5
exd5 38.e5 White is clearly better.
39 ... gd7 40.ge3 @g7 41.ga f6 42.h4
i.d4t
35.Wd3 ged8?!
35 . . . Wh4!? 36.Wxd6 Wg4 37.�e3 e5 38.f5 .id4
39.�xd4 exd4 40.Wxd4 White has two pawns
for the exchange and a positional advantage.
Adianto decides to reclaim some material
before the kingside threats become too
serious.
43.gxd4 exd4 44.V;Yxd4
35 . . . e5 36.f5 gxf5 37.exf5 (Also promising
is: 37.Wg3t �h8 38.exf5 Wf6 39.We3 �g8
40 . .ie4 White is clearly on top and can
continue to improve his position.) 37 . . . Wh4
(37 . . . e4 38.Wg3t �h8 39.�xd6 White wins
The rest of the game was probably played
under time shortage, and is not especially
important. Black has no compensation for his
pawn deficit and Karpov converts his advantage
without any problems.
400
The Prime Years
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
44...�e5 45.�fl ,he4 46.,he4 �xe4 47.fxg6 hxg6 48J�xf6 �bl t 49.cj;>h2 �a1 50J�e6 gf7
5 1 .�g3 gf6 52.�xd6 gxe6 53.�xe6 �xa5 54.�e7t cj;>h6 55.c5 �b5 56.�d6 a5 57.c6 'lWe2
58.�f4t cj;>h7 59.c7 �e6 60.'l1*/c4 �e5t 61 .cj;>h3 �f5t 62.cj;>g3 �e5t 63.cj;>fl �b2t 64.cj;>gl
�al t 65.cj;>h2 �e5t 66.g3 �b2t 67.cj;>h3 �bI 68.c8=ti' �hl t 69.cj;>g4 �dl t 70.cj;>g5 'lWh5t
71 .cj;>f6 ti'f3t 72.ti'f4 Wfc3t 73.�xc3 bxc3 74.Wfc7t
1-0
In 1 997 Karpov only took part in a few classical tournaments, instead preferring to concentrate
on rapidplay tournaments and other exhibition events. His results were certainly not on the same
level as they had been earlier in the nineties.
1997 Summary
Dos Hermanas (3rd-5th place) : 5/9 (+2 =6 - 1 )
Dortmund (6th-8th place) : 4/9 ( + 1 =6 -2)
Biel (2nd place) : 6Y2/ 1 0 (+4 =5 - 1 )
Match versus Adianto, Jakarta: Lost 1 Y2-Y2 (+0 = 1 - 1 )
Total 5 3 . 3 % (+7 = 1 8 -5)
.�
II Wins • Draws • Losses
199 8
Rating 2735 (6 in the world)
Karpov started the year by defending his FIDE World Championship title, which was contested
under the new knockout format. He was helped considerably by FIDE, as he was seeded directly
into the final. After a three-week series of elimination rounds, which finished on 30 December, it
was Anand who emerged as the second finalist after defeating Adams. All of Anand's qualification
matches took place in Groningen, but the final was in Lausanne.
By that time Anand outrated Karpov by forty five points, so he was definitely the stronger player
overall. The Indian grandmaster had a few other advantages: his mind was sharp after contesting
so many games over the previous few weeks, and had become accustomed to the time control
which involved an increment, which Karpov had not encountered in many previous events.
Nevertheless Karpov's advantages were more significant: Anand must have been tired after
contesting no less that seven elimination rounds, and he even had to travel between Groningen
and Lausanne. He must also have revealed a lot of his preparation during the knockout phase.
Karpov had the white pieces in the first game, and he surprised his opponent with a piece sacrifice
in the Meran. Anand did well to stay in the game, nevertheless Karpov obtained a winning
position which he eventually converted. The game is referenced in the notes to Game 34 on page
267. In Game 2 Karpov unveiled another opening surprise and used the Arkhangelsk Variation
against the Ruy Lopez. He got a decent position and sacrificed an exchange for two pawns. Later
in the complications he missed a win in time trouble, and Anand punished him with a series of
accurate moves to level the match.
In Game 3 Karpov took no chances and went for safety against the Semi-Slav. We will look at
the fourth game.
I Game �O I
Viswanathan Anand - Anatoly Karpov
Lausanne (4) 1 99 8
l.e4 c6 2.d4 d5 3.exd5 ad5 4.c4
Hitherto Anand had never played the Panov against Karpov. In their early meetings he used
sidelines such as 2.d3 and 2.lLlc3 d5 3.lLlf3, and later switched to 3.e5. Overall he had cracked
Karpov's Caro-Kann just once, and drew the other six games.
The Prime Years
402
Karpov used this move in approximately
half of his games from this position, favouring
6 . . . ib4 in the others.
17 ... hxg5 l S.hxg5 ie7 White's king is too
exposed for him to claim any advantage.
7.adS tiJxdS 8.i.d3 tiJc6 9.0-0 0-0 IO.gel
i.f6
Anand prepared well for the game and plays
a strong novelty. The threat of tLlxd5 cannot be
ignored, and highlights the fact that Black is
somewhat behind in development.
The alternative is 1 O . . . tLl f6, which Karpov
used as far back as 1 973 against Uhlmann.
1 l .i.e4 tiJce7 12.h4!?
This rare but dangerous move was first played
by Maya Chiburdanidze in 1 9S5. It does not
carry a direct threat, but is nevertheless a clear
statement of White's aggressive intentions.
8
7
6
5
4
13.Wd3!
1 3.ig5
This had been seen in one previous game. It
is not a bad move, but Black should be able
to equalize with accurate play.
1 3 . . . tLlxh4
1 3 . . . tLlxc3 14.bxc3 tLlxh4 1 5 .ixf6!? ( 1 5 .if4!?
White's strong bishops give him reasonable
compensation; 1 5 .ixh4 ixh4 1 6.�b l
2"1:bS 1 7.tLle5 White had compensation in
Hoogeterp - Van Rijn, Netherlands 1 995.)
15 ... Wxf6 1 6.2"1:b 1 tLlxf3t 1 7.ixf3 �b8
l S .Wa4 White's better development gives
him a promising initiative for the pawn.
1 4.ixh4 ixh4 1 5 .tLlxh4 Wxh4
3
2
1
a
12 ... tiJf5?!
b
c
d
e
f
g
h
Karpov wastes no time bringing the knight
over to help organize the defence against
Anand's aggressive idea.
1 2 . . . id7 is the most common move, and is
probably the better choice as well. 1 3 .'1Mid3
h6 ( 1 3 . . . tLl b4!? 1 4.ixh7t �hS 1 5 .Wc4 �xh7
[ 1 5 . . . tLlxa2 1 6.2"1:xa2 �xh7 1 7.ig5 Black's
position is somewhat inconvenient] 1 6.Wxb4
ic6 1 7.tLle4 White is slightly better according
to Timman.) 1 4.g4!? This aggressive move
has not yet been tested. ( 1 4.ig5 ic6 is okay
for Black.) 1 4 . . . tLl g6 1 5 .ixg6 fxg6 1 6.tLlxd5
exd5 1 7.g5 ( 1 7.Wxg6 ieS l S.Wd3 ixh4
Black regains the pawn with a good position.)
a
b
c
d
e
f
g
h
1 6.ixd5
1 6.tLlxd5 exd5 1 7.ixd5 ie6 l S.�e4 �f6
1 9.ixe6 fxe6 20.f3 White is fractionally
better, but Black should not have much
trouble holding this ending.
1 6 . . . exd5 1 7.tLlxd5 ie6 l S.g3 WdS 1 9.tLlf4
Wb6
Black also equalizes with 1 9 . . . Wd6.
20.d5 2"1:adS
Black has equalized.
403
Viswanathan Anand - Anatoly Karp ov
1998
1 4.'1M1'c4 a5
8
7
8
7
6
6
5
5
4
4
3
3
2
2
a
b
c
d
e
f
g
h
13 tl.hc3
Karpov decides to meet the threat by
exchanging knights, but now White strengthens
his centre and his bishop obtains the use of the
a3-square.
...
1 3 . . . g6?
Black has no time to strengthen his f5knight.
14.ixf5 ! exf5
1 4 . . . gxf5 1 5 .ih6 Ele8 ( 1 5 . . . tLl b4 1 6 .Wd2
l"Ie8 1 7.ig5 White may hurt Black's
inadequately covered king.) 1 6.tLlxd5 Wxd5
17.tLle5 Black's kingside is too weak.
1 5 .ih6 ig7 1 6.ixg7 <j;Jxg7 1 7.Ele5
Alternatively after 1 7.tLlxd5 Wxd5 1 8 .l"Ie5
\Wd6 1 9.Elae 1 White's pieces dominate.
1 7 . . . tLl f6
After 1 7 . . . ie6 1 8 .Elae l Black is in a
dangerous situation, for instance 1 8 . . . h6?!
1 9 .h5 is strong.
1 8.l"Iae l
Black will have a hard time completing
his development. 1 8 .d5 gives White some
advantage as well.
A few month later Portisch attempted to
improve on Karpov's play with:
1 3 . . . tLl b4!?
He failed to equalize, although this may have
due to a subsequent error.
a
b
c
d
e
f
g
h
1 5 .ig5
1 5 .a3 tLl d6 1 6.We2 tLlxe4 1 7.Wxe4 tLl c6
( 1 7 . . . tLl d5 1 8 .ig5 is unpleasant) 1 8 .ie3
id7 1 9 .tLlg5 ixg5 ( 1 9 . . . g6 20.d5) 2o.ixg5
f6 2 1 .ie3 White's advantage is small.
1 5 .if4!? tLl d6 ( 1 5 . . . tLlxd4 1 6.Elad l tLlxf3t
1 7.ixf3 Wb6 1 8 .ie3 Wa6 1 9 .Wc7 White
has excellent compensation.) 1 6.ixd6
Wxd6 1 7.a3 tLl c6 1 8 .l"Iad l Black's bishop
pair is a serious force, but it cannot match
the fact that all of White's pieces are in play.
1 8 . . . g6 1 9.d5 (Also promising is 1 9. tLl b 5
Wf4 [ 1 9 . . .W b 8 20.d5] 20.g3 Wg4 2 1 . tLl h2
and Black is getting pushed back.) 1 9 . . . tLle5
20.tLlxe5 ixe5 2 1 .dxe6 ixe6 22.Wb5 Black
is struggling.
1 5 . . . ixg5 1 6.hxg5 tLl d6 1 7.Wb3 tLlxe4
1 8 .Elxe4
a
1 8 . . . Ela6!?
b
c
d
e
f
g
h
The Prime Years
404
Portisch finds an imaginative way of
developing the rook. Nevertheless there was
an easier way to solve his opening problems:
I B . . . b6! 1 9 .a3 ( 1 9.lLle5 i.b7) 1 9 . . . i.b7! This
finesse is the key. 20.axb4 axb4 2 1 .l''lxa B
WxaB Black is fine.
1 9.a3 lLl d5 20.lLlxd5 Wxd5 2 1 .Wxd5 exd5
22.i"le5 i.e6 23.i"lc 1 i"lb6 24.i"le2 i.g4?!
24 ... i"lb3! would have given Black excellent
chances to hold the ending.
25 .i"lc5
White won a pawn and later the game,
Timman - Portisch, Frankfurt 1 99B.
14.bxc3
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
14... h6
Karpov's play is driven by the need to respond
to threats. This time he saves the h-pawn.
1 4 . . . g6?
Normally this move would strengthen the
kingside, but here it allows White to get
closer to it:
1 5 .i.xf5 !
White remove an important defensive piece
and changes the ratio of developed pieces
even more in his favour.
1 5 . . . exf5
1 5 . . . gxf5 1 6 .i.g5 b6 1 7.i"le3 i.b7 I B .lLle5
Black's king comes under heavy fire.
1 6.i.h6 i"leB 1 7.d5
The d-pawn will be hard for Black to live
with. 1 7.i.g5 i.e6 I B .c4 also gives White
some advantage.
1 7 . . . Wd7
Or 1 7 . . .i.d7 IB .i.g5 .
I B .i"lxeBt WxeB 1 9 .i"le l
White has an excellent position.
15.h5
The move itself is not a mistake, but
in conj unction with the following one it
squanders White's advantage.
1 5 .i.a3 lLl d6
After 1 5 . . . i"leB White has more than one
route to an advantage:
a) 1 6.i.xf5 exf5 1 7.i"lxeBt WxeB I B .1"le 1
WdB 1 9 .c4 White's d-pawn has the potential
to become strong. (After 1 9.Wb5?! ie6
20.Wxb7 i.d5 Black has nice play for the
pawn, as pointed out by Karpov.) 1 9 . . . b5!?
Karpov mentioned this interesting pawn
sacrifice. 20.cxb5 i.e6 2 1 .i.c5 Black has
some compensation but objectively White
must be better.
b) 1 6.h5 i"lbB 1 7.i.xf5 ( 1 7.lLle5 i.xe5 should
be okay for Black.) 1 7 . . . exf5 1 B .i"lxeBt 'lWxe8
1 9.i.d6 i"laB 20.i"le 1 Black has to be careful
as his queenside is undeveloped.
1 6 .i.h7t
White saves the important bishop.
1 6.h5 reaches the note to White's 1 6th move
in the main game.
1 6 . . . cj;JhB
a
b
c
d
e
f
g
h
Viswanathan Anand - Anatoly Karpov
1998
I 7.Elad l ! ?
White wants to open the position before the
enemy pieces come into play.
I 7.h5 is also promising, for instance:
I7 . . . b5?! 1 8.'lWe2! ( I 8 .�e4 ElbB 1 9 .�c5
Wic7 20.a4 Black will have problems coping
with his opponent's powerful bishops.)
IB . . . c;t>xh7!? Otherwise White will retreat
his bishop to c2 and create awkward mating
threats. 1 9 .�xd6 'lWxd6 20.'lWe4t �g8
2 1 .'lWxaB 'lWb6 22.'lWe4 �b7 23.'lWd3 �d5
Black's strong bishops give him chances to
resist, but ultimately White must be winning
this position.
I 7 . . . 'lWc7!
This was suggested by Karpov, and looks like
the best chance to equalize.
I7 . . . g6 is risky: I B .�xg6 fxg6 1 9.d5!
( l 9.'lWxg6? �g7) 19 ... ttJ f5 ( I 9 ... e5? 20.'lWxg6
if5 2 1 .'lWxh6t+-) 20.�xf8 'lWxfB 2 1 .dxe6
White has the upper hand.
I 8 .d5 EldB
I B . . . exd5 ? 1 9.'lWxd5 wins.
I 9.dxe6 ttJ c4
a
b
c
d
e
f
g
h
20.'lWxdBt!
The temporary queen sacrifice wins a pawn.
2o . . .ixdB 2 1 .ElxdBt 'lWxdB 22.exf7 �g4
23.fB='lWt 'lWxfB 24.�xfB c;t>xh7 2 5 . ttJ h2
White keeps an extra pawn, but his split
queenside pawns combined with the reduced
material offer Black good drawing chances.
Karpov was able to hold a similar type of
405
endgame against Portisch in the 1 975 Milan
play-off.
15 ... lLld6
Karpov tries to ease his
exchanging the e4-bishop.
suffering by
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
16.tlJe5?
This allows Black to exchange too many
pieces. Anand must have overestimated his
chances in the ensuing position .
1 6.�a3
This move would have posed an awkward
dilemma: Black must either attempt to
neutralize his opponent's initiative, or
sacrifice an exchange for slightly less than
full compensation.
After the game the players took questions
from spectators for a few minutes. When
asked why he rejected the development of
the bishop to a3 , Anand's response was short
and to-the-point: "I thought my move was
stronger."
1 6 . . . ttJxe4!?
The exchange sacrifice is arguably the best
chance to save the game.
After 16 . . . EleB 1 7.�h7t c;t>h8 I B .ttJe5 �xe5
1 9.Elxe5 Black is rather passive.
1 6 . . . �e7 1 7.�h7t �hB I B .ttJe5 is also
unpleasant for him.
The Prime Years
406
1 7.ixfB!
White should take the opportunity to win
material.
If 1 7.iWxe4 Ei:eB Black can gradually equalize
by developing his queenside pieces. I B .tLle5
( l B.c4 iWc7) IB ... iWc7 1 9 .Ei:ab l ( l 9.Ei:e3
id7) 1 9 . . . Ei:bB 20.tLlg4 ig5 Black is close to
solving his problems.
1 7 . . . tLlxf2 I B .mxf2 iWxfB
Black does not have full compensation for
the exchange, although he certainly has some
practical chances to save the game.
White also had a second promising path:
1 6 .ih7t! ? Avoiding exchanges. 1 6 . . . mhB
From here White has several ways to maintain
the initiative: 1 7.ia3 transposes to the line
1 5 .ia3 tLl d6 1 6.ihlt mhB 1 7.h5 in the
note to White's 1 5 th move above; 1 7.if4
and 1 7.a4 are both promising; 6nally 1 7.tLle5
is arguably the most appealing move of
all.
a
b
c
d
e
f
g
h
16 ... ttJxe4!
Of course Karpov takes the opportunity to
remove the dangerous bishop.
17.iWxe4
The alternative is: 1 7.Ei:xe4 ixe5 I B .Ei:xe5
id7! (This is stronger than Karpov's suggestion
of I B . . . b6, after which 1 9.iWg3 mhB 20.ia3
Ei:eB 2 1 .iWf3 gives White an edge in view 2 1 . . .f6
22.iWxaB fxe5 23.iWxa7.) 1 9.iWg3 mhB 20.ia3
Ei:eB 2 1 .iWf3 Now this move does not produce
a double attack, so Black can safely reply
2 1 . . . f6 with no problems.
17....b:e5!
This decision required 6ne judgment, as it
was important to determine that Black can
cope with the threats on the kingside.
18.dxe5
I B .iWxe5 iWd5 leads nowhere for White:
1 9 .iWg3 ( 1 9.iWe2 b6=; 1 9 .iWxd5 exd5 20.ia3
Ei:dB 2 1 .Ei:e7 b6 22.f3 ie6=) 1 9 . . . iWxh5 20.ia3
Ei:eB Black is 6ne, as 2 1 .Ei:e5 can be met safely
by 2 1 . . .iWg6 or 2 1 . . .f5 .
19JWe2
Anand played this move instantly, although
it is far from obvious if it is White's best.
1 9.iWf3 id7! (After 1 9 . . . iWd5 20.iWg3 mh8
2 1 .ia3 Ei:gB 22.Ei:ed l White can annoy the
queen a bit.) 20.iWxb7 iWeB 2 1 .ia3 ic6
22.iWa6 Ei:f7 Black has good compensation for
the pawn.
407
Viswanathan Anand - Anatoly Karp ov
1998
19.exf6 looks like the most logical move. I was
in Lausanne at the time, providing commentary
with Mikhail Gurevich. My former pupil Peter
Leko was assisting Anand, and the next day he
told me they had analysed the exchange on f6
and concluded that it was winning for White.
My co-commentator and I did not share that
assessment, and modern chess engines have
confirr�ed our opinion: 1 9 . . . 'lWxf6 20.i.e3'IWfS
21 .'lWh4 e5 Black is safe as White is unable to
direct his forces against the g7 -pawn.
23 . . . 2"1:£7 24.Wfa5 2"1:b7 25 .2"1:d6'IWxh5 26.2"1:xc6
'IWd 1 t leads to a perpetual check.
24.'lWxa7?
After 24.'lWd6 2"1:£7 2 5 .i.f4 2"1:d7 Black has
a pleasant game but White should not be
worse.
24 . . . 'lWxh5
19 ...i.d7
From now on Karpov will be in his element,
with opposite-coloured bishops and a slightly
better pawn structure.
20J:�dl
a
a
b
c
d
e
f
g
h
20 ...i.b5!
Karpov uses a simple tactical finesse to
improve his bishop.
21.�£3 �e8 22.i.f4
22.1&xb7?!
This greedy move is enough for equality at
best.
22 . . . i.c6 23 .'lWc7
23.1&a6!? 'lWxh5 24.i.xh6! 2"1:ac8 25 .i.e3'IWg4
26.1&f1 is roughly level.
23 ... 2"1:c8
b
c
d
e
f
g
h
2 5 . f3?
25 .2"1:d3 is necessary, although after 25 ... f4
26.f3'IWf5 27.'lWd4 i.d5 Black is in control.
25 . . . i.xf3!
Karpov gives this winning sacrifice in his
analysis.
26.gxf3 'lWxf3
26 . . . 2"1:xc3! may be even simpler, for instance
27.i.f4 'lWxf3 28 .2"1:d7 'lWg4t 29.cj;>h l cj;>h8
and the white king is defenceless.
27.i.b2! ?
This move resists a bit. Karpov mentioned
27.'lWd4 'lWg3t 28.�f1 2"1:xc3 when Black
wins easily.
27 . . . 2"1:c4! 28 .2"1:d4 'lWe3t 29.cj;>f1 f4!
The advance of the f-pawn is too much for
the defence to bear.
22 .. J�c8 23.�d4 �c4
Karpov prepares to exchange rooks in order
to reduce White's attacking chances. He was
already short of time, although the increment
gave him a certain amount of safety.
23 . . . 2"1:£7 was quite playable, and after 24.'lWg3
cj;>h8 25 .2"1:d6 Elfc7 26.1&g6 the position is equal.
The Prime Years
408
In his Chess Informant analysis, Karpov suggests
23 . . . ic6!? intending to transfer the bishop to
d5. During the game he was probably worried
about an exchange sacrifice on d5, but on
closer inspection Black has nothing to fear, for
instance: 24JMi'e2 id5 25 .1hd5? exd5 26.e6
El:c4 27.e7 At this point Karpov gives 27 . . . El:f7
with a long win, but 27 . . . El:f6! wins instantly.
24J�adl?!
It looks like Anand became too ambitious.
White can grab a pawn with 24.El:xc4 ixc4
25 .Wxb7, but after 25 . . . id5 26.Wxa7 Wxh5
27.We3 g5 Black has good compensation as
Karpov pointed out.
24.Wxb7 would have brought the draw closer:
24 . . . El:xc3 (Sacrificing with 24 . . . ic6?! 25 .Wxa7
is unnecessarily risky for Black.) 25 .Wxa7
Wxh5 The position is roughly equal.
Anand decides to reduce the pressure on
the c3-pawn by exchanging rooks rather than
retreating his bishop to a passive position.
25 hc4 26.a3 gc8 27.El:d4 <it>h7
.•.
Karpov makes his habitual improvement
of his king's position, but there was a more
incisive continuation available: 27 . . . id5!
2B .Wd3 (2B .Wh3 Wc7 29.Wd3 [29.El:d3 'lWc4]
29 . . . mhB Stopping White from taking on d5.
30.Wd2 Wxc3 3 1 .ixh6 mh7 32.ie3 'lWxa3
Black's extra pawn should be enough to win
the game.) 2B . . . Wc7 29.a4 cj;JhB 30.Wd2 Wff7
Black's chances are higher than in the game.
28.i.d2
2B.g4?! weakens the kingside, and after
2B . . . b6 29.id2 id5 30.Wf4 b5 Black has
excellent chances, as . . . El:fB is coming.
28 i.d5 29.Wfh3 b5!?
••.
29 . . . b6 was also reasonable, but perhaps
Karpov wanted to provoke his opponent's next
move.
30.a4 bxa4
30 . . . a6 3 1 .axb5 axb5 32.if4 El:aB would
have kept some advantage, but Karpov prefers
to create an outside passed pawn.
3U�xa4
8
a
b
c
d
e
f
g
h
24 Wf7
•.•
Karpov threatens to increase the pressure
against both of White's weak pawns.
24 . . . ic6 was also reasonable, and after 25 .Wh3
id5 26.El: l d3 Black keeps a small advantage
with 26 . . . Wb5 or 26 . . . b6.
25 Jhc4
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
409
Viswanathan Anand - Anatoly Karp ov
1998
3I...B!c4
Karpov is happy to exchange pieces. Without
the rooks White has virtually no swindling
chances, and the passed a-pawn gains in
strength.
at some point, for instance: 36.Wb4 .ixh5
37 . .ic3 Wd5 38 .Wb5 f4 39.f3 Wd l t 40.Wfl
(40.c;t>h2 .ixf3 !-+) 40 . . . Wxfl t 4 1 .c;t>xfl The
endgame is similar to that which occurs in the
game.
32Jhc4 .hc4 33.V;Vh4 .ib5
36 V;Vd3
Black can win a pawn with 33 . . . Wc7, but
he will pay the price as he must either lose
his a-pawn or expose his king to a perpetual
check: 34 . .ie3 .ie2 (34 . . . a6! ?) 3 5 .Wa4 .ixh5 (If
35 . . . Wxe5 36.We8! White threatens a perpetual
as well as taking the a7-pawn.) 36.Wxa7 Wxe5
37 . .id4 We l t 38.c;t>h2 e5 39 .We7! White has
equalized.
36 . . . Wdl t was also promising: 37.c;t>h2 .ic6
38 . .ib4 We2 39 . .ic3 a6 40 . .ib4 Wb5 Black
has excellent winning chances.
.•.
37.V;Vd4
Anand decides to swap queens. Keeping
them would also have been unpleasant for
him. At this point I got the impression he just
wanted the put this game behind him.
34.c4 .ie8 35.c5?
Anand pushes his pawn, but it was more
important to go after the dangerous a7-pawn:
35.Wd8! Wxh5 36.f3 (Also after 36 . .ie3 a6
37.Wd6 White is safe.) 36 . . . Wf7 37 . .ie3 a6
38.Wb6 Black loses his a-pawn and can kiss his
winning chances goodbye.
37 V;Vxd4 38 .hd4
•.•
•
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
38 a5!
•.•
The closer the pawn gets to promotion, the
more dangerous it becomes.
a
b
c
d
e
f
g
h
35 V;Vd7!
39.c6?
36 ..ic3
It looks like Anand lost his nerves and was
unable to think clearly. He probably wanted to
draw the black bishop away from the h5-pawn,
but the c-pawn is an important asset which
should not have been sacrificed so cheaply.
White can try keeping the queens on, but
Black will probably be able to force an exchange
39.f3 would have given some chances to save
..•
Karpov improves his queen rather than
taking the h5-pawn immediately.
410
Th e Prime Years
the game: 39 . . . a4 (39 . . . i.b5 40.i.b2 cj;lg8 4 l .cj;lf2 cj;lf7 42.'it>e3 'it>e8 43 .i.a3 'it>d7 44.c6t cj;lxc6
45 .i.f8 a4 46.'it>d4 i.f1 47.f4 White holds.) 40.i.b2 i.xh5 4 l .c6 i.e8 42.c7 i.d7 Karpov gives
these moves without evaluating the position. The ending certainly looks difficult for White, but
perhaps he can still draw it. I suspect that had Karpov managed to find a clear win for Black, he
would have shown how. . .
39 ...i.xc6 40.£3
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
40 ... f4!
Karpov isolates the h5-pawn. After taking it he will be able to create another passed pawn or
invade.
41..ib2 .ie8 42 ..icl a4!
Taking the f4-pawn will incur a heavy price, as it allows the a-pawn to get to the second rank.
43.i.xf4 a3 44..ie3 .ixh5 45.c;t; fl .ie8 46 ..id4 .ic6 47..ic3 a2
8
7
6
5
4
3
2
1
a
48.g3
b
c
d
e
f
g
h
1 998
Viswanathan Anand - Anatoly Karp ov
41 1
Waiting passively would not have hel ped. Had White left his pawns on g2 and f3, Black could
have won by putting his pawns on g5 and h4, fixing the white pawns on light squares, then
walking his king to the queenside.
48 ... h5 49.g4 h4
0-1
Having renewed a one-point lead, Karpov played solidly and got a slight edge against the Semi­
Slav. In the queenless position he spurned a repetition more than once, but never got a serious
advantage and Anand held the draw without too many problems. In the sixth and final game
Anand had to win with the white pieces. He tried the Trompowsky and a complex middlegame
ensued. The p ressure of the final game may have played a role, as Karpov blundered badly and
had to give up a piece. He got some pawns for it but was clearly lost and Anand converted his
advantage to force a tie-break.
The title was decided by two rapid games. In the first of them Anand wonderfully outplayed
Karpov with the black pieces and won a pawn. Then in a lost position Karpov used up most of
his time. Peter Leko actually thought he did it intentionally in order to make Anand more tense.
If it really was a deliberate ploy, it proved to be a psychological masterstroke as Anand proceeded
to misplay the endgame and squander the win. Perhaps frustrated, he then made some further
mistakes and ended up losing, which would have been unthinkable from the position he had at
the start of the endgame.
In the second rapid game Anand tried a variation on the Trompowsky with 1 .d4 d5 2.�g5 ,
but got no advantage. Later in the opening he inexplicably blundered two pawns, then sacrificed
a piece as his position became desperate. Karpov calmly took the material and won easily, thus
retaining his FlOE title. It seems that Anand's nerves were part of his undoing, although tiredness
must also have played a role.
***
Karpov's first tournament of the year was Wijk aan Zee. He drew his first six games against
Nijboer, Adams, Gelfand, Shirov, Timman and Van Wely. None of them were short draws, but
all were roughly balanced throughout, apart from the Shirov game, in which Karpov got into
trouble on the white side of a Fianchetto King's Indian and only survived after some inaccuracies
from his opponent.
In Round 7 Karpov suffered a setback against Salov, who played the 4.g3 .ia6 5.tt:lbd2 line
against the Queen's Indian and soon got an advantage, which he converted with the help of
some powerful middlegame play. In the next round Karpov's troubles almost doubled as he was
outplayed by Anand, but the Indian grandmaster missed a couple of wins and Karpov narrowly
managed to draw. Karpov was not so lucky in Round 9 against Topalov, and the Bulgarian
defeated him with a powerful attacking display.
Just as it seemed his tournament might turn into a disaster, Karpov showed his fighting spirit
and bounced back with wins over Judit Polgar and Van der Sterren, although it must be said that
both of them played somewhat below their usual level. The latter game is referenced on page 400
of the first volume.
412
Th e Prime Years
Karpov finished with quick draws against Piket and Kramnik. His final score of 6Y2/ 1 3 was not
bad, considering that he had been on 'minus two' and it took him until Round 1 0 to win a game.
Still, the two-point margin between Karpov and the joint winners Anand and Kramnik showed
that times had changed.
***
Karpov took part in various blitz and rapidplay events, but his next classical tournament was the
Rubinstein Memorial in the Polish city of Polanica Zdroj . In the first round he suffered a setback
against Michal Krasenkow, who played well and eventually won a pawn-up endgame with rook
and opposite-coloured bishops. He followed with a quick draw with Gelfand, then he got an
advantage against Shirov but missed a win and drew.
In Round 4 Karpov got back to fifty percent after outplaying Macieja on the black side of
a 4 . . .if5 Caro-Kann. He followed with three relatively uneventful draws against Markowski,
all and Leko. In Round 8 Karpov switched to the 4 . . . tt:ld7 Caro-Kann against Rublevsky, but
suffered a painful defeat after some powerful play from his opponent. In the last round Karpov
faced Ivanchuk, and after some simplifications the players agreed a draw in a balanced endgame.
Karpov finished with a disappointing score of 4/9, which left him in equal seventh place.
1 998 was a year of mixed fortunes for Karpov. He started the year by defeating Anand for the
highest title in chess, although he did have a huge advantage in being seeded directly to the final.
His tournament results were not great, and like the previous year he took part in more rapid and
simultaneous events then classical tournaments. However, it is worth mentioning his victory
in the very strong Cap d'Agde rapid tournament, where he defeated Boris Gelfand in the final.
Karpov's best years may have been behind him, but he remained a formidable player.
1998 Summary
FIDE World Championship match versus Anand, Lausanne: Drew 3-3 (+2 =2 -2) (Won rapid
tie-break 2-0)
Wijk aan Zee (6th- 1 0th place) : 6Y2/ 1 3 (+2 =9 -2)
Polanica Zdroj (7th-8th place) : 4/9 (+ 1 =6 -2)
Total 48.2% (+5 = 1 7 -6)
II Wins • Draws • Losses
1 999
Rating 27 1 0 ( 1 0 in the world)
Karpov started the year with an eight-game match against Jeroen Piket in Monte Carlo. In the
first game Karpov played the Queen's Indian and held a marginally worse position for a draw.
In the second Piket tried the Vienna Variation against the Queen's Gambit. Karpov sacrificed a
pawn for promising compensation and later missed a win and allowed the Dutchman to escape
with a draw. In Game 3 Piket switched to the 4.Wc2 Nimzo-Indian and got some advantage, but
Karpov managed to hold. Game 4 was a Catalan; Karpov got no real advantage and the game
naturally ended in a draw.
In Game 5 Karpov equalized in the 4.Wc2 Nimzo-Indian and the game remained balanced until
the end. In the next game the players repeated the same opening with reversed colours. Karpov
got a slightly better endgame but Piket defended well and drew. In Game 7 Piket opted for the
Petrosian Variation against the Queen's Indian and sharpened the position by castling long. The
players must have become desperately short of time, as at one point Piket left his rook en prise but
Karpov overlooked it. Shortly after, Piket missed a win and the game finally ended peacefully. In
the eighth and final game Karpov played solidly with 4.e3 and 5 . lt:l ge2 against the Nimzo-Indian.
Piket equalized easily but then played too optimistically and lost a pawn. But Karpov was short
of time and he took a move repetition. Overall the 4-4 result was a fair reflection of the play,
although it was disappointing that none of the eight games ended in a decisive result.
***
Karpov took part in the Amber rapid/blind event, but his first classical tournament was Dos
Hermanas, where he started well, drawing with both Kramnik and Topalov with the black
pieces and then outplaying Svidler from a slightly better endgame. In the fourth round he drew
comfortably against Anand with the Caro-Kann, then drew with Korchnoi in a game that was
always roughly level.
In Round 6 Karpov suffered a setback against Illescas. He failed to equalize in the Caro­
Kann and sacrificed a piece for three pawns, but did not quite get enough play and the Spanish
grandmaster converted his advantage. In the final three rounds Karpov drew against Gelfand,
Judit Polgar and Adams. All three were fighting games which remained roughly balanced until
the end. Karpov's final score of 4V2/9 was not a bad result against a strong field, and enabled him
to share fifth place.
Karpov's next classical tournament was Dortmund, where he faced Ivan Sokolov in the first round.
The Yugoslav-Dutch grandmaster has played forty seven games against the world champions,
scoring nine wins, nineteen draws and nineteen losses. Previously he had lost one game and
Th e Prime Years
414
drawn another against Karpov. The players
met in two subsequent games, both of which
ended in draws.
I Game �ll
Anatoly Karpov - Ivan Sokolov
Dortmund 1 999
l.d4 lLlf6 2.c4 g6 3.g3 .ig7 4..ig2 0-0
Sokolov is a Griinfeld expert, but he seldom
puts his pawn on d5 against the Fianchetto set­
up.
5.lLlc3 lLlc6 6.lLlf3 d6 7.0-0 �b8
7 . . . a6 is more popular. Some experts believe
the rook move to be more accurate, although
in most cases the two moves will transpose.
8.b3 a6 9.lLld5
According to the database this move was
introduced by Horowitz in 1 972, but was
subsequently ignored until Romanishin
took it up in the mid-eighties. By the time
of the present game it had become more
established, and nowadays it remains one of
the hot battlegrounds of the Fianchetto King's
Indian, with several hundred games on the
database.
9 lLle4
8
7
6
5
4
3
2
1 ������
a
b
c
d
e
f
g
h
1 1.e3
Karpov bolsters the d4-pawn and prevents
. . .f4.
1 1 ...e6
So far no other move has been tried here.
1 1 . . .e5 is possible, although after 1 2.�c2 �e8
1 3.!!ad l White has a slight initiative in the
centre.
12.lLlc3 lLlxc3
1 2 . . . b5!? is interesting.
13.i.xc3 fie7
Sokolov clears the d8-square for his knight.
Black could have gained space in a few
different ways, but any pawn advance would
have created some kind of target.
•.•
Sokolov plays a slightly unusual move. The
main line is 9 . . . tt:lh5 , which decentralizes the
knight but allows the black e-pawn to advance
to e4. Karpov faced this move at the Amber
rapid tournament earlier in the year, and
managed to overcome Shirov.
10 ..ib2 f5
Sokolov opts for the Dutch pawn structure,
which usually means that Black will postpone
the development of his queens ide until after he
has gained space in the centre.
1 3 . . . b5 1 4 .!!c 1 bxc4 1 5 .bxc4 �e7 White
maintains a slight plus after 1 6.!!b l or
1 6.d5 .
1 3 . . . e5 1 4.dxe5 ( l 4.d5!? Humans rarely
like to make this pawn advance against the
Leningrad Dutch, but computer programs
seem to like doing it, and usually evaluate it as
slightly better for White.) 1 4 . . . dxe5 1 5 .1Wd5t
Wh8 1 6 .�c5 ! White has a small initiative
in the centre thanks to his well-placed
queen.
1999
Anatoly Karp ov - Ivan Sokolov
41 5
18 ....ib7
a
b
c
d
e
f
g
h
I S . . . e5!?
This move is playable but risky, as it gives
White a target to attack.
1 9 .dxe5 dxe5 20.'lWb2 2"i:dS
After 20 . . . 2"i:eS 2 1 .a4 a5 22.2"i:ae l White has
a small edge.
2 1 .tt:lf3 axb5 22.cxb5 .ib7 23.tt:lxe5 tt:lxe5
24 . .ixe5 .ixe5 2 5 . fxe5 .ixg2 26.'it>xg2 2"i:d3
The disappearance of the minor pieces eases
the defence, but Black's problems are not yet
over.
14.b4!
Karpov starts to gain space.
14 ... lLld8
If 1 4 . . . b5 1 5 .'1Wb3 .ib7 1 6.a4 tt:ldS 1 7.2"i:fc l
Black comes under pressure o n the c-file.
2
15.b5!? lLlf7
After 1 5 . . . axb5 1 6.cxb5 c6 1 7.a4 .id7
lS.'lWb3 tt:lf7 1 9.2"i:fc l 2"i:fcS 20.tt:ld2 the
position is close to equal, but White's side is
more comfortable.
16.�b3 b6 17 . .!iJdl <i>h8
Sokolov prepares to advance his e-pawn. The
immediate 1 7 . . . e5 was playable, although after
lS.dxe5 Black is forced to play I S . . . tt:lxe5.
(Instead IS ... dxe5? 19 . .ib4 wins material as
19 . . . tt:ld6? loses to 20.c5t.) Nevertheless his
position after the knight recapture is quite
reasonable.
Other ideas include 1 7 . . . axb5 I S .cxb5 .ib7,
and 17 . . . .ib7 I S .a4 .ixg2 1 9 . <;t>xg2 a5 20.2"i:ac l
e5 2 1 .c5, both o f which are just fractionally
better for White.
18.f4!?
Karpov gains space and though he does not
stop . . . e5 altogether, he makes it less attractive
to carry out.
d
e
f
g
h
27.2"i:fd l !
This is the most testing.
27.e4 is playable, but after 27 . . . 2"i:bdS 2S.e6t
<;t>gS 29 .'lWe5 2"i:d2t 30.'it>gl 'lWa3! White
must bale out with a draw: 3 1 .2"i:f2! 2"i:xf2
32.e7! 2"i:eS 33 .'lWd5 t=
27 . . . 2"i:xe3
After 27 . . . 2"i:bdS 2S.2"i:xd3 2"i:xd3 29.e6t <;t>gS
30.'lWe5 White keeps some advantage.
2S.e6t 'lWg7 29.'lWxg7t <;t>xg7 30.2"i:ac l
White keeps some initiative although Black
should be able to draw.
19.a4
With this pawn advance Karpov prepares to
exchange on b7 followed by a quick bxa6 and
a5 to open the queenside.
19 .. i.xg2 20.<i>xg2 ga8
.
Sokolov reacts to the threat.
Th e Prime Years
416
;�
��
�
8.i�
7 ����BIt.f
r.,,3.,�.t�
5�lSm _,_
�������_
!�'h1""'�b'J""%�r�
3 � i§� r[j r[j
2 �m ""'�m""%� ��
�.�r.
6
4
1
a
b
c
d
e
f
g
h
The natural 22.e4! may well have been stronger:
22 . . . ttJ h6 (22 . . .l"i:fe8 23.d5 axb5 24.axb5 exd5
25 .exd5 Whi te has some advantage as his knigh t
has better prospects than its counterpart, for
instance it can aim for the e6-square.) 23.l"i:e2
(23 .d5!?) 23 . . . fxe4 24.ttJxe4 ttJ f5 Black's knight
emerges on a decent square, but White keeps
some advantage after 25 .ttJg5 or 25.d5.
Black could also have considered this waiting
move. The idea is that in certain positions
the king will not find itself in check after a
bishop exchange on c3 .
23.l"i:e2
23.c5? would be a mistake due to 23 . . . axb5
24.axb5 dxc5 2 5 . dxc5 �xc3 and since there
is no check, White loses the b5-pawn.
23.e4 fxe4 24.l"i:xe4 d5 25 .l"i:e2 dxc4 26.Wi'xc4
axb5 27.axb5 ttJ d6 28.�c6 �xc6 29.bxc6
l"i:fe8 30.l"i:fe l ttJ b 5 3 1 .�b2 l"i:a2 Black has
good counterplay.
23 . . . l"i:a7!?
23 ... d5?! is premature due to 24.cxd5 Wi'xd5
25 .�xd5 exd5 26.�b4! and White is a bit
better.
Black can also consider 23 . . . l"i:fc8 intending
. .. c6 or . . . c5 . White's best reaction looks to
be 24.l"i:c l .
24.l"i:c l
Most other moves would be met in the same
way.
24 . . . l"i:fa8
The potential counterplay on the a-file will
keep White busy, so Black is only slightly
worse.
22 ... axb5
23.axb5 ga7
a
b
c
d
e
f
g
h
2U�ae1!
Karpov switches t o the centre; now the
e4-push is constantly in the air, Instead after
2 1 .bxa6 l"i:xa6 22.a5 bxa5 23 .�xa5 l"i:fa8 Black
is in time to contest the a-file.
21...�d7
Sokolov removes his queen from the e-file,
presumably after deciding that lines such as
2 1 . . .e5 22.dxe5 dxe5 23.e4 were not to his
liking.
22.tLlf3!?
Karpov decides to prevent . . . e5 and postpone
any direct action in the centre.
Sokolov is a typical player who likes to go
forwards, and here he decides to play on the
a-file.
22 . . . 'tt> g8!?
After 23 . . . d5 Karpov mentions 24.c5
without any further comment, presumably
implying that the position is good for White.
Nevertheless after 24 . . . l"i:fb8 25 .c6 �e8 it will
not be easy for White to achieve anything.
1 999
Anatoly Karp ov - Ivan Sokolov
Perhaps a better reply would be 24.lDe5!?,
for instance: 24 ... lDxe5 25.dxe5 dxc4 26.lMfxc4
White will exert pressure against the c7-pawn,
and if Black exchanges it with . . . c6 then the
b6-pawn will be weak.
24.gal
Karpov does not intend to contest the a-file,
but he exchanges one pair of rooks in order to
prevent Black from having doubled rooks on
the open file.
24 gfa8 25.gxa7 gxa7
•••
8
7
6
5
4
3
2
417
closed.) 27 . . . lDxe5 28.dxe5 dxc4 29.lMfxc4 �f8
30.Eld2 White keeps the advantage, but it is
hard to say if it is truly dangerous or mostly
symbolic.
Maybe Black's most promising idea would have
been to improve his pieces without committing
himself to the . . . d5 push. For example:
26 . . . Wg8!?
The king sidesteps any potential checks on
the long diagonal .
27.Elc2 �f8 ! ?
Black prepares . . . d5 followed b y a possible
rook invasion on a3 .
28.d5!?
White prevents his opponent's idea and
hopes to secure the e6-square for his knight.
A quiet move like 28.Wf2?! would
allow Black to take over the initiative:
28 . . . d5! (28 . . . �e7 29.d5!) 29.cxd5 (29.lDe5 ?
lDxe5 30.dxe5 Ela3-+) 29 . . . lMfxd5 Black is
somewhat better.
8
7
a
b
c
d
e
f
g
h
26.gfl
Karpov guards against checks on the second
rank, in preparation for opening the queenside
with c5 .
26 WI'e8
•••
Sokolov prepares to double his heavy pieces
on the a-file. The text move is not a mistake in
itself, but the overall plan is faulty.
Black can try to be active in other ways as
well, for instance: 26 . . . d5!? 27.lDe5 (27.c5
Iifg8! [27 . . . bxc5? 28.dxc5 Black has no time
to win the b-pawn as the capture on c3
will come with check.] 28 .:1'k2 �f8 White
has a slight edge, but it is not easy to do
something with it as the position is rather
6
5
4
3
2
a
b
c
d
e
f
g
h
28 . . . e5!
28 . . . lMfe7 stops the knight from coming to d4,
but still does not equalize: 29.Ele2! (29.Wf2
e5) 29 . . . e5 30.lMfc2! Black has problems after
30 . . . e4 3 1 .lDd4 or 30 . . . �g7 3 1 .e4.
29.fXe5
After 29.Ele2 �g7 30.e4 (30.lMfc2 Ela3)
30 . . . fXe4 3 1 .Elxe4 IMff5 32.lMfb 1 Ela2t! Black
obtains unpleasant counterplay.
29 . . . dxe5 30.lDxe5IMfe8! 3 1 .lDxf7 Ela3!?
418
The Prime Years
The simple 3 1 . . . 'it>xf7 is also fine. In both
cases Black has enough counterplay to maintain
the balance.
7
27.c5 d5
5
Once
again
considering.
27 . . . 'it>g8!?
was
worth
6
4
3
2
28J�c2
Karpov places the rook on the file along
which he hopes to invade.
a
8
b
c
d
e
f
g
h
28 ...'Wfa8?
Sokolov makes a careless move and
relinquishes the pressure on the b5-pawn.
Black would have been much better off playing
28 . . . i.f6 or 2S . . . i.fS , both of which improve
his bishop slightly while eliminating the
problem of his weak back rank.
2S . . . 'it>gS!?
This move appears tactically dubious, but is
in fact playable.
29.cxb6
Obviously this and the next move are not
forced, but they clearly represent the critical
test of Black's last move.
29 . . . cxb6 30.i.b4
a
b
c
d
e
f
g
h
30 . . . 'lMrxb5!!
This move appears suicidal, but Black can
get away with it.
30 . . . ttJ dS? 3 1 .i.d6 Eia5 (3 1 . . .'lMrd7 32.ib8!
Eia5 33.ttJe5+-) 32.'lMrc3 White has excellent
winning chances in view of the imminent
invasion on c7.
3 1 .EicSt i.fS
Black can afford to give up this bishop as he
has counterplay against the white king.
32.i.xfS!
By offering a queen sacrifice, White causes
maximal problems for the defence.
After 32.EixfSt 'it>g7 33 .Eixf7t 'it>xf7 34.ttJe5t
'it>gS! 3 5 .'lMrb2 (3 5 . ttJ d3 'lMrc4) 3 5 ... 'lMra4 Black
will obtain counterplay on the first rank.
White can also postpone capturing the
bishop:
32.'lMrb2 Eia4 33.EixfSt 'it>g7 34.Eixf7t 'it>xf7
3 5 .ttJe5t 'it>eS 36.'lMrc2 Eixb4 37.'lMrcSt The
game ends with perpetual check.
32.'it>f2 Eia4 33.i.xfS 'lMrxb3 34.i.h6t ttJ d8
3 5 .EixdSt 'it>f7 36.ttJe5t 'it>f6 Once again
White must give a perpetual, with either
37.i.g5 t 'it>g7 3S .Eid7t or 37.ttJd7t rj;;e7
3S.i.g5 t.
32 . . . 'lMre2t
Black had better refuse the offer.
32 . . . 'lMrxb3? 33 .i.h6t ttJ dS
34.Eixd8t
'it>f7 3 5 .ttJe5t 'it>f6 36.EifSt 'it>e7 37.Eif7t
'it>dS 3S .Eixa7 Black avoids an immediate
checkmate, nevertheless he is in trouble;
there is no perpetual, and White retains
Anatoly Karpov - Ivan Sokolov
1999
419
serious attacking chances and more than
enough material for the queen.
33.mh3
In this position Black has a few interesting
ways to secure a draw.
8
7
6
5
a
4
3
2
a
b
c
d
e
f
g
h
33 . . . h6!
The tempting 33 . . . Wfl t? 34.'it>h4 h6 meets
with a beautiful refutation: 35 . .txh6t
(3 5 . e4?! is spectacular but does not quite
work: 35 . . . fXe4 [35 . . . dxe4? 36.'1Wxe6+-l
36 . .txh6t mh7 37.tLlg5t tLlxg5 3S . .txg5
'lWe2 39.g4 e3! 40.Wc2 We I t 4 1 .mh3 Wfl t
42.mg3 Ei:g7! After a series of only moves,
Black holds.) 35 . . . mh7 36.tLlg5t tLlxg5
(36 . . . mxh6 37.Wb4!+-) 37.mxg5 Wh3
b
c
d
e
f
g
h
3S.Ei:hSt!! mxhS 39.Wxb6 Black can resist
for a bit longer, but the position is losing for
him.
However, there was an alternative drawing
line in which the black king survives a near­
death experience: 33 . . . tLl g5t!
c
d
e
f
g
h
34.fXg5 (34.tLlxg5 Wg4t is a trivial draw.)
34 . . . Wfl t! 3 5 .mh4 Wxf3 36 . .th6t mfl
37.Ei:fSt me7 Amazingly White does not
have more than a perpetual, for instance:
3S .Wb4t md7 39.Wb5t me7! 40.WeSt
(40.Ei:eSt mfl=) 40 . . . md6 4 1 .WbSt mc6
42.Ei:cSt mb5 43.Ei:c5t ma6=
a
a
b
b
c
d
e
f
g
h
34.e4?
This attempt to avoid a perpetual leads to
disaster for White.
Instead White should settle for a draw with
something like: 34 . .txh6t (34.tLl e 1 Wg4t=)
34 ... mh7 3 5 . tLl g5 t tLlxg5t (or 35 . . . mxh6
36.tLlxflt Ei:xfl=) 36 . .txg5 Wg4t=
34 . . . fXe4
After 34 . . . dxe4?? 3 5 .Wxe6 Black's position
collapses.
35 . .txh6t
3 5 . tLl e 1 mh7 36 . .tb4 Wfl t 37.tLlg2 h5
3S.Ei:c2 tLlg5t wins.
35 . . . mh7 36.tLlg5t mxh6
The Prime Years
420
8
Sokolov takes away the c6-square from the
rook.
6
31.Wc3 E!:al
4
After 3 1 . . .i>g8 32.Wc8 if6 33.ttJe5 Black's
position falls apart.
7
5
3
32 ..ie7
2
a
b
c
d
e
f
g
32.Wc7! would have won even more quickly:
32 . . . E!:xc2t 33 .Wxc2 Wb7 34.id6 <;f;>g8 35 .Wfc7
And the b6-pawn falls.
h
37.ttJxf7t
Had the analogous position been reached
with the moves . . . wn t and <;f;>h4 already
inserted, White could have won with 37.Wb4,
but in the present position this move is easily
is refuted by 37 . . . Wh5 t 38.i>g2 E!:a2t. Such
nuances make the game of chess endlessly
fascinating.
37 . . . E!:xf7
Black is winning.
32 tiJf7
•.•
29.cxb6! cxb6
a
b
c
d
e
f
g
h
33.E!:xal!
Karpov finely j udges that his opponent's
active queen will not be able to do any damage
by herself.
33 ...\Wxalt 34.i>f1 \Wbl t
34 . . . Wc4t? 3 5 .Wxc4 dxc4 36.i>e2 is an easy
endgame for White.
a
b
c
d
e
f
g
h
Karpov exploits the fact that the b 5-pawn is
no longer attacked, by taking the opportunity
to improve his bishop while also opening the
c-file.
34 . . . h5 was more challenging, although Black is
still in trouble: 35 .Wc6 ttJh6 (35 . . . <;f;>h7 36.h3
ih8 37.ttJ gl ig7 38 .Wxb6 White wins a pawn
and the game.) 36.h3 Wb3 37.<;f;>e2 ttJg8 38.td8
Wb2t 39.ttJd2 if8 40.We8 <;f;>g7 4 1 .Wd7t <Jih8
42.ic7 Black is unlikely to survive.
30 tiJd8
35.i>f2 h6
30.i.b4!
•.•
Darmen Sadvakasov - Anatoly Karpov
1999
35 . . . Wa2t also fails to save the game: 36.ttJd2
h6 37.'kt>e2 'kt>h7 3S .Wc6 Wa5 39 .WeS ttJhS
40.�dS White wins as the b6-pawn falls, as
pointed out by Karpov in his Chess Informant
analysis.
36.We6 Wd3
After 36 . . . Wb2t 37.'kt>fl (37.'kt>e l 'kt>h7
3S.ttJ d2 e5 39.Wxd5 exd4 40.exd4 also wins)
37 . . . 'kt>h7 3S.ttJgl ! Wxh2 39 .Wxe6 White wins.
42 1
Sokolov resigned without waiting for the
knight to fall.
Karpov continued with three sharply-contested
draws against Kramnik, Topalov and Adams.
He finished the tournament with three less
exciting draws against Anand, Leko and
Timman. His final score of 4/7 was enough to
share third place with Anand and Adams.
Karpov's last tournament of the decade, and
indeed the millennium, was a four-player,
double-round-robin event in the Dutch
town of Hoogeveen. In the first round he
met an opponent whom he had never faced
before. Darmen Sadvakasov has played eight
games against the "three Ks" amongst world
champions: Karpov, Kasparov and Kramnik.
He has scored two wins, three draws and three
losses. In his three subsequent encounters with
Karpov he scored two wins and a draw.
a
b
c
d
e
f
g
h
37.liJgl!
I Game �ll
Darmen Sadvakasov - Anatoly Karpov
Karpov elegantly stops Black's counterplay
to seal his victory.
Taking the knight was also winning, albeit in
a slightly less straightforward way: 37.WeSt
<;t>h7 3S .Wxf7 Wc2t 39.'kt?e l We I t 40.'kt>e2
�c4t 4 1 .'kt?d2 Wa2t (4 1 . .. Wfl 42.�fS Wf2t
43.'kt?c3 Wxe3t 44.'kt>b4+-) 42.'kt?e l Wb l t
43 .'kt?e2 Wxb5 t 44.'kt?f2 Wb2t 45 .'kt?g l We I t
46.'kt?g2 White escapes the checks and wins.
37 ....bd4
A desperate try, but there was nothing else.
38.We8t
White wins a piece and the game.
38 ... �g7 39.Wffit �h7
1-0
Hoogeveen 1 999
l .e4 e6 2.d4 d5 3.ti:Jd2
According to the database the Kazakh
grandmaster had previously played the Panov
Variation.
3 ... dxe4 4.liJxe4 liJd7 5.liJg5 liJgf6 6.i.d3 e6
7.liJ 1£3 i.d6 8.We2 h6 9.liJe4 liJxe4 10.Wxe4
We7
Karpov also played 1 O . . . ttJ f6 and 1 O . . . c5, but
the queen move is the one he employed most
frequently.
I l .Wg4 �ffi
Almost all of Karpov's classical games in
this variation ended in draws. Karpov scored
poorly with it in blind and rapid games, but
The Prime Years
422
subsequent analysis must have restored his
confidence in Black's position.
29.Ei:xd8 bxa5 0- 1 Topalov - Karpov, Dubai
(rapid) 2002.
12.0-0
13 ... b6
This is the usual move. In her rapid match
against Karpov in Budapest 1 998, Judit Polgar
scored a win and two draws with 1 2..�e3 .
12 ... c5
Sometimes Black plays 1 2 . . . b6 first, but he
invariably challenges the opponent's centre at
some point
1 3 . . . c4?! is risky: 1 4 .�e2 (The immediate
1 4.�xc4?? loses to 14 . . . �e7! .) 14 . . . b5 1 5 .a4 b4
1 6.�xc4 �b7 1 7.liJd2 liJ f6 1 8 .�d3 �d5 Black
got some compensation for the pawn and
eventually won in Kummerow - Speelman,
Bundesliga 1 999, but not many players have
followed in the creative Englishman's footsteps,
probably because of the strong 1 6. liJ d2! which
questions Black's play.
14..ie4
a
e
f
g
h
13JWh4
It is natural to move the queen off the
diagonal of the enemy bishop. 1 3 .c3 is the
most popular move, and 1 3 .Ei:e l is another
sensible alternative.
In 2002 Karpov twice faced 1 3 .b3 against
Topalov: 1 3 . . . e5 1 4 . dxc5 liJ xc5 1 5 .�f5 h5
1 6.1Wh3 e4! (This was Karpov's improvement
over the first game, which continued 1 6 . . . liJ e6
1 7.Ei:dl �e7 1 8 .1Wg3 when Black failed to
equalize and went on to lose, Topalov - Karpov,
Cannes 2002.) 1 7. liJ d4 �xf5 1 8 .liJxf5 �e5
1 9.�a3 g6 20.liJe3 �g7 Karpov refuses to take
the exchange and simply finishes developing.
2 l .Ei:ae l liJ e6 22.liJc4 liJ g5 23 .1We3? (23 .1Wh4
�f6=) 23 . . . �xh2t 24.�hl �f4 25 .1Wc3t
f6 26.�b2 1Wc6 27.Ei:d l Ei:hd8 28 .1Wa5 b6
a
b
c
d
e
f
g
h
14 J:�b8!?
••
Karpov postpones the bishop exchange and
deviates from 14 . . . �b7, with which he lost a
blindfold game against Lautier in 1 998.
15JMl
This was a new move a t the time.
1 5 .b3 leads to an interesting position in which
Black has a few different options:
a) 1 5 . . . liJ f6? is asking for trouble: 1 6.dxc5 bxc5
1 7.�b2 liJxe4 1 8 .1Wxe4 �b7 1 9.1Wg4 f5 20.�h5
�xf3 2 l .1Wxf3 c;l;f7 22.1Wh5t g6 23.1Wh4 ie7
24.1Wh3 �f6 2 5 .�xf6 �xf6 26.Ei:ae l e5 27.�c3
Ei:he8 28.f4 Ei:b4 29.g4! Ei:d4 30.1Wh3 �g7
Darmen Sadvakasov - Anatoly Karp ov
1999
3 1 .fXe5t �xe5 32.�xe5 i>xe5 The position
deserves a diagram, especially considering that
Karpov actually managed to survive with such
an 'active' king . . .
a
b
c
d
e
f
g
423
�e7 I S .1Mff4 �d6 1 9.1Mfh4 Y2-Y2 Leitao Karpov, Buenos Aires 2000.
c) Apart from the above drawing line Black has
another reasonable move: 1 5 . . . c4!? 1 6.bxc4
�a6 1 7. tLl d2 tLl f6 l S .�d3 �f4 ( 1 S . . . g5 !?
1 9.1Mfh3 It>g7 is also interesting) 1 9. tLl f3 �xc 1
20.�axc 1 �xc4 2 1 .�xc4 1Mfxc4 Black appears
to be dangerously behind in development,
but he held the draw with ease: 22.tLle5 1Mfc3
23 .Wff4 �cS 24.�fd l i>gS 25 .h4 b5 26.d5
exd5 27.�xd5 tLlxd5 2S.1Mfxf7t Y2-YZ Milos Vescovi, Sao Paulo 2000.
h
33.1Mfg3t f4 34.1Mfe l t It>d6 3 5 .c3 �d3 36.1Mfe4
Wlxc3 37.1Mfxg6t i>c7 3S .1Mff7t i>b6 39.1Mfxf4
a
a
b
c
d
b
c
d
e
f
g
h
15 ... c4!?
e
f
g
h
39 . . . lt>a5! The way Karpov uses his king
is remarkable. Objectively his position is
probably still losing but he manages to
create enough obstacles for his opponent.
40.1Mfc7t (40.lt>hl ! ?) 40 . . . lt>b4 4 1 .1Mfb7t It>a3
42.1Mfxa7t (42.1Mfa6t It>b2 43.1Mfxh6 It>xa2
Black gets counterplay on the queenside.)
42 . . . It>b2 Black's position is still difficult but
he eventually managed to hold it, Timman Karpov, Bali 2000. By the end of the game his
king had returned to a5 again!
b) The next time he reached this position
Karpov improved with a much safer
continuation: 1 5 . . . �e7! 1 6.1Mff4 �d6 1 7.1Mfh4
Karpov p revents his opponent from opening
the d-file.
16.tLle5!?
White should aim to open the position as
quickly as possible, as Black enjoys good long­
term prospects thanks to his control over the
d5-square.
The alternative is 1 6.d5 e5, which has so far
resulted in three draws without a win on either
side.
16 ... �f6 17.i.f3 i.b7 18.,hb7 gxb7 19.b3
Sadvakasov continues trying to open the
position in the hope of exploiting his better
development.
The Prime Years
424
After 1 9 . .if4 tLJd5 20 . .ig3 .ixe5 2 1 ..ixe5
Wd7 22.1"i:e l f6 23 . .ig3 <;t>f7 24J:l:e2 b5 Black
obtained a pleasant position in Odeev Sargissian, Istanbul (01) 2000.
is doing well after 22.Wf3 We4 or 22 . .ie3
�h7.) 2 1 . . .Wc2! 22.1"i:ac 1 Wf5 Black should be
fine.
21 ..if4 Wfe7
This tempo-gaining move is an important
part of Karpov's plan.
22.Wfg3
a
b
c
d
e
f
g
h
19 cxb3
•••
Black opens a file, but in return he wins
time to catch up on development. If he can
organize his position successfully then he will
have excellent prospects against the isolated
pawn.
So far nobody has tried 1 9 . . . c3 with the idea
of keeping the position closed. Indeed, it looks
risky to allow the white rook to appear on the
third rank, and in the long run the c3-pawn
might become weak.
20.cxb3 tlJd5
Karpov puts his knight on its dream square,
while preparing to win a crucial tempo by
threatening to exchange queens.
The greedy 20 . . . .ixe5? leads swiftly to disaster:
2 1 .dxe5 Wxe5 22 . .if4 Wb2 23.1"i:ac 1 �g8
24.1"i:c8t <;t>h7 25 . .ixh6!+20 . . . �g8!? Trying to improve the king first has
never been tried, but it is interesting. 2 1 ..if4
(2 1 .tLJc4 tLJd5 22 . .id2 .if4 Black is safe;
2 1 .Wg3 is well met by 2 1 . . .Wc2! and Black
a
b
c
d
e
f
g
h
22 ....he5!?
Karpov takes an ambitious and somewhat
risky decision to play for the win. With the
superior minor piece and the d4-pawn as a
target, his long-term chances are excellent.
The problem is that his pieces are temporarily
uncoordinated, which means that before he
can reach his dream endgame, he will have
to suffer and tread carefully to avoid the
immediate threats.
Despite Karpov's eventual success in the present
game, no-one else has been brave enough
to follow in his footsteps. In all three of the
subsequent games that reached this position,
Black preferred 22 . . . tLJxf4 which leads to a
slightly worse but tenable position, and indeed
all three ofthose games were drawn. The first and
most high-profile of them continued: 23.1Wxf4
<;t>g8 (23 . . . g5 ! ? 24.We4 �g7 is interesting but
has not yet been tested.) 24.We4 f5 25.1We3
Darmen Sadvakasov - Anatoly Karp ov
1999
mh7 26.l''1ac 1 .txe5 27.'1Wxe5 E:dS 2S .E:c6 E:d5
29 .Wxe6 Wxe6 30.E:xe6 E:bd7 Black regained
the pawn and easily held the draw, Timman Seirawan, Bali 2000.
23 .he5 �g5
•
Nothing else is really playable.
24.�d3
Obviously White should keep the queens on
the board.
425
26.�a6
In a sense, the pressure is on White, as he
must try to achieve something before Black
can finish develop ing, otherwise he will face
a difficult defence with his isolated pawn.
Sadvakasov's move is a good one, but there
were other possibilities.
Doubling the rooks on the c-file does not
achieve much: 26.E:c4 f6 27.E:dc 1 �f7 2S ..tg3
(2S .Wf3 E:eS) 2S . . . WaS 29.f3 ltJ e7 Black is
somewhat better.
24 E:d7 25.E:ac1
•.•
Trying to invade with the queen on the c-file
is not dangerous: 25 .Wc4 �gS 26.Wc6 We7
Black covers his weaknesses and is ready to
finish unravelling his kingside.
8
7
6
5
4
3
2
26.a3!?
a
1
a
b
c
d
e
f
g
h
25 �d8?!
••.
Karpov retreats his queen in order to keep his
queenside together. It is j ust about playable, but
it was not the most straightforward solution.
25 . . . �e7! was simplest. Black can solve his
problems by leaving his queen on the kingside
and quickly bringing the second rook into play.
26.E:e l E:hdS (26 . . . E:eS!?) 27.h4 (27.We4 mfS ;
27.Wh7 ltJ f6) 2 7 . . . Wg4 2S.E:e4 Wg6 Black
achieves full coordination and it is White
who will have to start thinking about
defending.
b
c
d
e
f
g
h
This is a subtle move. The point is that Black
is likely to play . . . f6 sooner or later, so White
wants to be able to attack the e6-pawn with
E:c6 without fearing the reply . . . ltJ b4.
26 . . . ltJ e7
26 . . . f6 27 . .tg3 mf7 (27 . . . ltJ e7 should be met
by 2S.We2!, defending the d l -rook, when
White is fine.) 2S.E:e 1 E:eS 29.E:c6 White is
not worse as the e6-pawn is vulnerable.
If Black tries to play as he does in the game
with 26 . . . �gS , then White replies 27.We4. If
Black wants to organize his position then he
will have to play . . . f6 at some point, which
falls in with White's plan.
27.We4 f6
After 27 . . . mgS 2S.E:c3 ltJ g6 29 . .tg3 White
obtains good counterplay on the queens ide
as the black knight is far away.
The Prime Years
426
28.�f4 e5
28 . . . <j;lf7 is met by 29.Ele l .
29 .�e3 <j;l f7 30.V*'c2 exd4 3 1 .V*'c4 t Eld5
32.Elxd4
The position is equal.
26 ... tiJe7
Black's position looks passive, but for the
moment he is doing enough to cover all the
important squares.
8
7
6
5
4
3
2
30.�g3 V*'xc4 3 1 .bxc4 ttJ e3 32.Eld3 tiJf5
33 .�fL. The position is equal.
29.Wfa4 Wfe8
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
30.i.b8?
This is a serious mistake which gifts Black
two free tempos with which to finish his
development.
1
a
b
c
d
e
f
g
h
27.Wfc4
With a lead in development and a bishop
versus a knight, White should have taken
the opportunity to open the position: 27.d5!
f6 28 .�b2 exd5 29.V*'e2 White has excellent
compensation, for instance: 29 . . . Wf7 30J'1e l
Ele8 3 1 .V*'e6t <j;lf8 32.�xf6 Eld6 33 .�xe7t
Elxe7 34.V*'f5 t Black should be able to hold
this inferior endgame, but he will have to be
careful.
30.V*'xa7 This is the most straightforward
continuation, although j ust about any other
reasonable move would have given White a
playable position as well. 30 . . . ttJd5 3 1 .iWa4!
V*'e7 32.�e5 (Another idea is 32.�g3 :ga7
33 .V*'b5 Elxa2 34.Ele l V*'f6 3 5 .V*'d3! when the
generally desirable . . . g6 will weaken the dark
squares, which should give White adequate
play.) 32 . . . Ela7 33.V*'c6 (or 33.V*'c4 :gxa2
34.V*'d3) 33 . . . Elxa2 34.Ela l Elxa l 3 5 .Elxa l Wh7
36.V*'c2t f5 According to Karpov the position
is equal.
27... <;t>g8
Karpov prepares to shuffle his king and
rook.
28.i.c7 Wfa8
Karpov decides to keep the queens on the
board for the time being. It is not without risk,
but exchanging them would have improved
White's pawn structure: 28 . . . V*'c8 29.f3 ttJ d5
30 .. .'kt>h7 31 .i.c7
Sadvakasov must only now have realized
that 3 1 .�xa7? was unplayable due to 3 1 . . .:gb7!
32.V*'a6 V*'a8 33 .V*'d3t g6 when Black wins a
piece.
31 ...tiJd5 32.i.e5 Wfe7
Karpov prepares to develop his h8-rook.
Darmen Sadvakasov - Anatoly Karp ov
1999
33.\Wc4 \Wg5
a
b
c
d
f
e
g
h
34.\Wd3t
This queen check does not harm White, but
nor does it improve his position in any way.
More constructive would have been:
34.a4!
With this move white enables his rook to
utilize the c4-square. Furthermore he may
be glad for the opportunity to exchange
the queens ide pawns in order to escape to
a tenable endgame with three pawns against
four on the kingside.
34 . . . �hdB
After 34 . . . a5 ?! 3 5 .�e2 �hdB 36.�c6 White
obtains counterplay.
35 .�e2 CiJe7 36.�c4 CiJf5 37.h3
427
The direct 37 . . . f6 is critical, but does not
quite work: 3B . .ic7! �xd4 39.�dxd4!
(39 . .ixdB? ? CiJh4! 40.g3 �d5-+) 39 . . . �xd4
40.�xe61he position is equal.
3B .g4!?
White should try to disrupt his opponent's
pieces.
After 3B .'i!fh2 f6 39 . .ic7 �Bd7 40.�e4 �g6
41 . .ibB e5 Black wins the d-pawn.
3B . . . f6 39 . .ic7 �Bd7 40.�e4 �g6 4 1 .'i!fh2
CiJd6 42.�xg6t <Jixg6 43 ..ixd6 �5xd6
43 . . . �7xd6 is also possible, although after
44.<Jig3 Black must avoid 44 . . . e5 ? in view
of 45.dxe5 , when the attack on the d6-rook
enables White to ignore the pin. Instead
Black can keep some winning chances with
44 . . . a6! intending . . . b 5 , when White must
either lose the d-pawn or risk his rook
becoming locked in a cage on b4.
44.g3 e5
The d-pawn falls, and Black has excellent
winning chances.
a
b
c
d
e
f
g
h
34 ...\Wg6 35.\We2
a
b
c
d
e
f
g
h
37 . . . �d5!
Black should crack the defence slowly.
3 5 .�xg6t!?
Going for the endgame is not necessarily
losing as Capablanca also held Flohr in a
similar and famous endgame from Moscow
1 93 5 . Here the situation is slightly different
as all four rooks are still on the board.
Nevertheless it is not easy for Black to break
The Prime Years
428
his opponent's defence. At the same time, it is
understandable that one would be reluctant
to play such a position against an endgame
wizard like Karpov.
35 . . . Wxg6 36.f3 13:hd8 37.Wf2
Another possibility is: 37.13:d2 Cfj e7 (37 . . . f6
38 .i.g3 Wf5 39.13:c4 h5 40.h4 Cfj e7 4 1 .i.c7
13:c8 42.i.g3 White seems to be okay.) 38.13:c4
b5 (38 . . . Cfj f5 39.g4) 39 .13:c7 f6 40.13:xd7 13:xd7
The endgame looks dangerous for White. In
the aforementioned classic game, the Cuban
genius defended by advancing his a-pawn
two squares, but here a4 would be risky as
Black has already played . . . b 5 .
3 7 . . . f6
After 37 . . . Cfj e7 38 .g4 f6 39 .i.c7 13:c8 40.i.g3
White seems to hold.
38 .i.g3 h5 39 .13:d2 Cfje7
If 39 . . . \t>f5 40.13:c4 Cfje7 then 4 1 .i.c7 once
again looks like a good move.
40.13:e2 \t>f5 4 1 .13:ce l 13:xd4 42.13:xe6 Cfj d5
43. \t> n
Black i s clearly better, but not necessarily
winning.
36 ... h5!
Karpov stops the h-pawn from advancing
and fixes it as a weakness.
37J3d3 �g4!?
With little time on the clock, Karpov goes
for safety.
The more direct approach was 37 . . .f6!1,
intending to act in the centre at once. 38 .i.g3
Cfj e7! (Karpov mentions the line 38 . . . e51!
39.dxe5 ! Cfj f4 40.i.xf4 13:xd3 4 1 .exf6 gxf6
42.13:c6 when despite his extra exchange, Black
has no advantage.) 39 .13:dd 1 Cfj f5 40.'lWxe6
Cfjxd4 4 1 .'lWe3 13:e8 42.13:xd4 13:xe3 43.Elxd7
13:e4 44.13:xa7 f5 Black has good winning
chances.
38.�d2?
Sadvakasov gambles by keeping the queens
on the board. The decision is understandable,
but ultimately incorrect as his position goes
from clearly worse to losing outright.
35 .. J3hd8
8
7
6
5
4
3
2
1
a
difficult for White. The most resilient defence
would have been 36.f3 intending to sit and
wait, although this would obviously have been
unpleasant as well.
b
c
d
e
f
g
h
36.h4?
White tries to do something active on the
kingside, but he gets nowhere and the h-pawn
will become weak. 36.13:c4 f6 37.i.g3 e5! is also
White should have exchanged queens:
38.'lWxg4 hxg4 Karpov only evaluated the
endgame as better for Black, which suggests
that he believed White could still hold it.
39.13:c4 (39.\t>h2 f6 40.i.g3 Cfj b4 4 1 .Ele3
Cfjxa2 42.Elc6 Cfj b4 43.13:cxe6 13:xd4 Black
is a healthy pawn up.) 39 . . . \t>g6 (39 . . . Cfje7
40.d5 [40.13:g3 f6] 40 . . . 13:xd5 4 1 .13:xd5 Cfj xd5
[4 1 . . .13:xd5 42.i.b8] 42.Elxg4 f6 43 .i.g3
Black is a bit better.) 40.i.g3 \t>h5 4 1 .f3 f5
42.fxg4 t \t>xg4 Black is certainly better, but
it is hard to tell if his position is technically
winning.
38 f6 39.i.g3
.•.
a
b
c
d
e
429
Darmen Sadvakasov - Anatoly Karp ov
1999
f
g
h
a
b
c
d
e
f
g
h
39 ... .!lJe7
45.Wc2?!
Finally Karpov gets the chance to attack
the d4-pawn. It is his reward for enduring the
unpleasant period that went before.
White commits an inaccuracy. 45 .Wc8
would have given better practical chances,
although the end result should be the same:
45 . . . Wxh4 (45 . . . We5 46.Wc3 :§:dl t 47.'it?f2
Wxc3 4B.:§:xc3 :§:d2t is possible, but White
still has some drawing chances here.) 46.WhBt
<±>g6 47.WeBt <±>h7 4B .We3 :§:d7 Black wins in
a similar way as in the game.
40.We3?
This move loses quickly, but White lacked
a decent alternative, for instance: 40.:§:c4
e5 (40 . . . b5 is less convincing: 4 1 .:§:c5 :§:xd4
42.:§:xd4 :§:xd4 43 .Wc2t ttJ f5 44.:§:xb5 :§:d l t
45.<±>h2 Wd4 46.:§:xf5 exf5 47.'�·xf5t g6
48 .We6) 4 1 .d5 Wg6 42.d6 ttJ f5 43.:§:c6 Wf7
White soon loses the d-pawn.
40 ... gxd4?!
Karpov wins the pawn, but not under the best
possible conditions. Instead 40 . . . e5! is simply
winning. With more time on the clock Karpov
would not have missed such a possibility.
45 ... Wxh4
Karpov collects a pawn and his pieces
remain better placed. Winning the position
still requires some technique, but fortunately
this is not something Karpov lacks.
46.gdl Wf4 47.gd3
41.gxd4 gxd4 42.£3 tiJf5!
Simplifying to a won endgame.
43.Wxe6
Sadvakasov finds the only way to prolong the
game. Less resilient would have been: 43.fxg4
CLlxe3 44.�f2 (44.gxh5 :§:d2) 44 . . . :§:dl t 45.:§:xd l
CLlxdl 46.gxh5 ttJxf2 47.<±>xf2 <±>h6 48.g4 f5
49.<±>f3 fxg4t 50.<±>xg4 e5 Black wins.
a
b
c
d
e
f
g
h
The Prime Years
430
55".Wle5!
47... h4!
Preparing to use the extra pawn aggressively.
48.�hl g6!
Karpov makes more room for his king.
Karpov centralizes his queen while reducing
White's checking options - it is textbook
stuff.
56.Wldl �h4!
Karpov uses his king actively while avoiding
the threat.
49.Wlc3 ihd3 50.Wlxd3
8
57.�f2 Wlh2t 58.�e3 Wle5t?!
Taking the second pawn was a simpler path
to victory: 5 8 . . . Wxa2! 5 9 .Wd4t I!?g3 60.Wg4t
I!?h2 Black wins.
7
6
5
59.�f2 f5 60.a4?
4
Sadvakasov could have offered sterner
resistance by driving the active king backwards:
60.Wh I t! Iifg5 6 1 .Wc l t Wf4 62.Wd 1 Wh2t
63.I!?e3 Iifh6 64.Wa 1 Wd6 65.I!?f2 Black still
has to work for the point.
3
2
1
a
b
c
d
e
f
g
h
50".h3!
This temporary pawn sacrifice opens White's
kingside.
5 1.gxh3 Wlg3 52.Wld2t �h5 53.Wle3 Wlxh3t
54.�gl Wlf5!
Defending the f-pawn and threatening to
win a second pawn by checking on b 1 .
6o".Wlg3t 61.�e3 Wle5t 62.�f2 Wlh2t
63.�e3
8
7
6
5
4
55.Wle2
3
2
8
1
7
a
6
b
c
d
e
f
g
h
63".g5!
5
4
Karpov advances his extra pawn and blocks
one of the potential checking diagonals.
3
64.Wld5
2
1
a
b
c
d
e
f
g
h
After 64.We 1 t I!?h3 65 .Wfl t Wg2 66.'.Wal
f4t 67.I!?e4 Wc2t 68.I!?e5 I!?g3 Black soon
wins the f-pawn and the game.
1999
64
.••
Darmen Sadvakasov - Anatoly Karpov
43 1
�gl t 65.@e2 @g3!
0-1
Black wins the f3-pawn and indirectly defends the f5-pawn, as taking it would allow an
immediate queen exchange.
In the next round against Judit Polgar, Karpov chose the 4.�c2 line against the Nimzo-Indian.
Judit sacrificed two pieces and a thrilling gam e ended in a perpetual check. In Round 3 Karpov
faced Tim man, and found himself on the black side of the same opening, via a slightly unusual
move order. The Dutchman exploited his two-bishop advantage m asterfully, and won a fine game
of which Karpov himself would have been proud.
In the first round of the second half, Karpov played an innocent opening with an early queen
exchange against Sadvakasov. He managed to obtain real winning chances, but the young Kazakh
grandmaster showed his resiliency and held a draw. Karpov's return games against Polgar and
Tim man were both drawn quickly. Karpov's fifty percent score left him in third place behind the
co-winners Tim man and Polgar.
1999 Summary
Match versus Piket, Monaco: Drew H (+0 =8 -0)
Dos Hermanas (5th-6th place) : 4Yzl9 (+ 1 =7 - 1 )
Dortmund Ord-5th place) : 4/7 (+ 1 = 6 -0)
Hoogeveen (3rd place) 3/6 (+ 1 =4 - 1 )
Total 5 1 .7% (+3 = 2 5 -2)
III Wins • Draws • Losses
2000
Rating 2696 ( 1 1 i n the world)
Karpov started the new millennium with a match against Bacrot in Cannes. The two-game
classical part of the match ended in a 1-1 tie after two hard-fought draws in which neither side
was able to get the upper hand. The young French talent won the rapid section 2-0, but Karpov
took revenge by the same score in the blitz.
After taking part in the Amber rapid/blind tournament, Karpov took part in a match against
the reigning Women's World Champion Xie Jun, comprising four classical games and two rapid
ones. In the first game Karpov got a small edge in a Samisch King's Indian, but the Chinese
grandmaster defended well. In the second Karpov got a winning advantage, but let it slip away
and the game ended in a draw. In Game 3 Karpov obtained a clear positional advantage. At
one point he allowed his opponent a chance to get back into the game, but Xie Jun missed her
opportunity and went down without much resistance. The last game was a quick draw, so Karpov
won the match by the score of 2Y2-1 liz. He also won the rapid part of the match, winning the
first game and drawing the second.
Karpov remained in China for another match, this time against the top-rated Chinese male
player, Ye Jiangchuan. This time there were only two classical games. The first was a peaceful
draw in the Caro-Kann. In the second Ye Jiangchuan equalized in a Fianchetto Grunfeld. Karpov
obtained a slight edge in the middlegame but it was not enough to win. Karpov won both rapid
games.
***
Karpov's first classical tournament of the new millennium was the Japfa Classic, on the Indonesian
island of Bali. He started with a nice win over Gunawan, then drew quickly with Seirawan. In the
next two rounds he was pressing against Milos and Judit Polgar, but had to be content with two
draws. Then he scored a good win over Adianto.
In Round 6 Karpov had to suffer against Timman but eventually achieved a draw after some
remarkable defensive play in the endgame, as shown in the note to White's 1 5th move in Game
52 at the end of the previous chapter. In Round 7 Karpov sacrificed a pawn against Ehlvest, and
punished the Estonian grandmaster who was unable to find the right solution to his problems.
In the next game he fought hard to bring down Win Lay Zaw of Myanmar, but eventually had
to settle for a draw. In the ninth and final round Karpov got into trouble against Khalifman, but
eventually managed to hold the draw. His final score of 6/9 was enough to share second place
with Khalifman, half a point behind Polgar.
The Prime Years
434
Karpov's next classical event was the Najdorf
Memorial in Buenos Aires. In the first round he
got into trouble against the much lower rated
Diego Flores, but managed to win after his
opponent spoiled his position. In the second
round Karpov got just a marginal edge against
Milos, but managed to turn it into something
significant and won a nice middlegame with
opposite-coloured bishops. He maintained his
perfect start against the local player Pierrot.
Perhaps Karpov wanted to pay tribute to
Miguel Najdorf, as he used the legendary
Polish-born Argentine grandmaster's patented
Sicilian to win a nice game.
In Round 4 Karpov dropped his first half
point against Judit Polgar, who sacrificed a lot of
material before the wild game ended peacefully.
After a quick draw with Leitao, Karpov seemed
to have good chances against Milov but the
Russian-born Israeli-Swiss grandmaster held a
draw. Karpov's next two games against Short
and Bologan were drawn, although both were
hard-fought. Unfortunately Karpov suffered
a disappointing end to the tournament after
losing on time in a winning position against
Ricardi. He finished in fourth place with a
score of 5 Yz/9.
***
Karpov's final event of the year was the Cap
d'Agde rapid tournament. I have mostly
avoided discussing rapid events in detail, but
this was a high-calibre tournament in which
Karpov performed well and scored some
impressive victories.
One of his opponents in the qualifYing
group was Viktor Bologan. The Moldovan
grandmaster has contested twelve games
against world champions, scoring one win,
eight draws and three losses. He only ever
played one classical game against Karpov,
which was drawn.
I Game �3 1
Anatoly Karpov - Viktor Bologan
Cap d'Agde (rapid) 2000
l .d4 c!Llf6 2.c!Llf3 g6 3.g3 .ig7 4..ig2 0--0 5.c4
d6 6.0-0 c!Llc6 7.c!Llc3 a6 SJ�i:el
Bologan is an expert on the King's Indian,
and he has reached the position after Black's
seventh move in dozens of games. The text
move is one of Karpov's pet lines; see Game 45
and the accompanying references.
S .. J:i:bS 9J::l: b l .id7?!
This move is a bit slow. Black's best bet may
be 9 . . . tDa5 1 O.'lWa4 b6 as used by Shirov; see
page 369.
10.e4
Karpov reached this posmon in one
subsequent game. His choice of response may
well have been influenced by the situation of
the tournament in which he was participating.
The game continued: 1 O .b4 e5 1 1 .dxe5 tDxe5
1 2.tDxe5 dxe5 l 3 .b5 i.e6?! 1 4.i.a3 !!e8
1 5 . bxa6 bxa6 1 6.i.c6
8
7
6
5
L=;, ///" //'
4
3
2
a
b
c
d
e
f
g
h
1 6 . . . tD d7? ( l 6 . . . 'lWxd l 1 7.E!:exd l E!:ecB was
better, but Shirov evidently wanted to keep
more complexity in the position.) 1 7.E!:xb8
'lWxbB I B . tD d5 'lWdB 1 9 .'lWa4 i.fB 20.i.xf8
<;t>xfB 2 1 .E!:dl Wg7 22.tDxc7 (22.'lWa5! was
2000
435
Anatoly Karp ov - Viktor Bologan
even stronger.) 22 .. :�xc7 23 .�xd7 E:e7
24.�xe6?! Karpov only needed to draw this
game to qualifY for the final. Under normal
circumstances I have no doubt that he would
have preferred the stronger and more ambitious
24.�c6!.) 24 . . . E:xe6 25 .\Wb4 E:c6 26.E:c 1 White
went on to win, Karpov - Shirov, Prague (blitz
- 4.4) 2002.
10 ... e5 H .d5 lLle7
8
7
6
5
4
3
2
After 1 5 . . . ttJ f6 1 6. ttJ d2 b6 1 7.cxd6 cxd6
1 8 .b5 White's queenside play is too fast.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
16.c6!
By closing the queens ide Karpov slows down
his attack, but on the other hand he creates
two long-term targets on b5 and c7.
16 ... i.c8 17.axb5 axb5 18.i.c1!
a
b
c
d
e
f
g
h
12.c5!
Karpov takes the opportunity to gain space
on the queenside, as occurs in many variations
of the King's Indian.
Th e bishop had n o future o n a3 , s o it returns
to its original square in order to impede the
. . . f4 advance.
18 ... lLlf6
12 ... lLle8 13.b4 h6
Bologan attempts to improve over a game
from a few years before, which continued:
13 . . . <;t>h8 1 4.�a3 ( 1 4.a4!?) 14 . . . f5 1 5 .b5 axb5
1 6.tLlxb5 fxe4 1 7. ttJ g5 ttJxd5 This complicated
position eventually resulted in a draw, Soppe
- Zapata, Paulinia 1 997. The Moldovan
grandmaster prefers to prepare . . . f5 without
allowing the enemy knight to go to g5 .
14.a4
a
b
c
d
f
g
h
Karpov intends to carry out a full-scale
queens ide assault.
19.exf5!
14... £5 15.i.a3 b5
With this exchange Karpov takes most of the
poison out of Black's kingside attack.
e
The Prime Years
436
19 ... tiJxf5!
Bologan correctly decides that giving up the
e4-square is the lesser evil. He will get some
compensation in the form of the d4-square.
23.ctJxe5 (After 23 . ctJ d2 ctJ d4! Black becomes
active.) 23 . . . dxe5 24.ctJxb5 'lWe7 (not 24 .. .'�f7?
25 .j,xe5) Black is very much in the game.
21...g4?!
1 9 . . . gxf5 gives Black more control over the
central squares, but his pawns are rather
cumbersome. 20.ctJh4! e4 (After 20 . . . 'lWe8
2 1 .�al 'lWf7 22.�a7 White has strong pressure
on the queens ide.) 2 1 .j,e3 (Also after 2 1 .h3
'lWe8 22.'lWb3 ctJ h7 23 . ctJ e2 White controls
several important squares.) 2 1 . . .ctJ g4 22.j,d4
ctJe5 23.f3 White breaks up his opponent's
pawns and maintains a clear advantage.
20.i.b2!
Karpov refuses to give up the d4-square
cheaply. Less strong is 20.ctJd2 ctJ d4 2 1 .j,b2
j,f5 when Black gets some counterplay.
20 ... g5
a
b
c
d
e
f
g
Bologan gains space on the kingside and
hopes to improve the g7 -bishop. The idea is
logical but it was not the best choice.
2 1 . . .ctJ d4 was possible, although after 22.tLJe2
ctJxe2t 23 .'lWxe2 j,f5 24.�al White has a small
but pleasant positional advantage.
Black's most promising idea was:
2 1 . . .'lWe7!
Intending to transfer the queen to 0, from
where it will eye the vulnerable pawns on d5
and f2.
22.�al
22.ctJb3 'lW0 23 .'lWd2 (23 .�al tLJ d4)
23 . . . 'lWh5 Black will follow up with . . . tLJe7
and . . . j,h3, with active play.
22.ctJf1 'lW0 23.ctJe3 ctJ d4 24.�al (24JW
�a8) 24 . . .j,g4 2 5 . ctJ xg4 ctJxg4 26.�f1 IWf5
27.ctJ e4 White is just a little better.
22 . . . 'lW0
22 . . . ctJ d4 23 . ctJ ce4 ctJxe4 24.ctJxe4 �f5
25.�a7 is good for White.
23.�a7 ctJ d4 24.ctJ de4 ctJxe4 25.ctJxe4 g4
26.'lWd2
White keeps the advantage, but to a lesser
extent than in the game.
h
21.tiJd2!
It is not easy to read Karpov's intentions.
It may appear that he wants to put a knight
on e4 quickly, but for the moment he is more
concerned with reducing his opponent's active
possibilities.
Going after the b5-pawn was not fully clear:
2 1 .�al 'lWe8 (2 1 . . .'lWe7 22.�a5) 22.�a5 g4
a
b
c
d
e
f
g
h
2000
Anatoly Karp ov - Viktor Bologan
22.tlJb3!
Taking the d4-square from the enemy knight.
22 h5 23.i.cl!
.•.
Karpov prevents the enemy bishop from
becoming active on the c 1 -h6 diagonal.
23 ... llJh7 24.tlJe4
Preventing . . . tLlg5 is logical, but it was not the
only strong idea. White could also have gone
after the c7-pawn immediately with 24J'h l !?,
for instance 24 ... tLl g5 25 .E1a7 E1f7 26.i.xg5
"\Mfxg5 27.Wie2 and White has a convincing
advantage.
24 Wie8?!
..•
Bologan is only thinking about the kingside,
but he would have done better to invest a
tempo improving his situation on the opposite
flank. After 24 . . . E1aB! White's advantage would
have been smaller than in the game: 25 .Wie2
ttJd4 26.tLlxd4 exd4 27.i.f4 Black is somewhat
worse, but he is very much in the game.
437
27. tLl c3
27.Wid3 ?! is less promising: 27 . . . tLl d4 2B .i.e3
(After 2B.tLlxd4 exd4 29 .i.b2 i.f5 Black is
active.) 2B . . . tLl f3t 29.i.xf3 gxf3 30.tLl bd2
i.f5 Black's kingside play is dangerous.
27 . . . Wif6
27 . . . tLl d4 2B.tLlxd4 exd4 29.E1eBt E1fB
30.E1xfBt tLlxfB 3 1 . tLl e2 White wins
material.
27 . . . tLl g5 2B .i.xg5 Wixg5 29.Wid3 Wif6
30.E1e2 Black is in trouble.
2B .E1e2 i.h6
2B . . . h4 can be met by 29 .Wid3 .
29.i.xh6 tLl xh6 30.Wid3 i.f5 3 1 .Wie3 tLlg5
32.E1b7
White wins the b5-pawn. Black's kingside
counterplay should not be underestimated,
but White should be able to handle it.
27 tlJg5 28.llJxg5
•.•
Obviously White should exchange the knight
before it causes trouble on f3 or h3.
28 Wlxg5
.•.
8 ��;--'���
7
8
7
6
6
5
5
4
4
3
3
2
2
a
b
c
d
e
f
g
h
25.gal!
The main object of attack is the c7-pawn.
25 Wlg6 26.ga7 gO 27..ib2!?
•..
Karpov probably played this move to prepare
"\Mfa! and E1aB . It was also possible to go after
the b5-pawn:
1
a
b
c
d
e
f
g
h
29.Wlal
Karpov's play is reminiscent of some of
his games from the sixties, when he was
willing to send his queen to the far reaches
of the board in order to accomplish his
objectives.
The Prime Years
438
He could have also have deployed his strongest
piece in the centre by means of 29.We2 h4
30.W.e4, keeping a slight plus.
29 Wf6
•..
After 29 . . . h4 30.Wa5 Wd8 3 1 .W.c 1 W.h6
32.W.xh6 ltJ xh6 33 .W.e4 'kt>g7 34.:8e2 White is
a bit better.
30J::g f1 ?!
This is not the best place for the rook.
30.:8a8? would have been disastrous in view of
30 . . . ltJxg3!.
The strongest continuation was: 30.:8e2! h4
3 1 .Wa5 ! It is important to draw the black
queen away from the f-file. (White has no time
to exchange rooks: 3 1 .:8a8? :8xa8 32.Wxa8 :8f8
33 .Wa5 [or 33 .Wb8 ltJe3!J 33 . . . ltJ e3! 34.fXe3
h3 Black has a dangerous attack.) 3 1 . . .hxg3
32.hxg3 Wd8 33.:8a8 :8xa8 34.Wxa8 :8f8
3 5 .Wa5 The b5-pawn will soon fall.
31.. ..iffi?
Bologan prepares to put his heavy pieces on
the h-file, but misses a golden opportunity to
seize the advantage:
3 1 . . .W.h6!
Out of nowhere the a7-rook has become
tactically vulnerable.
32.We l
Defending the e3-square.
32.:8a2 ltJ e3 Black wins an exchange and
keeps his attack.
32.ltJ c5 Black can ignore this cheeky sacrifice:
32 . . . hxg3 33.hxg3 ltJxg3!-+
32 . . . hxg3
After 32 . . . ltJ e3 33.:8xc7! White remains in
the game.
33.hxg3 ltJxg3
Black is winning.
32.Wdl!
Karpov probably noticed the danger after
his previous move and immediately brings his
queen back into the defence.
30 h4
•.•
32 ...WgS 33 ..icl Wh5
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
a
b
c
d
e
f
g
h
31..ie4?
Karpov forgets about his king for a moment.
He could still have kept an edge by transferring
his queen to the centre. 3 1 .We 1 ! ltJ d4 (3 1 . . .W.h6
32.We2) 32.ltJxd4 exd4 33 .We8t :8f8 34.Wh5
Black has numerous weaknesses.
1
a
b
c
d
e
f
g
h
34.We2!
The queen defends along the second rank,
thus neutralizing Black's play on the h-file.
34 .. J::g h7 35.:8el
2000
439
Anatoly Karp ov - Viktor B ologan
This move does not spoil anything, but
White could have carried out the same plan
more efficiently with 3 5 .j,d2!?
the back rank was less convmcmg: 40J'1:a8
l"i:xa8 4 1 .l"i:xa8 ltJ e7 Black keeps his position
together.
35 �h8 36.i.d2 i.h6 37.i.xh6 'lWxh6
40 'lWg5 41.'lWd2 'lWh5 42.ga2
.•.
.•.
42.ltJ c5 ! ? was interesting but hardly
necessary: 42 . . . dxc5 43 .bxc5 ltJ h6 44.l"i:xf7
ltJxf7 45 .d6 White's central pawns give him
the upper hand.
A promising alternative was: 42.ltJ a l ! ? White
has time to go after the b5-pawn. 42 . . . Wfg6
43.ltJc2 cj;Jg7 44. ltJ a3 White is winning as
Black has no defence against j,d3 followed by
taking the b5-pawn.
42 1L1h6?
..•
a
b
c
d
e
f
g
h
38J�eal !
Threatening to create serious problems on
the back rank.
38 ... hxg3 39. fxg3 gO
The rook escapes the pin and moves to the
open file.
a
b
c
d
e
f
g
h
40.gfl !
Karpov plays across the whole board and
switches to the kingside, setting up another
pin in the process. Attempting to attack on
Now a comedy of errors starts. Bologan's
position is so bad he can hardly move, so it was
understandable that he wanted to relieve the
pressure on the f-file, nevertheless his chosen
move can be refuted immediately.
42 . . . cj;Jg7 was better, but after 43.l"i:f2 ct?g8
44.l"i:al cj;Jg7 45.l"i:afl l"i:a8 46.j,d3 Black will
lose the b5-pawn sooner or later.
a
b
c
d
e
f
g
h
43.'lWd3??
It is a pity after playing such a great strategic
game Karpov misses the simple finish: 43.l"i:xf7!
ltJxf7 44.Wff2! ct?g7 45.l"i:a7 White wins the
The Prime Years
440
c7-pawn and invades decisively. We should
remember that this was a rapid game, so the
players were probably short of time by now.
Black continues to improve his pieces. 49 .1!ge4
(49.'lMfd3 'lMfe3-+) 49 . . . lU e7 White is in trouble
as the d5-pawn is weak.
43 i.f5!
46 ... lLle3?
•..
Having suffered for so long, Black is suddenly
right back in the game.
44 .bf5?
This move only helps the black knight to
come back into play. White should have settled
for 44J'l:af2 1"!:bfB 45.lUd2 'lMfg6 46.'lMfe2 when
the position is equal.
.
8
Black gets rid of the pin but completely
squanders his advantage.
46 . . . 'lMfh3! would have led to decisive threats
against the white king: 47. lU d2 (47.1!ge2
lUxg3-+; 47.'lMfd3 lUh4-+) 47 . . . lU e3 4B .1!ge2
lUxfl 49.1"!:xf7 1"!:xf7 50.lUxfl 1"!:f3 5 l .b5 1!9h7
Black is winning.
8
7
7
6
6
5
5
4
4
3
3
2
2
1
a
b
c
d
e
f
g
h
46.'lMfxb5!?
Objectively this move is a mistake which
leads to a losing position, but practically it
turned out to be the right choice as it forced
Black to make an immediate decision.
46.lU d2? was unsatisfactory due to 46 . . . lU d4.
Objectively White could have offered sterner
resistance by waiting passively, but this would
have given Black the opportunity to organize
his position and strike at a moment of his
choosing. For example: 46.'lMfe2 1"!:f6! 47.'lMfd3
White plans to shuffle his queen back and
forth, and take the b5-pawn only when Black
moves his knight. 47 . . . i>g7! 4B .'lMfe2 'lMfg5
1
a
b
c
d
e
f
g
h
49 Y*lf3?
•.•
This leads to a hopeless endgame. Black
should have kept the queens on: 49 . . . lUxd5
50.'lMfxg4 lU e7 5 l .'lMfh4t i>g7 52.lUa5 lUd5
Black has good drawing chances as White's
king is exposed.
50.Y*lxa gxf3 5 1.lLl d2
Karpov takes no risks, although the flashy
5 l .lUc5!? was winning as well: 5 1 . . .lUxd5
52.lUa6 e4 5 3 .'tt> f2 'tt> g7 54.b5 'tt> f6 (54 . . . 'tt> g6
5 5 .lUxc7 lU b6 56.lUeB d5 57.lUd6-+)
5 5 .lUxc7 lU b6 56.h4 d5 57.h5 d4 5B.lUe8t
'tt> e7 (5B . . . i>g5 5 9 . lU d6) 5 9.h6 d3 60.h7 d2
6 l .hB='IMf d l ='IMf 62.'lMfe5t White wins.
Anatoly Karp ov - Viktor Bologan
2000
5I. ttJxd5 52.b5 <j;>g7
44 1
Karpov centralizes his king before pushing
his kings ide pawns.
••
59 ... ttJd4 60.e7t <j;>d7
a
b
c
d
e
f
g
h
53. <j;>f.Z!
Naturally Karpov uses his king to capture the
f-pawn, thus optimizing both of his remaining
pieces.
a
b
c
d
e
f
g
h
6I .g4
Finally the kingside pawns begin to march,
and the end is nigh.
61...ttJe6 62.h4 ttJe5t 63.<j;>f5 e4 64.h5 e3
65.ttJxe3 <j;>xe7 66.h6
1-0
The h-pawn cannot be stopped.
a
b
c
d
e
f
g
h
56.b6
Exchanging pawns is not a problem for
White. His passed c- and h-pawns are a long
way apart, and Black is unable to deal with
both of them.
56 ... cxb6 57.ttJxb6 ttJb5 58.ttJd5t <j;>d8
59.<j;>e4
Karpov remained undefeated in the qualifying
group and tied for first place with Hamdouchi
on 5/7. In the first elimination round he
started with two draws against Zhang
Zhong, then beat him in the third and fourth
games.
His opponent in the semi-final was Alexey
Dreev, whom he had never played before.
The former World Junior Champion has
played twenty classical games against the
world champions, scoring two wins, eleven
draws and seven losses. He drew four classical
games against Karpov, with no wins on either
side.
The Prime Years
442
10 ... a6?!
I Game �4 1
Anatoly Karpov - Alexey Dreev
Cap d'Agde (rapid) 2000
l.c4 c6 2.e4
When Karpov wants to play this line against
the Caro-Kann , he usually opens with l .c4.
Dreev has a narrow repertoire for a top-level
player, which would have made him easy to
prepare for. Had he been stronger in this area,
he might even have become a contender for
the World Championship.
2 ... dS 3.cxdS cxdS 4.exdS llJf6 S.llJc3 llJxdS
6.lLlf3 llJc6 7..ibS?!
According to the database this move was
first used by Keres in 1 967. Statistically it has
been the most popular choice in the position ,
which seems strange as i t i s n o t particularly
dangerous, and Karpov never used it again.
7.d4 is likely to transpose to a well-known
variation of the Panov after 7 . . . .tg4 or 7 . . . e6.
Provoking the exchange justifies White's play.
Dreev subsequently switched to 1 O . . . Wd6,
losing one game but winning two others.
1 O . . . .td7 and 1 O . . . .tf6 are also more reliable
than the game continuation .
1 1 ..ixc6 bxc6 12.llJeS
It was worth considering 1 2. ttJ a4!? in order to
avoid the second of the possibilities mentioned
in the following note.
12 ....ib7?!
It seems that Dreev underestimated the
difficulties associated with his backward
c-pawn.
1 2 . . . Wc7 has been played in a few games, but
after 1 3 .ttJxd5 cxd5 14 . .tf4 Black must be
careful as he is behind in development.
Black's best bet may well be 12 . . . ttJxc3!?,
which has the advantage of blocking the c-file:
1 3 .bxc3 Wc7 1 4 . .tf4 .td6 Black should be
able to obtain a reasonable game after freeing
his position with . . . c5, Bojkov
M. Berg,
Germany 2003.
-
7... e6
7 . . . g6 is rarely played, nevertheless Black
scores well with it.
8.0-0 .ie7 9.d4 0-0 10.�el
a
b
c
d
e
f
g
h
2000
Anatoly Karpov - Alexey Dreev
13 .. J�� e8
After 1 3 . . . id6 1 4 .id2 Wh4!? ( l 4 . . . Wc7
1 5 .Wg4 lLl f6 1 6 .Wh4 lLl d7 1 7.if4 White
keeps an edge.) 1 5 .ia5 White's grip on the
queens ide is unpleasant.
443
the most of his advantage. 20 . . . l::1 dc8 (Also
after 20 . . . l::1 d7 it will not be easy for White to
convert his extra pawn.) 2 1 .Wb3 Wa6 22.Wa4
Wb6 Black's positional compensation gives
him realistic chances to hold the position a
pawn down.
14.lLld3!
Karpov solidifies his grip over the c5-square.
14 a5
•••
Dreev was probably hoping to move his
bishop to a6 or even a8 at some point in
the future. Nevertheless the move also has
a downside
as the a-pawn becomes more
vulnerable.
b) I prefer 1 9 .Wg4! with the idea of building
White's position on the kingside. This idea
is stronger, and also more in keeping with
Karpov's style. 1 9 . . . ixe5 ( l 9 . . . ic8 20.h4)
20.dxe5 <tt> h 8 2 1 .h4 White keeps a clear
positional advantage and his kingside play is
not easy to neutralize.
19.9xe5
15.lLlae5 We7 16.i.d2 ga8 17.gcl gfd8
18.liJe5
Karpov continues to target the c-pawn.
Of course Karpov maintains the pressure on
the backward c-pawn. 1 9 .dxc5? ia6! would
stop the knight from coming to d6, and Black
should be fine.
19 ... a4
By this point it looks like Dreev had already
decided to sacrifice the c6-pawn.
a
b
c
d
e
f
g
h
18 ,he5
•••
Black faces a difficult dilemma. Should he
defend passively, or should he sacrifice the
c-pawn in a bid for counterplay?
The main alternative was 1 8 . . . if6, after which
White must decide whether to snatch the
c-pawn or continue to improve his position.
) 1 9 .1Llxb7 Wxb7 2o.lLlxc6 is possible, but one
gets the impression that White has not made
a
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
20.Wf3!
20.Wg4 was also attractive, but Karpov's
move is more versatile. The queen eyes the
f7 -pawn, clears the path for the rook to slide
across to c l , and in some positions the queen
may even swing to a3 .
The Prime Years
444
20 Y;lTe7
The last of the above points is nicely
illustrated in the following line: 20 . . . f6 2 1 . lLl c4
Wfd7 (After 2 1 . . .:B:e8 22.Wfa3! :B:a6 23.lLla5
Black's queens ide is under heavy pressure.)
22.Wfa3! Surprisingly White can use the queen
to collect the a4-pawn. 22 . . . :B:a6 23.:B:a5 The
a4-pawn perishes, and the weakness on c6
remains.
24.Wfa3
24.lLla5 Wfd7 25 .Wfd3 :B:a7 26.:B:2c4 lZlb6
27.:B:c 1 (27.:B:b4 Wfc7) 27 . . . lLld5 White has
decent chances, but Black still resists.
24 . . . Wff6 2 5 . lLl e5 Wfe7 26.:B:2c4 f6 27.lZl d3
Wfa7 28.lLlb4 lLlxb4 29.:B:xb4
Black probably loses the a4-pawn, but the
opposite-coloured bishops offer him some
drawing chances.
21 .:B:ec1
22.:B: lc2
2 1 .Wfa3 is not quite as convincing here:
2 1 . . . lLl f6 (After 2 1 . ..lLlc7 22.�e3 lLlb5 23 .Wfb4
Wfe8 White is somewhat better.) 22.Wfb4 h6
(22 . . . lLl d 5 ? 23.:B:xd5 +-) 23.:B:b5 Wfxb4 24.:B:xb4
White's advantage is smaller than in the game.
Karpov postpones capturing the pawn and
improves his position first. Nevertheless after
22.lLlxc6 �xc6 23.:B:xc6 Wfd7 24.Wfa3 :ga7
White has decent winning chances.
•.•
22 :B:a6 23.h3!?
•.•
Karpov continues to postpone the capture
on c6.
8
7
6
5
4
a
b
c
d
e
f
g
h
21...:B:db8?!
Dreev decides to abandon the c6-pawn in
order to search for counterplay on the b-file,
but on this occasion passive defence would
have been more stubborn:
2 1 . . .:B:a6
Black's position will remain unpleasant, but
it is not easy for White to find a way through.
Here is a possible continuation:
22.:B: l c2 :B:c8 23 .lLlc4 �a8
Also after 23 . . . :B:d8!? 24.lLla5 �a8 25 .Wfd3
Wfa7 26.lLlxc6 �xc6 27.:B:xc6 :B:xc6 28.:B:xc6
h6 it will not be easy for. White to convert
his extra pawn.
3
2
1
a
b
c
d
e
f
g
h
23 f6?!
•••
Dreev loses patience and forces Karpov to
take the pawn, but in doing so he weakens the
e-pawn.
24.ttJxc6 :B:xc6 25.:B:xc6 Y;lTd7??
It is difficult to tell what happened to Dreev;
forgetting to take the rook is j ust one of those
freakish blunders which usually only occur a
few times in a player's career.
2000
Anatoly Karp ov - Alexey Dreev
Let's see what might have happened if Black
had played correctly.
25 . . . �xc6 26.l''&xc6 'lWeB
Black cannot regain his pawn, as 26 . . . 1"i:xb2?
27.1"i:cBt 1itf7 2B.'lWh5t leads to mate.
27.1"i:c2 'lWg6 2B.1"i:cl
26.'i'g3 e5 27.:a6e5 :ae8
Had this position occurred under a classical
time limit, Dreev would probably have resigned
by now.
28.dxe5 £Xe5 29.'i'g4 'i'f7 30.'i'xa4
Winning a second pawn.
8
30 ... :af8 31 .'i'h4 h6 32.i.e3 :aa8 33.a3 :aa6
7
6
8
5
4
7
3
6
2
5
a
b
c
d
e
f
g
4
h
2B . . . 1"i:xb2?!
Black should prefer a quiet move like
2B . . . h6, when it is hard to determine
whether White can convert his advantage
into a win.
29.'lWa3!
After 29.1"i:cBt cj{f7 30.'lWa3 'lWb l t Black get
some counterplay.
29 . . . 1"i:c2
29 . . . 1"i:bB 30.'lWxa4 is a similar story.
30.'lWxa4
White restores his material advantage and
has good winning chances.
a
445
b
c
d
e
f
g
h
3
2
1
a
b
c
d
e
f
g
h
34.:ae8t!?
Karpov gives back the exchange, knowing
that his two-pawn advantage is more than
enough to ensure victory. The remaining moves
can pass without comment.
34 ... i.xe8 35.:axe8t �h7 36.'i'e4t :ag6
37.b4 'i'e6 38.:ad8 tlJc3 39.'i'a8 :af6 40.:ae8
'i'd6 41.:ah8t �g6 42.:ad8 'i'e7 43.'i'e8
'i'xe8 44.:axe8 tlJd5 45.b5 �f5 46.a4 �e4
47.a5 tlJ e7 48.:ae7
1-0
In the second game Karpov equalized in a
Petrosian Queen's Indian. Dreev launched
an attack which was objectively unsound,
but Karpov made some mistakes and lost. In
the third game Karpov had the black pieces
again, and equalized following an early queen
exchange. He then proceeded to outplay his
opponent and won a nice endgame. In the
fourth game Karpov played too overtly for a
446
The Prime Years
draw with the white pieces, and found himself in serious trouble. But Dreev let him off the hook
and he managed to escape into a drawn rook ending which he went on to hold.
Karpov faced Mikhail Gurevich in the final. In the first game the Ukrainian-born Belgian (now
Turkish) grandmaster played a dubious line of the Semi-Slav and Karpov destroyed him with a
brutal attack in the centre and kingside. In the second game Gurevich built a dangerous attack;
in response Karpov sacrificed an exchange but was unable to hold the position. In the third game
Karpov had a slight edge in a stonewall structure, but he sacrificed a pawn unnecessarily and paid
the price as Gurevich took the lead. In the final game Karpov tried to generate winning chances
on the black side of an English, but Gurevich played well and the long game ended in a draw.
Thus Karpov finished the tournament as the losing finalist.
2000 Summary
Match versus Bacrot, Cannes: Drew 1 - 1 (classical games) (+0 =2 -0)
Match versus Xie Jun, Guanzhou: Won 2Yz-l Yz (classical games) (+ 1 3 -0)
Match versus Ye Jiangchuan, Shenyang: Drew 1 - 1 (classical games) (+0 =2 -0)
Japfa Classic, Bali (2nd-3rd place) : 6/9 (+3 =6 -0)
Najdorf Memorial, Buenos Aires (4th place) : 5 Yz/9 (+3 =5 - 1 )
=
Total 6 1 . 5 % (+7 = 1 8 - 1 )
[j Wins • Draws • Losses
200 1
Rating 2679 (20 in the world)
Karpov's first tournament of the new year was Linares, where he started with the black pieces
against Shirov. He played a new move in the 4 .. .tZl d7 Caro-Kann and got a better position with
excellent winning chances, but Shirov eventually escaped to a drawn ending with rook against
rook and knight. Karpov followed with two fairly short and uneventful draws with Grischuk and
Judit Polgar, then had to settle for a draw with Leko after failing to make progress in a slightly
better double-rook-endgame. In the fifth round Karpov faced Kasparov, who prepared the sharp
4.ltJc3 line in the Advance Caro-Kann. Karpov was always in trouble and lost badly.
In the return game against Shirov, Karpov got nothing against the Queen's Gambit Accepted
and the game was always balanced until the draw was agreed. In the next game Grischuk followed
Kasparov's example and chose the 4.ltJc3 variation against the Caro-Kann. Karpov chose a quieter
response and came close to equalizing, but eventually blundered in time pressure and lost. In
Round 8 Judit Polgar sacrificed a pawn for promising compensation, but then went astray. Both
players committed a few more inaccuracies, but eventually Karpov prevailed. In the next game
Karpov drew solidly against Kasparov; the game is referenced in the note to White's twelfth move
in Game 56. In the last round Karpov was under a bit of pressure against Leko but was able to
hold the draw.
***
After taking part in several rapid and simultaneous exhibition events, Karpov took part in a
Europe versus Tatarstan match in Kaluga, where he made two fairly quick and uneventful draws
against Dreev.
His next tournament was the Najdorf Memorial in Buenos Aires. In the first round he equalized
with the Petroff against Milos. The Brazilian grandmaster tried launching an attack, but Karpov
got the better of the complications and won a piece followed by the game. In the next two games
Karpov drew fairly peacefully against Short and Felgaer. In Round 4 he got no advantage against
Mecking, but the former world championship candidate made a serious mistake and soon went
down.
In Round 5 Karpov equalized easily against Korchnoi and a quick draw ensued. In the next
round he played the Petroff against Xie Jun and sacrificed a pawn. He always had enough
compensation and a draw ensued. In the seventh round he faced Judit Polgar and got a slight edge
in a queenless middlegame, which he eventually converted to a win. In Round 8 Karpov faced
the young prodigy Radjabov, with whom he was presently sharing the lead in the tournament.
Karpov was pressing for a win with the black pieces in a major piece endgame, but Radjabov was
able to hold the draw.
The Prime Years
44B
Karpov's opponent in the ninth and final
round was Pablo Ricardi. The Argentine
grandmaster played four games against world
champions, scoring one win, one draw and two
losses. This was his last encounter with Karpov.
They met once before, in the final round of
the previous year's Najdorf Memorial, when
Karpov lost on time in a winning position.
Anatoly Karpov - Pablo Ricardi
l .d4 lLlf6 2.c4 e6 3.lLlc3 d5
It looks like Ricardi wanted to surprise
Karpov. His usual defences are the Nimzo­
Indian and the Griinfeld, and a year earlier
against Karpov he preferred the Benko
Gambit.
4.cxd5 exd5 5.i.g5 lLlbd7 6.e3 c6 7.i.d3
i.d6
8
7
6
5
4
3
2
1
c
d
e
Black wants to play . . . tO g6 followed by ... h6
to exchange the g5-bishop.
9.tLle5
f
9... lLl g6
In the stem game Lj ubojevic preferred
9 . . . We7 and eventually drew.
Buenos Aires 200 1
b
8 ... lLlffi
According to the database this aggressive
move was first played by Portisch in 1 97B.
I Game �� I
a
Karpov always developed his knight to f3 in
the exchange variation.
g
h
Karpov has enjoyed great success against the
main lines with the bishop on e7. This was the
only time he encountered the development of
the bishop to d6.
The Yugoslav grandmaster later switched to
9 . . . Wb6 with unfortunate results. 1 0.0-0
Lj ubojevic tried two moves here:
a) The first game continued 1 0 . . . '.Wxb2
1 1 .l"Ic1 tO g6?! ( l 1 . . .�xe5!? 1 2.dxe5 ctJ 6d7
deserved attention; but not 1 2 . . . tO g4? 1 3 .'.Wa4!
transposing to Hjartarson - Ljubojevic below.)
1 2.f4 ( l 2.�xf6 gxf6 1 3 . ctJ g4�) 1 2 . . . 0-0 1 3J''lc 2
Wb6 1 4.�xf6 gxf6 1 5 .tOg4 �xg4 1 6.'.Wxg4
<j{hB?! ( l 6 . . . l"IaeB!?) 1 7.l"Ib l Wc7 I B .ctJxd5
WdB 1 9 .tOc3 �xf4 20.l"Ie2 l"IeB 2 1 . tO e4 �xe3t
22.l"Ixe3 Wxd4 23.l"Ibel tOe5 24.Wf5 1 -0
Timman - Ljubojevic, Hilversum 1 9B7.
b) Two years later Ljubojevic tried to improve
with 1 O . . . �xe5 but suffered an even more
brutal defeat: I l . dxe5 ctJ g4?! ( l 1 . . . ctJ 6d7)
1 2.Wa4! Wxb2 This whole plan is much too
risky. 1 3 .l"Iac 1 a5 ? 1 4 . tO b 5 ! tO e6 1 5 . ctJ d6t
<;t>fB 1 6.Wxg4 Wxe5 1 7.tOxcB h5 I B .ctJb6
hxg4 1 9 .tOd7t <j{eB 20.tOxe5 tOxg5 2 1 .E1b l
Black is a piece down and soon had to resign,
Hjartarson - Lj ubojevic, Tilburg 1 989.
10.f4
Karpov strengthens his knight on e5.
10 ...V;¥b6?!
Anatoly Karpov - Pablo Ricardi
200 1
449
Ricardi plays a novelry, which was never
repeated. The idea of letting his opponent
double the pawns leads to a passive position.
1 0 ... 0-0 and 1O ... �e7 are played here.
8
7
6
5
4
a
3
b
c
d
e
f
g
h
15 ... gac8
2
a
b
c
d
e
f
g
h
1 1 .�c2
Unlike Timman and Hjartarson, Karpov
does not feel the need to sacrifice the b2pawn. The queen is well-placed on c2 where
she exerts an influence over both sides of the
board.
A knight transfer to d6 is often a good plan
in this pawn structure, but here it is not easy
to accomplish: 1 5 . . . c;!;>g7 1 6.a3 h6 1 7.b4 lLl e7
1 8 .lLla4 Wc7 1 9 . 1Ll h4 White is somewhat
better.
16.g3
Maybe Karpov was considering playing e4
at some point.
16 ... @g7 17.gacl
1 1...0-0 12 ..bf6
Karpov is happy to create a structural
weakness in his opponent's camp.
1 2.0-0! ?
White can also continue developing and
concentrate on bringing more pieces into
the attack.
1 2 . . . :ge8 1 3.:gae l Wc7
1 3 . . . �e7 1 4.f5 lLlxe5 1 5 . dxe5 lLl g4 1 6.f6
White's kingside initiative is dangerous.
14.�xf6 gxf6 1 5 .lLlxg6 fxg6 1 6.�xg6
White wins a pawn.
12 . gxf6 13.ll)i'3 ge8 14.@fl
Karpov defends his backward pawn and
moves his king off the e-file.
..
14 .
.
.i.g4 15.ghe1
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
17 ...�d8
It is hard to suggest an active plan for Black.
17 . . . :ge7? allows White to play 1 8 .f5 ! lLl f8
1 9. lLl xd5 ! , winning.
The Prime Years
450
1 7 . . .Ek7 1 8 .<j;lg2 :gce7 1 9.'lWfl .txf3t 20.i>xf3
lLl f8 2 1 .:ge2 White has the better pawn
structure and more active pieces.
20 :gce7 21 .�cdl
17 . . . :gg8 1 8 .lLl d2 :gce8 1 9.i>g2 <j;lh8 ( 1 9 . . ..td7
20.h3 'lWd8 2 1 .'lWd l .tc8 22.'lWh5 White
remains in control.) 20.h3 .tc8 2 1 .lLl f3 White
undoubtedly has the advantage, although it is
hard to say how serious it is.
21...i.c7 22.�e2 Wfd6
•..
Karpov makes sure Black will not be able to
become active in the centre with . . . c5.
8
7
6
5
4
3
2
1
a
b
c
d
f
e
g
h
23.<;!{fl!
a
b
c
d
e
f
g
h
With this prophylactic king move Karpov
prevents the possibility of23 . . . lLl h4t intending
24.gxh4 �xe3t .
18.i.fS i.xf3?!
Ricardi chooses the less desirable exchange.
In the resulting position White's bishop will be
a powerful force.
The lesser evil was 1 8 . . . .txf5 1 9.'IWxf5
.tb4 20.e4 when White's position is clearly
favourable, but Black has more chances than
in the game.
19.<;!{xf3
Naturally Karpov is not tempted by 1 9 . .txc8?
'lWxc8 20.i>xf3 'lWh3 when Black has enough
counterplay to maintain the balance.
After the text move the king is safe enough
on f3, as it has enough protection and Black is
in no position to carry out an attack.
19 .. J:�c7 20.a3!
Karpov takes away any hopes Black may have
had to swap his somewhat paralysed bishop for
the strong white knight.
23 a6 24.i.g4 i.b6 25.WffS
•.•
Karpov brings the queen closer to the enemy
king.
25 ...i.a7
8
7
6
5
4
3
2
1
a
26.<;!{f3!
b
c
d
e
f
g
h
Anatoly Karpov
200 1
Superb play with the king. The main point
of this move is to enable the d 1 -rook to move
to a different file, as the trick based on . . . l'l:xe3
and . . .�xd4 will not work.
26 ... bS?!
Ricardi does not want to wait passively,
but by going for counterplay he weakens his
position.
Better was 26 . . .tLlf8 27.l'l:de 1 <;t>g8 28.h4 ltJ e6
29.�g2 <;t>g7 30.Wh5 when Black's position is
unpleasant, but still less problematic than in
the game.
-
Pablo Ricardi
45 1
White remains in control and will soon start
exerting pressure on the c-pawn.
31 .i.f3 �4e7 32.�d!
Karpov homes in on the weak c6-pawn like
a shark that smells blood.
32 ...�d7 33.lLlcs �h3 34.i.g2 �hS?!
The queen achieves nothing on the kingside
and will soon find herself cut off from the
action. 34 . . . Wc8 was better, although 3 5 .Wd2
wins the b4-pawn for no compensation.
8
27.�d3
Karpov brings his strongest piece back
to guard the queens ide, anticipating his
opponent's play.
27 ... aS
7
6
5
4
Ricardi keeps going forward.
3
28.c;t>f2 b4 29.axb4 axb4 30.llJa4
2
1
a
b
c
d
e
f
g
h
3S.h3!
Karpov prevents the black queen from
returning to the queenside.
3S ... llJf8 36.llJb3 �e6 37.£5!
Forcing the enemy rook off the e-file.
37... �d6 38.i.f3
a
b
c
d
e
f
g
h
30 ... l'l:e4?
This turns out to be a waste of time.
In the event of 30 . . . h5 3 1 .�f5 h4 Black
reduces his number of pawn islands, but after
32.l'l:g1 the open g-file is more important.
30 . . . l'l:a8 was more resilient, but after 3 1 .l'l:al
(3 1 .l'l:c l ?! �xd4!) 3 1 . . .l'l:ee8 32.b3 l'l:h8 33 .�h5
Small tactics also work for White - taking
the h-pawn would cost Black his queen.
38 ...�h6 39.llJaS!
Karpov plays across the whole board. The
attack on the c6-pawn is too much to bear.
39 ... i.b8 40.h4!
Karpov not only safeguards the h-pawn, but
also paralyses the enemy queen in the process.
452
The Prime Years
40 Jks 41.V!lfa6!
••
Having caged the enemy queen on the opposite flank, Karpov goes after the weak c6-pawn.
41..J�c7 42J�xc6
Finally the doomed pawn perishes, and its colleagues on b4 and d5 will soon follow.
42 .. J�a7 43.V!lfb6 tLld7 44.V!lfxb4
1-0
In this hopeless position Black resigned.
Korchnoi and Radjabov both drew, so Karpov's final score of 6Yz/9 was enough to claim first place
outright. This was Karpov's last ever tournament victory at a classical time control. According to
my calculations, starting with his 1 966 Trinec victory he has won sixty five individual classical
tournaments, including thirteen in which he shared first prize. No other player in chess history
has won so many high-level tournaments.
Karpov's next event was the Spanish team championship, where he defeated Matamoros Franco
and drew against Giorgadze and Cifuentes Parada.
Karpov's final event of the year was the FIDE World Championship, which once again took the
form of a knockout tournament. For much of chess history, the World Champion enjoyed many
more privileges than in most other games and sports. Take the 1 998 version of the same event,
where Karpov was seeded directly into the final, whereas his challenger Anand had to undergo
a gruelling series of elimination matches. In 200 1 Karpov received no such privilege, and he
had to start in the first round j ust like everyone else. He drew both of his classical games against
Zhang Pengxiang of China, then lost both rapid tie-break games. Thus the reigning champion
was eliminated in a shocking upset. The title was eventually won by Ruslan Ponomariov, who
defeated Ivanchuk in the final.
2001 Summary
Linares (2nd-6th place) : 4Yz/ 1 0 (+ 1 =7 -2 )
Europe - Tatarstan match, Kaluga (versus Dreev) : Drew 1 - 1 (+0 =2 -0)
Najdorf Memorial, Buenos Aires ( l st place) : 6Yz/9 (+4 = 5 -0)
Spanish Team Championship: 2/3 (+ 1 =2 -0)
FIDE KO World Championship (versus Zhang Pengxiang) : Drew 1 - 1 (+0 =2 -0) (Eliminated
in rapid tie-break.)
Total 57.7% (+6 = 1 8 -2)
Iii! Wins • Draws • Losses
2002
Rating 2693 ( 1 6 in the world)
Karpov played one game in the Belgian Team Championship, where he defeated Kim Le Quang,
FIDE Master.
His first major tournament was the Nao Masters in Cannes. He started with a draw against
Fressinet, followed with a win over Nataf and draws against Bareev and Gelfand.
His opponent in Round 5 was Peter Leko. The Hungarian grandmaster has so far contested
1 68 games against all the world champions from Karpov to Anand. He has won fifteen of those
games, drawn 1 20 and lost thirty three. This was his last regular game against Karpov. Previously
the two players had met eleven times, with the score standing at two wins to Karpov, with nine
draws and no defeats.
a
I Game �6 1
Anatoly Karpov - Peter Leko
Cannes 2002
l.d4 liJf6 2.liJa g6 3.g3 .ig7 4..ig2 d5 5.c4 dxc4 6.ltJa3
The next time Karpov faced this variation he deviated surprisingly early: 6JWa4t ltJ fd7 7.Wxc4
ttJ b6 8 .Wc3 ltJ a6 9 . 0-0 c5 1 0.e3 0-0 1 1 . ltJ a3 j,f5 1 2.�d1 cxd4 1 3 .exd4 ltJd5 1 4 .Wb3 ltJ ab4
1 5 .j,d2 a5 1 6.ltJe5 j,e6 1 7.Wa4 j,xe5 1 8 .dxe5 j,g4
a
b
c
d
e
f
g
h
1 9 .j,h6!? j,xd 1 20.�xd 1 �e8 2 1 .ltJb5 Wd7 22.a3 e6 23.h4 Wc6? Maybe Karpov's original play
took too much energy from the Dutch grandmaster and here he makes a serious mistake. (23 . . . ttJc6
24.We4 We7 25.ltJd6 �ed8 is unclear, as Ftacnik pointed out.) 24.axb4! axb4 25 .j,xd5 exd5
26.Wxb4 White easily converted his material advantage, Karpov - Van Wely, Wijk aan Zee 2003 .
The Prime Years
454
6 c3
•••
Compared with 6 . . . 0-0 and other
alternatives, the text move leads to a change
in the pawn structure which tends to result in
more complex play.
7.bxc3 0-0 8.0-0 c5
I think that one of the reasons why Leko was
able to break into the world's elite at such a
young age was that his opening repertoire was
modern. Your author taught this line to him
when he was eleven years old. The repertoire I
taught him was partly based on the preparation
of Karoly Honfi. Leko tends not to say much
about his early years, but the fact that he was
able to use his childhood repertoire against
great champions speaks for itself
9.e3
According to the database this move was first
played by Panno in 1 987.
9 ... ftJc6 10JWe2 .ifS l l .�dl 'Wfb6
8 � �
�,)�;.
7 ��
, � �W� ' m '
6 " " '%Br� " '''%_fiY-�
Leko's previous game continued: 1 2.4Jc4
'lMfa6 1 3 .iJ1 4J a5 1 4 . 4J ce5 'lMfxe2 1 5 .ixe2
ftJ e4 1 6.ib2 2"1fd8 1 7.id3 ftJ d6 1 8 .ixf5 gxf5
1 9.ia3 b6 Black had equalized and the game
was later drawn, Nikolic - Leko, Bastia 200 1 .
Against Kasparov the previous year, Karpov
tried 1 2 .ib2 but achieved very little: 1 2 . . . Ei:fd8
1 3 . ftJ c4 'lMfa6 1 4.a4 4J a5 1 5 .ifl lLl e4 1 6.4Jfd2
4Jxd2 1 7.lLlxa5 'lMfxa5 1 8 .2"1xd2 cxd4 1 9.cxd4
Ei:ac8 The position was equal and a draw
ensued, Karpov - Kasparov, Linares 200 1 .
I n two subsequent games Karpov opted for:
1 2. 4J d2
On both occasions he obtained a large
advantage, but this can mostly be attributed
to his opponents' overoptimistic moves.
1 2 . . . cxd4 1 3 .cxd4 2"1ac8 1 4.h3
The following year Karpov deviated with:
1 4 . lLl b3 2"1fd8 1 5 .id2 h5 1 6.ie l g5?!
1 7.lLlc4 'lMfb5 1 8 .2"1ac l e5? 1 9.dxe5! 4Jxe5
20.2"1xd8t 2"1xd8 2 1 . 4J d4 White obtained a
winning position, Karpov - Illescas Cordoba,
Lanzarote 2003.
14 ... 2"1fd8 1 5 . 4J b3 lLl e4 1 6.ib2 g5 1 7.g4 ig6
1 8 .lLl c4 'lMfb5 1 9 .ifl 4J b4 20.4J bd2
�� I.
"�
�� L '.�� �.
��r�
�r�" " %�r�
��%'0%
tfl�
3 �
f':\
tfl�
tfl� ""2-1
�
%'0%
%'0%
"
2 /� � "",%� iid'£�
""� ��7;; �." " %f&"if" "
1 �
�:� �
5
4
"//
a
b
c
d
e
f
g
h
12.�h4?!
Leko had reached the same position the
previous year, so this rare continuation
may well have been prepared by Karpov in
advance. Despite his eventual success in this
game, the idea is unimpressive and it never
caught on.
a
b
c
d
e
f
g
h
20 . . . 4J d6??
Judit commits a fatal blunder.
After 20 . . . 2"1c7 or 20 . . . 4Jxd2 Black would
have been at least equal.
2 1 .4Jxd6 'lMfxe2 22.lLlxc8!
2002
Anatoly Karp ov - Peter Leko
Th e youngest Polgar sister probably
overlooked that White does not have to take
back the queen immediately.
22 . . .'IWxfl t 23.lLl xfl
White went on to convert his extra exchange,
Karpov ]. Polgar, Moscow (rapid) 2002.
-
12 ...J.g4!
Leko reacts in the most principled way.
13.0?!
1 3 .�f3 was safer, but this would hardly have
been a serious try for an advantage.
45 5
15.J.b2
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1 5 ... c!lJb6
This is not a bad move, but more precise
would have been I 5 . . . �e6! I 6.�fl l"i:fd8 . Black
stands better here, although the magnitude of
his advantage is a matter for debate.
16.J.f1
Karpov is happy to exchange queens. I 6.fxg4
\Wxc4 gives Black no problems.
16 ... V9xc4 17.V9xc4 c!lJxc4 18.hc4 c!lJ a5
19.J.e2 J.e6
Black also has a comfortable game after
I 9 . . . cxd4 20.cxd4 (20.fxg4? dxc3) 20 . . . �e6.
14.c!lJc4
I 4 .�b2?! is inadvisable due to: I 4 . . . lLlxe3!
I 5 .d5 (Other moves are worse, for instance
I 5 .fxg4?! lLlxd I I 6.l"i:xd I cxd4 or I 5 .dxc5?!
'&xc5 I 6.\Wf2 �e6 I 7.lLlc2 g5 , with a clear
advantage to Black in both cases.) I 5 . . . lLlxd I
I 6.l"i:xd I �d7 I 7.lLlc4 \Wa6 I 8 .dxc6 �xc6
White is in trouble.
14 ...V9a6!
After I 4 . . . lLlxc3? I 5 .\Wd3 \Wc7 I 6.V9xc3 �e6
I 7.�fl White has an edge.
a
b
c
d
e
f
g
h
The Prime Years
456
23 ... b5 24J�d2 �Ue8
20.d5
Karpov certainly obtained no advantage
from the opening, and had to play carefully
to avoid becoming significantly worse. With
this move he keeps the c-file closed in order to
limit Black's active prospects.
20 .id7 21 .e4 e6
..•
Leko decides to undermine the centre. Other
reasonable possibilities included 2 1 . . . b5!? 22.f4
lLl c4 and 2 1 . . .f5 ! ? 22.exf5 g5 23 . lLl g2 �xf5 .
22. tLl g2
Having escaped from a dubious opening
with a roughly equal position, Karpov sets
about improving his pieces. First there is the
knight, which has been sidelined ever since the
questionable novelty on move twelve.
a
b
c
d
e
f
g
h
25.<.t>fl
Karpov brings the king closer to the centre.
25 tLlb7
.•.
22 ... exd5
22 .. .f5 ! ? , opening the centre and making it
harder for White to use his king, was worth
considering. 23.lLlf4 fXe4 is fine for Black.
Ftacnik's 25 . . . c4! ? 26.l::k l lLl b7 is a sensible
suggestion.
26.c4
It makes sense for White to exchange his
passive bishop.
8
7
26 b4
.•.
6
26 . . . �h6 should be met by 27.lLlf4, rather
than 27.f4 ? lLl d6.
5
4
27.hg7 <.t>xg7
3
2
1
a
b
c
d
e
f
g
h
23.exd5
23.Ei:xd5? only helps Black: 23 . . . �c6
24.Ei:xc5 b6 25 .Ei:g5 Ei:fd8! White has problems.
Incidentally, Ftacnik suggested 25 . . . Ei:ad8 but
I prefer to use the other rook so that �a3 will
not win a tempo. Besides, the other rook will
be happy enough on c8.
a
b
c
d
e
f
g
h
Anatoly Karpov
2002
-
Peter Leko
457
28J�b2
31.tLlfl!
After 2B.d6? E'1adB the d-pawn is much more
of a weakness than a strength.
When i t comes t o identifying a weakness in
the enemy position, there are few players who
can rival the twelfth World Champion. With
this move he homes in on the vulnerable c5pawn.
28 ... liJd6
The knight stands well here, but Black could
also have gone for an immediate queenside
advance: 2B . . . a5 ! ? 29.a3 E'1ebB Black has a
decent position, and he retains the option of
meeting axb4 with . . . cxb4 when his knight will
find an ideal home on c5 .
29.a3 a5
Black could also have considered 29 . . . bxa3!?
30.E'1xa3 a5, hoping that his piece activity would
compensate for the structural weakening:
3 1 . liJ e3 (3 1 .E'1b6? E'1xe2t 32.<j;lxe2 tLlxc4 wins.)
3 1 . . .E'1a6! (3 1 . . .E'1edB 32.E'1b6!) 32. tLl d l a4
33 . tLl c3 The position is balanced.
31..J�eb8
Ftacnik recommends 3 1 . . .tLl b7?! 32.tLld2
'tt> e 5, but it turns out that White can embarrass
the adventurous king: 33.tLle4! E'1abB (33 . . . f5 ?
is refuted by 34.f4t! 'tt> xe4 3 5 .E'1dl followed by
36.�f3 mate!) 34.E'1b3 a4 3 5 .E'1e3 Black's king
is in trouble.
However, the immediate king improvement is
not such a bad idea: 3 1 . . .<j;le5 32.tLld2 E'1ecB
Black should be all right here.
32.axb4
30.liJe3 'ii? f6 ?!
Activating the king looks natural, but it was
not the most accurate. 30 . . . bxa3? would also
have been a mistake in view of 3 1 .E'1b6!.
However, Karpov's suggestion of 30 . . . E'1ebB!
would have given Black a comfortable game:
3 1 .axb4 (Otherwise Black advances his a-pawn,
for instance 3 l .tLl fl ?! a4! And White has
problems.) 3 1 . . .axb4 32.E'1ba2 E'1xa2 33.E'1xa2
E'1b6 Black is doing fine.
8
7
6
5
4
3
2
1
8
7
6
5
4
3
a
b
c
d
e
f
g
h
32 ... axb4
After 32 . . . cxb4?! 33.c5 tLl f5 34.c6 �cB
3 5 .�d l E'1b5 (35 . . . tLl d6 36.�a4) 36.�b3 E'1c5
37. tLl d2 White's pawns are more dangerous
than Black's.
2
1
a
b
c
d
e
f
g
h
On the other hand 32 . . . E'1xb4!? 33.E'1xb4 cxb4
is not so bad: 34.c5 tLl f5 3 5 . c6 �cB With only
one pair of rooks on the board Black is not
worse, as he can centralize his king without
The Prime Years
45 8
White having as much firepower with which
to attack it.
33J3ba2
Karpov gains control over the a-file.
35.g4!
Thanks to some precise calculation, Karpov
has seen that he has time to drive the bishop
away.
35 ... b3?
Black had to play 35 . . . id7, after which
36.ttJ d2 b3 37.:gb2 transposes to 34 . . . b3 as
analysed above.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
34 i.f5
.•.
This is not a mistake in itself, but it seems to
have been based on a miscalculation.
34 . . . :gb6 3 5 . ttJ d2 icS! Black must take action
against the knight transfer to b3. 36.:ga5! (The
hasty 36.ttJ b3?! can be met by 36 . . . :ga6! 37.:gb2
ttJ b7 when Black gets himself organized.)
36 ... ttJ b7 37.:gaS if5 3 s .id l White keeps a
small initiative.
34 . . . b3 3 5 .:gb2 if5 (35 . . . ia4 36.'it>e3 :geSt
37. r;t>d3 ttJ f5 [37 . . . h5 3 S . ttJ d2 :ge7 39.:gb 1
id7 40.g4 'it>g5 4 1 .h3 White wins the b3pawn and Black's kingside play is not quite
sufficient.] 3S .g4 ttJ d4 3 9 . ttJ g3 The c5-pawn
is weak.) 36.g4 id7 (36 . . . ic2 37. ttJ e3 White
wins a pawn, though winning will not be easy
due to Black's blockade on the dark squares.)
37.ttJd2 ia4 3S .'it>e3 White is somewhat
beuer, although Black has a surprising
way to create counterplay with 3S . . . 'it>g5!
39.f4t 'it>h4.
a
b
c
d
e
f
g
h
36.:ga6!
Leko probably overlooked this clever move
and only calculated 36.:gb2.
36 ...b2 37.�d2 rll e7 38.gxf5 bl=� 39.ttJxbl
:gxb1 40.f6t!
It is useful to have a pawn so close to the
promotion square.
40 ... rll d7
a
b
c
d
e
f
g
h
Anatoly Karpov - Peter Leko
2002
41 Jk6!
It was not easy to foresee that Karpov would
eventually pick up the c5-pawn with his rook
rather than his knight.
41..J�b2!
Leko goes for active counterplay based on
pinning the bishop. If Black tries the feeble
4 1 . . . tLl b7? then 42 . .tfl ! wins easily.
459
50.h4
50.l::!: g 8 'it>xf7 5 1 .l::!: xg5 tLl c 1 gives good
drawing chances.
5 0 . . . gxh4
Another line is 50 . . . h6 5 1 .hxg5 t hxg5
52.l::!: g8 'it>xf7 5 3 .l::!: xg5 tLl c 1 when it is hard
to say if Black can hold.
5 1 .d6 tLl c 1 !
With a check o n e2 coming soon, Black has
decent drawing chances.
42J�xc5 tLlf5?
Leko plays too impatiently. He could have
obtained excellent drawing chances with the
subtle move:
42 . . . l::!: a2!
Black's defence is based on the fact that it is
far from easy for White to free his rook and
break the pin.
43.l::!: c6
After 43.'it>e3 tLl f5 t 44.c.t>d3 l::!: a 3t 4 5 . 'it>d2
l::!: a2t White makes no progress.
43 . . . tLl f5 44.l::!: b 6 tLl d4 4 5 .l::!: b 7t c.t>d6
8
7
6
5
4
3
2
1
a
8
7
b
c
d
f
e
g
h
43.l::!: b 5!
6
5
4
3
2
a
b
c
d
e
f
g
h
46. 'it>e3!?
Giving up the bishop is arguably the best
winning chance. 46.l::!: e 7 is also not bad:
46 . . . c.t>c5 47.l::!: e 5 (After 47.c.t>e l tLlxe2
48.l::!: xe2 l::!: a l t 49.'it>f2 c.t>xc4 50.d6 l::!: d l
Black wins the d-pawn and has good chances
to hold.) 47 . . . tLlxe2 48.l::!: xe2 l::!: a6 It is difficult
to tell whether White's extra pawn is enough
to win the position.
46 . . . tLlxe2 47.l::!: xf7 c.t>e5 48.l::!: f8 g5! 49.f7 'it>f6
Karpov forces a rook exchange. In the
resulting endgame his pawns will be splintered,
but the sheer number of them will be too much
for the defence to bear.
In his Chess Informant analysis Karpov
mentions a second supposedly winning line,
but in fact its consequences are less than clear:
43.l::!: a 5 ! ? tLl d4 44.l::!: a7t c.t>d6 45.'it>e3! tLlxe2
46.l::!: xf7 'it>e5 47.l::!: f8
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
The Prime Years
460
47 . . . lO c l ?! 4S.f4t <;t>f5 49.f7 Elb7 (After
49 . . . Ele2t 50.�d4 <;t>f6 5 1 .c5 White's pawns
are irresistible.) 5 0.c5 Ele7t 5 1 . �d4 �f6 52.c6
White's pawns are too strong.
However, Black can improve with 47 . . . g5!
which leads to a virtually identical situation to
the analysis of 42 . . . Ela2! in the note to Black's
42nd move. In that case it is far from clear if
White can win.
43 ... a:xb5
Black has no choice bur to enter the losing
endgame. Instead after 43 . . . Ela2 44.Elb7t <;t>eS
45 .c5 lO d4 46.�e3 lOxe2 47.c6 White wins
easily.
44.cxb5 rtic7
Alternatively after 44 . . . <;t>d6 45.b6 lO h6
(45 . . . lO d4 46.ic4) 46.ia6 lO gS (46 . . . lO f5
47.ics lO d4 4S .ie6 wins.) 47.b7 <;t>c7 4S.d6t
White promotes a pawn.
Leko saw no reason to continue this hopeless
position. This game showcased a few ofKarpov's
great qualities. He showed great strategic
vision, especially when planning the knight
transfer to b3 to attack the weak c5-pawn.
Later he demonstrated his sharp calculating
ability with 36.Ela6! in the endgame.
InRound 6 Karpov was pressing against
Laurier, but the Frenchman skilfully defended
an isolated-pawn middlegame to hold the
draw. Then he suffered his first defeat against
Topalov, in a game referenced briefly on page
422, in the notes to Game 52. Karpov drew
his next game quickly with Morozevich. In the
final round he got into trouble against Bacrat.
The young French talent allowed him to escape
to a somewhat worse endgame, bur Karpov
erred and finally lost. He finished in equal fifth
place with a score of 4Y2/9.
***
45.J.d3
Karpov clears a path for his king.
45 ... �d6 46.rtie3 liJeS 47.rtid4 liJxf6
Black reclaims one pawn, but it is not enough
to affect the final result.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
4S.J.c4 liJd7 49.d6t rtixd6 50 ..ixf7
1-0
h
For the rest of the year Karpov only took part
in rapid and exhibition events. At the Dubai
World Cup he reached the quarter-final where
he lost to Kiril Georgiev.
Karpov's last ever world-class result at
an individual event came at the Eurote!
tournament in Prague. In the first round he
defeated Nigel Short on the white side of a
sharp Nimzo-Indian in a mere twenty moves,
then safely held a draw with the Petroff in the
return game.
In the second round Karpov faced his
successor Kramnik. In the first game he he!d
the black side of a Nimzo-Indian with relative
ease. We will look at the second game.
12 ... dxe4
I Game �1 1
Anatoly Karpov - Vladimir Kramnik
Prague (rap id
-
46 1
Anatoly Karp ov - Vladimir Kramnik
2002
2.2) 2002
l.d4 ttlf6 2.c4 e6 3.ttla b6
Kramnik plays the main line, which has the
safest reputation.
12 . . . dxc4 1 3 .tLlxc4 j,b7
This is a playable, though slightly riskier
alternative.
1 4 .Vlid3!?
When playing against Kasparov, Kramnik
was absolutely correct to choose openings which
demanded a subtle positional understanding.
Against Karpov, the same approach was not
so effective. The Queen's Indian seems like an
especially ill-advised choice, as Karpov played
this opening hundreds of times on both sides,
and understands its subtleties like no other
player.
4.g3 i.a6 5.b3 i.b4t 6 ..idl i.e7 7.i.g2 c6
8.i.c3 d5 9.ttlbdl
Karpov normally prefers the main line of
9.tLle5.
9... ttlbd7 10.0-0 0-0 1 1.gel c5
Later the same year Tkachiev deviated and
obtained an easy draw: 1 1 . . .j,b7 1 2.e4 dxe4
1 3 .tLlxe4 c5 1 4.tLlxf6t j,xf6 1 5 .l"i:e3 Vlic7
1 6.dxc5 Y2-Y2 Karpov - Tkachiev, Cap d'Agde
(rapid) 2002.
12.e4
a
b
c
d
e
f
g
h
c
d
e
f
g
h
This was Karpov's novelty, which brought
him a nice victory.
1 4.e5 is more common.
1 4 . . . cxd4
The next year Polugaevsky prepared an
improvement and held a draw relatively
easily: 1 4 . . . l"i:cS!? 1 5 .l"i:ad 1 b5 1 6 .j,a5 VlieS
1 7. tLl cd2 cxd4 1 S .e5 tLl c5 1 9.Vlixd4 tLl d5
20.b4 tLl a6 2 1 .a3 Vlid7 22.Vlixa7 tLl c5 23 . tLl e4
tLlxe4 Y2-Y2 Karpov - Polugaevsky, Reggio
Emilia 1 992.
1 5 .tLlxd4 tLl c5 1 6.Vlic2 a6 1 7.l"i:ad 1 Vlic7
a
a
b
1 S .j,d2!
b
c
d
e
f
g
h
462
The Prime Years
Karpov obtains an advantage with powerful
piece-play.
I B . . . tb cd7
I B . . . tbh5!? deserved consideration.
1 9.�f4 'lWc5 20.�c 1 !
After 20.�e3 'lWc7 2 1 .e5 tb g4 22.tbxb6
'lWxb6 23 . tb f5 �c5 24J''1xd7 �xe3 2 5 . tb xe3
White keeps some advantage, but Karpov's
move looks even stronger.
20 . . . 'lWc7
8
Black's posmon was already difficult, and
the double-rook-endgame was not the way
to alleviate his problems.
29 . . . :B:a5 30.�d4 keeps the advantage.
29 . . . b 5 ! ? may be the best chance: 30.E!b7
�c3 3 1 .�c5 :B:cB 32.:B:c7 :B:aB 33.f4 Black's
position is difficult, but his drawing chances
are better than in the game.
30.�xc5 bxc5 3 1 .:B:c7 :B:a3 ?
Black had t o keep his rook on the second
rank with 3 1 . . .:B:b2 or 3 1 . . .:B:e2.
7
6
5
4
3
2
a
b
c
d
e
f
g
h
2 1 .e5!
Karpov cleverly draws his opponent into a
pin and wins a pawn.
2 1 . . .tbd5 22. tb e3 'lWxc2 23.tbdxc2 :B:acB
24.�xd5! exd5 2 5 . tb xd5 �xd5 26.:B:xd5 :B:xc2
27.:B:xd7 �b4 2B.:B:ed l :B:xa2
Black has won back the pawn but White's
initiative persists.
29 .�e3
a
b
c
d
e
f
g
h
32.e6! fxe6 33.:B:dd7
1 -0 Karpov - Khalifman, Reykjavik 1 99 1 .
Despite temporarily having an extra pawn,
Black's position is hopeless and Khalifman saw
no point in playing on.
13.c!iJxe4 i.b7
Karpov faced the other main line of
1 3 . . . tb xe4 in a prior rapid game: 1 4.E!xe4
�b7 1 5 .:B:e3 �f6 1 6.dxc5 tbxc5 1 7.tbe5 �xg2
I B .�xg2 'lWc7 Black required j ust eight more
moves to secure the draw, Karpov - Anand,
Frankfurt (rapid) 1 999.
14.tLlfg5 cxd4 1 5.,ixd4 tLlxe4
a
29 . . . �c5 ?
b
c
d
e
f
g
h
Kramnik opts for immediate simplifications.
The main move is 1 5 . . . 'lWc7, which Karpov had
already used eighteen years prior to the present
game: 1 6.tLlxf6t �xf6 1 7.�xb7 'lWxb7 I B .tLle4
�xd4 1 9.'lWxd4 :B:adB 20.:B:ad l 'lWaB 2 1 .Wc3
tLl bB 22.tbf6t Yz-Y2 Kasparov - Karpov,
Moscow ( 1 B) 1 9B4. Maybe Karpov would have
2002
463
Anatoly Karpov - Vladimir Kramnik
deviated with 22.Wf3!?, although according to
the database this has only led to a long list of
draws and one defeat for White.
19.cxd5
Now Black must make a difficult decision
between a few unpleasant continuations.
16.tLlxe4 Wfc7
a
a
b
c
d
e
f
g
h
17.c!lJc3!
Karpov introduces a novelty. Objectively it
should not be especially strong, but over the
board it proved venomous enough to trouble
a great player. In the one previous game that
reached this position, Panno had used 1 7.WhS
to defeat Zarnicki in 1 992.
17 ... �ad8?
It looks like Kramnik overlooked Karpov's
next move. Instead 1 7 . . . i.xg2 1 8 .mxg2 Wc6t
brings Black close to equality, as demonstrated
in a few subsequent games.
18.tLld5!
With this small tactical stroke Karpov secures
the advantage of the bishop pair and changes
the pawn structure in his favour.
18 i.xd5
•.•
1 8 . . . exdS? is worse: 1 9 .1:"1:xe7 dxc4 (Or
19 . . .Wd6 20.We2 intending l:"1:d1 with a large
advantage.) 20.Wg4 g6 2 1 .Wh4 hS 22.i.dS !
�xdS 23.l:"1:xd7 Wxd7 24.Wf6 White will
deliver checkmate, as pointed out by Ftacnik.
19 e5
b
c
d
e
f
g
h
•..
Closing the centre is natural.
1 9 . . . �cS ? was worse: 20.�b2 (Also after
20.i.xcS tLlxcS 2 1 .b4 tLl b7 22.l:"1:c l Wb8
23.WhS Black has serious difficulties.)
20 ... exdS 2 1 .b4 i.d6 22.l:"1:c l Wb8 23.WxdS
White is close to winning.
The best chance looks to be: 1 9 . . . tLl f6!
Interestingly this move was missed by most
commentators. It leads to a position which is
certainly worse for Black, but probably still
tenable. 20.l:"1:c l W d6 2 1 .i.xf6 (After 2 1 .l:"1:c6
Wd7 22.i.xf6 �xf6 23 .Wc2 exdS 24.l:"1:c7 Wd6
Black should not lose.) 2 1 . . .i.xf6 22.dxe6 fxe6
23.Wxd6 l:"1:xd6 24.i.h3 l:"1:d2 2S .i.xe6t <j;>h8
26.l:"1:ed 1 l:"1:xa2 27.l:"1:d7 Black will probably
have to suffer for a good while, yet he has
decent drawing chances.
20.�cl Wfb8
After 20 . . . �cS ?! 2 1 .i.xcS bxcS 22.d6 Wb6
23 .WdS The d-pawn is strong and Black is
passive.
20 . . . Wd6?! also fails to solve Black's problems:
2 1 .l:"1:c6! ? (2 1 .i.b2 is also promising: 2 1 . . .Wb8
The Prime Years
464
leaves White a tempo up on the game, and
2 1 . . .W'f6?! 22J'k7! is even worse.) 2 1 . . .1.WbS
22.'lWd2 f5 23 . .tc3! The point of this move
will soon be revealed. 23 . . . .td6 24.'lWg5
tLlc5 (24 . . . a5 25 . .tfl ) 2 5 . f4 White is in a
commanding position, and since his bishop
retreated to c3 instead of b2, he does not have
to worry about the enemy knight landing
on d3 .
29.f3 White's bishop pair gives him an
ongoing advantage.
24 . . . a5 25.a3 f5 26.Elc4 'lWe7
26 . . . .txf2t? loses, as after 27.'it>xf2 tt'lc5
2S .'lWc6 tLl d3t 29.me2 tLlxb2 30.'lWxd6 Elxd6
3 1 .Elc2 Black's knight is trapped.
27.b4 axb4 2S.axb4
White executes his plan of opening the c-file,
and keeps some advantage.
2I ..tb2 i.c5!
22 ... a5 23.Elc4!
Kramnik takes away the c6-square from the
rook. 2 1 . . J'kS gives White a pleasant choice
berween 22.ElxcS ElxcS 23 . .th3 which is
strong, and 22.Elc6!? which may be even more
effective.
8
Continuing with the plan.
23 .'lWe2 'lWd6 24.b4!? was interesting though
hardly necessary: 24 . . . axb4 25 .axb4 .txb4
26.Elc6 'lWe7 27.Elal tLl f6 2S .'lWc4 .tc5 29.l"1e l
White has compensation for the pawn.
8
7
7
6
6
5
5
4
4
3
3
2
2
1
a
b
c
d
e
f
g
h
1
a
22.a3
Karpov wants to open the c-file. He succeeds,
although not without some help from his
opponent.
22.'lWe2!?
Perhaps transferring the queen to the
queens ide was even stronger.
22 . . . ElfeS 23.'lWb5 'lWd6 24.'lWa4!
This subtle queen manoeuvre supports the
plan of a3 and b4.
24.Elc4 a5 25 .'lWc6!? is also promising.
24.'lWc6 f5 25 .'lWxd6 .txd6 26 . .th3 ElfS
(26 . . . g6 27.g4!) 27.Elc6 tLl bS 2S.Elc4 tLl d7
b
c
d
e
f
g
h
23 . .f5?
.
Kramnik fails to prevent his opponent's plan,
and in the process his queen remains stuck in
a passive position.
23 ... b5 was possible, but driving away the rook
does not solve all of Black's problems: 24.l"1c2
'lWd6 25 .'lWd3 .txa3 (25 . . . l"1bS? 26 . .th3!)
26 ..txa3 'lWxa3 27.'lWxb5 White has the upper
hand on the queenside.
However, after the correct 23 . . . 'lWd6! Black's
position would have been perfectly playable:
Anatoly Karpov - Vladimir Kramnik
2002
24.Wa l (After 24J::!: e2 E!:feS 25 .b4!? [25 .Wc 1
h6] 25 . . . axb4 26.axb4 �xb4 27.E!:c6 We7
White has fair compensation for the pawn,
but Black can certainly live with it.) 24 . . . E!:feS
25.E!:dl Now after 25 . . . We7 or 25 . . . tLl f6 I do
not see a convincing way for White to improve
his position.
465
27 . . . b5 2S.E!:dl (2S .E!:a l ! ? is also promIsmg;
2S . . . tLle5 29 .Wc3 E!:cS [29 . . . E!:d7 30.E!:xd6;
29 . . . E!:feS 30.E!:aa6!] 30.E!:a5 Wb7 3 1 .E!:xb5
Wxb5 32.E!:xcS Wxd5 33 .�fl White's b-pawn
is dangerous.) 2S . . . E!:f7 (2S . . . tLl e 5 ? 29 .�xe5
�xe5 30.d6+-) 29 .�d4 White keeps some
advantage, as pointed out by Ftacnik. 29.�fl
is also promising.
24.b4 axb4 25.axb4 i.d6 26.�d2!
Karpov relieves the rook of the burden of
defending the b4-pawn.
28.i.d4
Karpov wants to tie the knight to the defence
of the b6-pawn.
26 e4
.•.
After 26 . . . b5 27.E!:c6 tLl b6 2S .�fl the b5pawn is vulnerable.
If Black sits and waits then White can steadily
improve his position: 26 . . . E!:f7 27.E!:c6 E!:ffS
2S.E!:al E!:f7 29.E!:a6 Next White will activate
his bishop with �h3 or �fl .
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
28 tiJe5?
.•.
Kramnik plays impatiently and sacrifices a
pawn without j ustification.
White remains in control after both
2S . . . �e5 29 .�e3 and 2S . . . b5 29.E!:a l .
a
b
c
d
e
f
g
h
27Jk6!
Karpov tightens the screw by finding a more
active square for the rook. He also restricts the
movements of the black queen.
2 S . . . E!:f7!? Black can also consider waiting.
29.E!:ec 1 (Also interesting is 29.E!:a l ! ? f4
30.E!:fl ! with the point that after 30 . . . e3
3 1 .fXe3 fXg3 32.hxg3 White is better.) 29 . . . E!:d8
(29 . . . f4? 30.�xe4! wins as Black cannot take
the bishop.) 30.�e3 i>hS 3 1 .E!:al White has
the initiative on the queenside.
27 J::!: de8
.•
The natural 27 . . . �e5 is strongly met by
28.d6! 'it>h8 29.E!:dl E!:f6 30.Wc 1 h6 3 1 .Wa l
when Black has problems keeping his position
together.
29Jhb6
Karpov does not refuse the gift.
29 �d8
••.
The Prime Years
466
On 29 . . . lt:l c4 Ftacnik points out that 30J''1xb 8
It:lxd2 3 1 .Elb6 Eld8 32.Eldl It:l c4 33.Elc6 wins
for White. He is right, although 30.Wg5 ! Wa7
3 1 .Elal Wfl 32.Elc6 is even more convincing.
34 . . . ElcS ?! is also insufficient: 3 5 .b6 1'!b8
36.Elb l Elb7 (36 . . . lt:l d7 37.b7 It:l c5 3S.1'!b5
It:lxb7 39.Wa7+-) 37.Wc3 'i!;fl 3S .WcS White
wins more material.
30J�xd6?!
35.gb! gb6
Perhaps Karpov was getting short of time and
wanted to simplify. He keeps a clear advantage,
but he could have won more convincingly
with 30.Ela6! Elfl 3 1 .b5, when Black is unable
to blockade the b-pawn.
Kramnik bases his defence on blocking both
passed pawns.
3o ...Wlxd6 31 .�c5 WId?
After 3 1 . . . Wf6 32.�xfS ElxfS 33.b5 Wb6
34.Elb 1 White keeps his extra pawn.
32.hf8 Elxf8 33.Wld4
The cheeky 35 . . . lt:l c6!? looks pretty but fails to
achieve much: 36.Wa l ! ? White abandons the
d-pawn in order to lend stronger support to
the b-pawn. (36.Wc4 is also possible: 36 . . . liJe5
37.Wa2 c;t>fS 3S .�h3 g6 39.�f1 Black's king
is too exposed.) 36 . . . lt:le5 (36 . . . lt:l e7 37.�fl )
37.b6 Wxd5 3S.b7 It:l d7 39.�f1 White is
winning.
Karpov improves the queen.
8
33 ...Wld6
33 . . . lt:l d3 does not achieve much, and after
34.Elb l White is ready to advance his pawns.
7
6
5
8
4
7
3
6
2
5
1
4
a
3
b
c
d
e
f
g
h
36.h3!
2
1
a
b
c
d
e
f
g
h
Having drawn both of Black's heavy pieces
into passive blockading roles, Karpov prepares
to take action on the kingside.
34.b5!
36 ... h6?!
The b-pawn is a force which cannot be
ignored, so Black will have to deploy his pieces
in defensive positions j ust to cope with it.
Kramnik creates an escape square for his
king, but weakens the light squares.
34... gb8
34 . . . lt:l d7? 3 5 . Elc 1 ! is virtually hopeless for
Black.
36 . . . h5!? was worth considering: 37.1/!Vc3
'i!;fl (37 . . . c;t>h7 3S .WcS It:l d7 39.WeS liJ f6
40.We6 Wxe6 4 1 .dxe6 It:l eS 42.f3+-) 3S .1/!Vc8
'i!;f6 39 .WaS White keeps his extra pawn and
Anatoly Karpov - Vladimir Kramnik
2002
467
keeps good winning chances, with l"1al coming
soon.
'kt>f7 42.Wxb7t �e6 43.Wc8t ttJ d7 44.b7
White has excellent winning chances.
36 . . . g6
This would also have been more resilient
than the game continuation.
37.�h l !
Ftacnik mentions 37.l"1b4 without giving any
further indication as to how White should
win the position. I do not see how White
exactly proceeds after 37 . . . �f7.
37.h4 l"1b7 38 .h5 gxh5! (38 ... �f7 39.h6 Wf6
40.Wc5 Wb6 4 1 .Wc8 l"1b8 42.Wc3 Black's
king is too open.) 39.b6 h4 40.gxh4 ttJ d7
4 1 .l"1b3 l"1xb6 42.l"1g3t 'kt>f7 43.l"1g7t 'kt>f8
44.l"1g5 Wf6 Black has chances to hold.
37.g4 l"1b7 38.gxf5 gxf5 39.l"1b3 �f7 40.b6
ttJ d7 4 1 .Wh8 is unpleasant for Black, but
the limited material gives him some chances
to draw.
37.g4!
Karpov proceeds with his kingside action
before Black can get fully prepared to meet it.
37... fxg4?
With this unfortunate move Kramnik opens
the position for the enemy bishop.
37 . . . g6 would have weakened Black's kingside
to some extent, but it was the lesser evil.
38 .l"1b3 'kt>f7 Black remains clearly worse and
will have to tread carefully to survive, yet there
is still no clear win in sight.
8
7
6
5
4
3
2
a
b
c
d
a
e
f
g
h
37 . . . l"1b8
After 37 . . . �g7? 38 .ifI �g8 39.l"1al followed
by l"1a6 White wins.
On the other hand 37 . . . ttJ d3!? 38 .ifI
Wc5 39 .Wxc5 tLlxc5 gives Black reasonable
drawing chances.
38.b6
Another possibility is 38 .ifI ttJ f3 39.Wc3 .
Now 39 . . .f4? is wrong because of 40.'kt>g2!,
but after the superior 39 ... We5 I do not
see a clear way for White to convert his
advantage.
38 . . . l"1b7 39 .Wc3 Wxd5 40.l"1b5 Wxb5 4 1 .Wc8t
b
c
d
e
f
g
h
38.Wxe4?
Karpov presents his opponent with a tactical
opportunity. To be fair, it would not have been
easy to anticipate the improvement noted at
move 39, especially in a rapid game.
The correct continuation was 38.hxg4! ttJxg4
39.Wxe4 when White wins easily.
38 gxh3!
••.
It looks like this move helps White by
bringing the bishop into the attack, but
sometimes general principles must be cast
aside due to specific tactics.
The Prime Years
468
38 . . . lLl f3t!? 39.ixf3 gxf3 would also have given
Black some practical chances, for instance:
39 . . . lLl f3t!!
White has two ways to react, but neither
is enough to win the game against accurate
defence.
a) White can take the knight.
40.'lWxf3 'lWg6t 4 1 .�h2
8
7
6
a
b
c
d
5
e
f
g
4
h
40.h4 (Other continuations include 40.1"i:b3
'lWc5 4 1 .1"i:xf3 'lWxb5 42.�h2 'lWd7, and
40.'lWe6t 'lWxe6 4 1 .dxe6 �f8 42.�h2 �e7
43.�g3 �xe6 44.�xf3; in both cases White
has good winning chances but the game is
not yet over.) 40 . . . 'lWc5 4 1 .h5 'lWf8 42.'lWe5
White should be winning, but plenty of work
remains.
39.hh3
a
b
c
d
e
f
g
h
39 ... gb8?
This is a losing mistake. Kramnik was
probably short of time (after all, this was a
rapid game!) , otherwise he would most likely
have found the following beautiful saving
move:
3
2
1
a
b
c
d
e
f
g
h
4 1 . . .1"i:f6!
Black must avoid 4 1 . . .'lWxb l ? 42.ie6t E1xe6
43.dxe6 'lWxb5 44.'lWf7t �h7 4 5 . e7 when he
loses.
42.ie6t
After 42.'lWh l 1"i:xf2t 43 .ig2 'lWd6t 44.�gl
'lWc5 4 5 . �h2 'lWd6t 46.�h3 'lWa3t White
must settle for a draw.
42 . . . �h7!
42 . . . �h8? loses to the sweet 43.1"i:c l ! , but
after the correct move Black has no problems.
b) White can also decline the knight:
40.�g2 'lWh2t 4 1 .�xf3
4 1 . �fI ?? lLl d2t wins for Black.
4 1 . . . 'lWxh3t
4 1 . . .1"i:f6t? 42.�g4 g5 43 .'lWe3 wins.
42.�e2 'lWh5t
Black can also consider 42 ... 'lWc3 when he
has reasonable drawing chances.
43 .�fI
White's king will be too exposed if it walks
to the queenside: 43 .�d3 1"i:b8 44.�c3
1"i:e8 45 .'lWd3 1"i:c8t 46.�b4 (46.�b3 1"i:c5)
46 . . . 'lWh4t 47.�b3 1"i:a8 Black has enough
counterplay for a perpetual.
Anatoly Karpov - Vladimir Kramnik
2002
469
7
The rook ending is hopeless: 43 . . . Wxe6
44.dxe6 cj;lgS 45 .b7 \t>fS 46.f4 g6 47. cj;lf2 cj;le7
4S.!l:b6 cj;ldS 49.e7t cj;lxe7 50.!l:xg6+-
5
44.!l:c1!
8
6
4
The rook is best on the c-file where it helps
to attack the king.
3
2
44 JWa6
Of course Kramnik avoids 44 . . . Wxb6
45 .!l:cSt \t>h7 46.WgSt when White wins the
rook.
•.
a
b
c
d
e
f
g
h
43 . . J'l:bS!
Black has to use the rook to harass the king.
44.b6 !l:eS! 45 .Wg2 Wf5 !
This i s the most accurate, although 45 . . . We2t
46.\t>gl Wc2 47.!l:fl Wb3 4S.d6 Wxb6
should also be drawing for Black.
46.!l:d l Wh5 47.f3 !l:e3 4S.cj;lf2 !l:b3
Black has enough counterplay.
40.i.e6t �h8
a
b
c
d
e
f
g
45J��c6! �U8
45 . . . cj;lh7 46.Wf5 t \t>hS 47.!l:xh6t gxh6
4 S .We5t wins.
46.�e7 �g8 47.b7 �b5
47 . . . Wa4 does not help, as after 4S .We6
Wd l t 49.cj;lh2 Wh5 t 5 0 .cj;lg2 Wg5 t 5 1 .cj;lfl
Black runs out of checks.
h
41.b6!
Karpov exploits the instability of Black's
knight to push his pawn a step closer to its
promotion square.
41 ...tlJd7 42.i.xd7
42.b7! lLl c5 43 .Wg6 looks even simpler.
a
b
c
d
e
f
g
h
50 ...�b4
Did Kramnik blunder, or did he play this
on the one-in-a-million chance that Karpov
might overlook that his queen was en prise?
5 1.�xb4
1-0
470
The Prime Years
In the next round Karpov faced Morozevich.
In the first game he tried the Petroff, but failed
to equalize and had to defend a depressing
major piece endgame. But the young Russian
grandmaster faltered, and Karpov emerged
with an extra pawn in a queen endgame,
which he duly converted. In the second game
he outplayed the young Russian grandmaster
and agreed to a draw.
In the semi-final Karpov faced Shirov. In
the first game Shirov played the Queen's
Gambit Accepted and ventured an interesting
exchange sacrifice. He got some initiative but
Karpov defended well and held the endgame.
In the second game Karpov used the Petroff
again. He came close to equalizing, then got
into some trouble but managed to draw the
endgame, thus forcing a blitz play-off. In the
first game Shirov built a dangerous-looking
attack against the Petroff, but Karpov was able
to neutralize it. Shirov then blundered with an
unsound piece sacrifice, and Karpov converted
his advantage easily. In the second blitz game
Karpov won convincingly, as shown in the note
to White's tenth move in Game 53, page 434.
In the final Karpov faced Anand, who
by that time had surpassed even Kasparov
when it came to rapidplay. In the first game
Karpov improved on his first game against
Shirov and came close to equalizing, but
Anand retained some nagging pressure.
Karpov tried too hastily to free himself, and
Anand was able to win a pawn, which he
eventually converted in the endgame. In the
return game Anand equalized with the Semi­
Slav, and went on to obtain a clear advantage
before agreeing a draw to guarantee first
prize.
last world-class result at an individual event. It
came more than thirty years after his first such
result, at the Alekhine Memorial in Moscow
1 97 1 .
***
Karpov played in several more rapid events
later in the year, most notably Cap d'Adge
and Corsica, both of which saw him finish as
the losing finalist, against Gelfand and Anand
respectively.
Karpov's final event of the year was a four­
game rapid match against Kasparov in New
York. In the first game Kasparov played the
Griinfeld dynamically, sacrificing a pawn for
active play. Karpov could have settled for an
equal position but instead he ambitiously
sacrificed an exchange for two pawns. The
game remained complex and both players made
mistakes, but eventually Kasparov prevailed.
In the second game Kasparov introduced a
strong novelty against the Petroff and later
obtained two extra pawns. Karpov defended
stubbornly and obtained some counterplay,
while Kasparov ran short of time and then
suffered a shocking collapse, sacrificing his
queen without good reason, then blundering a
piece when he still had good chances to defend
with a fortress.
The third game was mentioned briefly on
page 4 1 5 of the first volume. Now we will look
at it in more detail.
I Game �8 1
Anatoly Karpov - Garry Kasparov
New York, rapid (3) 2002
Despite his eventual defeat in the final, the
Eurotel result was a great achievement for
Karpov. It was a huge event where virtually all
the top players in the world were participating.
As mentioned previously, this was Karpov's
l.d4 llJf6 2.c4 g6 3.tLlc3 d5 4.tLla ig7
5.if4
Prior to this match Karpov had only played
this line once before, in Game 1 of his third
2002
Anatoly Karp ov - Garry Kasp arov
47 1
13,lUxdl gfd8 14.@f1
championship match against Kasparov.
Evidently Karpov had worked on it and
prepared it for the present match.
Karpov sensibly centralizes his king in
anticipation of the endgame.
5 ... dxc4
14 ... tlJd7?
In the aforementioned game Kasparov played
5 . . . c5 and drew comfortably. In the first game
of the present match he castled, but the same
position was soon reached.
Kasparov makes his first step on the way to
his eventual demise. Black's simplest solution
was 14 . . . Elac8, bringing his last undeveloped
piece into play. Play might continue 1 5 .'kt>e2
and now after 1 5 . . . ttJ e8 or 1 5 . . . ttJ d7 it is
unlikely that even Karpov would have been
able to squeeze anything out of White's tiny
advantage.
6.gel
This had never been played before, but
Karpov probably just wanted to confuse
his opponent slightly before returning to
mainstream theory.
1 5.i.xg7 @xg7
6... 0-0 7.e3
7.e4 is the other main move, but leads to a
more dynamic battle which would have played
into Kasparov's strengths.
7... tlJbd7
In the first game Kasparov went for a different
approach: 7 . . . �e6 8 . ttJ g5 �g4 9.f3 �c8 1 0.�xc4
c6 1 1 .Wb3 e6 1 2.ttJge4 ttJ d5 1 3 .�xd5 cxd5
14.ttJd6 ttJ c6 1 5 .ttJxb7 0 5 .0-0!?) 1 5 . . . Wh4t
16.�g3 Wh6 1 7.ttJe2 �xb7 1 8 .Wxb7 ttJa5
1 9 .Wb4 ttJ c4 20.Elxc4!? Karpov avoided
equality with this exchange sacrifice. 20 . . . dxc4
2 1 .\t>f2 Elfc8 22.Elc1 �f8 23 .Wa4 Wg5 24.Elxc4
The position is dynamically balanced, although
Black eventually triumphed after mistakes
on both sides, Karpov - Kasparov, New York
(rapid - 1 ) 2002.
8 .L:c4 c5 9.dxc5 ttJxc5 10.0-0 i.e6
•
As usual, Kasparov plays the opening
putposefully.
1 1.i.xe6 ttJxe6 12.i.e5 \Wxdl
The queenless position should be fine for
Black, although he could also have avoided it
with 1 2 . . . Wa5 (or 1 2 . . . Wb6!?) 1 3 .Wb3 ttJ c5
14.Wc4 ttJ cd7 with equality.
16 ... ttJb6!
Kasparov makes the right decision and
undertakes doubled pawns in order to relieve
the pressure.
After 1 6 . . . \t>f8? 1 7. ttJ c7 White gets a big
advantage and will soon win material.
1 6 . . . ttJ dc5!?
This is riskier than the game continuation,
but not entirely bad.
The Prime Years
472
1 7.b4
1 7.ti'Jxe7 is less convmcmg: 1 7 . . . lt'l d3
I B J'k2! It'l b4! (On I B . . . <;!;>f6 the beautiful
1 9 . 1t'l cB! causes problems.) 1 9.1"kd2 Elxd2
20.Elxd2 It'lxa2 2 1 .Eld7 ElbB 22.lt'ld5 EldB
Black should be okay in this endgame.
Had this position occurred in the game,
Karpov may well have opted for the simple
1 7.\t>e2!?, improving the king and preventing
the knight invasion on d3. In this case White
retains a pleasant and risk-free advantage.
1 7 . . . lt'l a6
18.a3 gxdIt
I B . . . b5 also fails t o solve Black's problems:
1 9 .ElxdB ElxdB 20.<;!;>e2 <;!;>f6 (20 . . . b4 2 1 .axb4
White's extra pawn may be doubled, but it still
gives him decent winning chances.) 2 1 .lt'ld2!
White plans to play It'l e4 and It'l c5 to exchange
the defensive knight. 2 1 . . .<;!;>g7 22.lt'le4 l"i:d5
23.lt'lc3 White will gain control of the d-file
soon.
19.9xdl gc8 20.<i!?e2 <i!?f6!
Kasparov makes the right decision In
activating his king.
20 . . . l"i:c2t was riskier, although it may still have
been playable: 2 1 .Eld2 Elel 22.b4! Restricting
the enemy knight. (After 22.lt'le5 It'l c5 23.b4
It'l e4 24.l"i:d7 It'l d6 25.l"i:xe7 \t>f6 26.lt'ld3 l"i:c2t
27.<;!;>d l Ela2 Black is active enough.) 22 . . . <;!;>f6
23.Eld7 Elc2t 24.\t>e l l"i:c7 25.l"i:xc7 1t is hard to
tell whether White can exploit his opponent's
doubled pawns.
a
b
c
d
e
f
g
h
I B.b5?!
Th i s appears critical, but i t does not quite
achieve the desired result.
I B .a3 maintains a slight plus for White.
I B . . . lt'l ac7
I B . . . lt'l ac5? 1 9.1t'lxe7 is a safe extra pawn for
White.
1 9.1t'lxc7 Elxd I t 20.Elxd l CfJxc7 2 1 .Eld7 ElcB
22.b6
After 22.Elxe7 \t>f6 23.Ele5 It'le6 Black has
compensation for the pawn.
22 . . . axb6 23.Elxe7 lt'l e6 24.Elxb7 Elel t 25 .<;!;>e2
It'l c5 26.Elxb6 Elc2t
Despite his two-pawn deficit, Black is active
enough to hold a draw.
17.lt'lxb6 axb6
Karpov must have been happy to obtain a
position with a slight but permanent advantage
thanks to Black's doubled pawns.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
21.llJel!
Preventing a rook invasion o n c2.
21. .. gc4
Kasparov decides to avoid a rook exchange.
It was worth considering 2 1 . . . lt'l c5 , although
White can keep some advantage with
22.Eld4!.
2002
Anatoly Karp ov - Garry Kasp arov
2 1 . . .bS!?
With this active move Black hopes to
exchange one of his weak pawns, but
achieving it is easier said than done.
22.tod3
22.l'!d7 to cs (22 . . . l'!c7 23.l'!xc7 toxc7
24.<;t>d3 <;t>e6 2 S . to f3 is similar.) 23.l'!d4
to e6 24.l'!dS (24.l'!d7 to cS) 24 . . . l'!cS 2S.l'!d7
l'!c7 26.l'!xc7 toxc7 The knight endgame is
somewhat unpleasant for Black, but he has
decent chances to hold it.
22 . . . l'!c4
22 . . . l'!c2t 23.l'!d2 l'!xd2t 24.<;t>xd2 to dS
2 S . cj;>c3 to c6 26.tocS wins a pawn.
23.l'!d2 to cs 24.toxcS l'!xcS 2S .b4 l'!gS
2S . . . l'!c3 26.l'!d7 wins a pawn.
2S . . . l'!hS 26.g4! is strong.
26.g3 <;t>e6 27.e4
White continues to press.
473
was evidently not satisfied by this type of
endgame.
Another idea is to block the seventh rank
with 22 . . . ltJ c7 and then chase the rook away:
23.ltJd3 <;t>e6 24.l'!dS! ltJ dS 2S .l'!hS (2S .<;t>d2
fS 26. ltJ c l ltJ f6 27.ltJ e2 ltJ e4t 2S.cj;>d3 also
gives White some advantage.) 2S . . . hS 26.cj;>d2
White remains on top, but his advantage is
smaller than in the game.
23.E:c7!
Karpov creates an unpleasant pin and
threatens to win material with b4. Black
cannot avoid losing a pawn.
23 ... b5
22J�d7
a
b
c
d
e
f
g
h
24.£3!
a
b
c
d
e
f
g
h
Before winning a pawn Karpov takes away
the e4-square from the black knight.
22 ... tlJc5?
24 ... e5
True to his style, Kasparov opts for active
defence, but the position does not justifY it.
Instead he should have gone for one of the
following alternatives:
24 . . . ltJ a4 does not help: 2S.l'!xb7 l'!cS
(2S . . . ltJxb2 26.l'!b6t e6 27.l'!xbS ltJ a4
2S .l'!b4 After exchanging rooks White wins
comfortably.) 26.b4 l'!c6 27.e4 gS 2S.<;t>e3
l'!c3t 29.cj;>d2 White keeps the extra pawn and
a winning position.
22 . . . l'!c7 23.l'!xc7 toxc7 24. ltJ f3 <;t>fS 2 S . <;t>d3
White's advantage is unquestionable, although
it is hard to say if he can win. Kasparov
25.M
The Prime Years
474
Karpov sees no further improving moves, so
he collects the b7-pawn.
White is a healthy pawn up, and should win
without too many problems.
25 ... �a4 26.E!xb7
27.e4 i>e6
Karpov eats the first b-pawn, and the second
one could soon be on the menu.
26 ... E!c6
26 . . . lLl c3t!?
This move contains more poison than one
might expect.
27.md3 1"i:c8
8
7
6
5
4
3
a
2
b
c
d
e
f
g
h
28.h4?!
L " . j·,",,"'/·
a
b
c
d
e
f
g
h
28.e4!
2 8 . lLl c2 lLl b l ! gives Black counterplay.
2 8 .1"i:b6t!? is reasonable: 28 . . . <j;Jg7 29.e4 1"i:c7
30.g3! This subtle move prepares f4 to vacate
the f3-square for the knight. (30.lLlc2 lLl b l !)
30 . . . h5 (30 . . . g5 ? weakens the f5-square:
3 1 . lLl c2! lLl b l 32.lLl e3+-) 3 l .f4 White has
decent winning chances. (But not 3 1 .lLl g2?
due to 3 1 . . . lLl b l !.)
28 . . . <j;Je6 29 .f4!?
29.g3 is also good enough: 29 ... mf6
(29 . . . h5?! 30.1"i:b6t me7 3 l .f4+-) 30.lLlc2
lLl b l 3 1 . lLl e3 White succeeds in improving
his knight, which should enable him to win
the game.
29 . . . lLl b l
2 9 . . . exf4 30.lLlf3 gives White excellent
winning chances.
30.1"i:xb5 1"i:c3t 3 1 .me2 exf4 32.lLld3 1"i:c2t
33.<j;Jf3 lLlxa3 34.lLl xf4t
Karpov anticipates the arrival of the enemy
rook on the second rank, and removes a
potentially vulnerable pawn. Nevertheless he
should have taken the opportunity to improve
his knight first with 28.lLld3!.
28 ...h5?
Kasparov fails ro take the opportunity with
which he was presented.
28 . . . lLl c3t! 29.me3 1"i:a6
29 . . . lLl d l t is tricky, but not quite sufficient
against accurate play: 30. md3 1"i:d6t 3 1 . <j;Jc2!
lLl e3t 32. <j;Jc3! lLldl t 33.<j;Jb3 White has
utilized an unusual triangulation to avoid the
checks. 33 . . . lLl e3 34.1"i:xb5 1"i:d l 3 5 . lLl c2+30.lLld3 1"i:xa3 3 1 .1"i:b6t <j;Je7 32.<j;Jd2 lLl b I t
3 2 . . . f6 33 .1"i:b7t wins.
33.<j;Jc2 1"i:c3t 34.mxb l 1"i:xd3 3 5 . 1"i:xb5 <j;Je6
36. <j;Jc2 1"i:e3
White keeps some winning chances with
his extra pawn, but in a rook ending this is far
from a guarantee of victory.
29.tiJd3!
Karpov corrects the previous error and
activates his knight.
475
Anatoly Karp ov - Garry Kas p arov
2002
36.gb6t �e7
29 .. Jk2t
After 29 . . . f6 30.<;t>d2 E:c3 3 1 .E:xb5 E:xa3
32.E:a5 ! the pin decides the game.
30.�e3 f5
Taking the g2-pawn is too slow: 30 . . . E:xg2
3 1 . E:xb 5 f6 (3 1 . . .E:e2t 32.i>xe2 lLl c3t 33.i>e3
lLlxb5 34.a4+-) 32.E:a5 lLl c3 33.lLlc5t i>e7
34.E:a7t <;t>d8 3 5 .a4 White wins easily thanks
to his two connected passed pawns.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
37. c!tJ xe5
Karpov wins a second pawn, and the outcome
of the game is decided.
37 ... ge2t
37 . . . g5 also fails to save the game: 38.i>f1
E:g3 39.lLlg6t <;t>d8 40.e5 gxh4 4 1 . lLlxh4+a
b
c
d
e
f
g
h
31.g3!
Karpov finds the simplest way to neutralize
Black's activity.
38.�fI gxe4 39.c!tJxg6t �f7 40.�fl c!tJ d5
After 40 . . . lLl b l the finish might have been
4 1 .a4 lLl d2 42.a5 lLl c4 43.E:c6 lLl a3 44.a6
E:xb4 45.a7 E:a4 46.lLle5t i>g7 47.E:c7t i>h6
48 .E:b7 and White wins.
31. .. fxe4 32.fxe4 gg2 33.gxb5
Karpov's plan prevails, and he obtains two
connected passed pawn. 33. <;t>f3?! fails to make
any progress after 33 . . . E:d2.
41 .gc6
33 ... gxg3t 34.�d2 gg2t 35.�el!
Karpov finds the right way to escape the
checks.
35 ... c!tJc3
At long last the black knight gets a say in
the proceedings, but it is too late to save the
game.
35 . . . E:g4 36.E:xe5t <;t>d6 37.E:d5t i>c7 3 8 . lLl f2
E:xh4 39 .E:a5 is winning for White.
a
b
c
d
e
f
g
h
476
The Prime Years
41. tlJxb4?!
.•
Sacrificing the knight is a desperate try. 4 1 . . .ttJ e7! was the best chance, as White must play
accurately to win the rook ending: 42.ttJxe7 (There is also 42.1"k4!? 1"!:xc4 43.ttJe5t 'ifie6 44.tLlxc4
tLl f5 45.a4 when White should be winning.) 42 . . . 'ifixe7 43.'ifig3 1"!:e3t 44.'ifif4 1"!:xa3 45.'ifig5 1"!:d3
46.1"!:c5 (46.'ifixh 5 ? 1"!:d5t! enables Black to escape.) 46 . . . 1"!:d4 47.1"!:b5 'ifid6 48.'ifixh5 'ific6 49.1"!:c5t
'ifid6 50.1"!:b5 'ific6 5 1 .1"!:b8 'ific7
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
52.1"!:h8!! The only way to win. 52 . . . 1"!:xb4 53.'ifig6+-
42.axb4 1"!:xb4 43.tlJe5t!
Karpov ensures that his last pawn remains on the board.
43 ... cJig7 44.1"!:c4 1"!:b5 45.tlJd3 cJif6 46.1"!:c5
1-0
In the final game Karpov made a mistake in the Petroff, but Kasparov missed a relatively easy
win. Karpov then got an edge but agreed a draw to win the match. Karpov had a slice of good
fonune, but deserved his success, and did well to survive the games where Kasparov dominated
the opening. The world number one was clearly not at his best; perhaps he had not fully regained
his confidence after his defeat to Kramnik two years earlier.
2002 Summary
Belgian Team Championship: I I I (+ 1 =0 -0)
Nao Masters, Cannes (3rd-7th place) : 4Yzl9 (+2 = 5 -2)
Total 5 5 .0% (+3 = 5 -2)
g Wins • Draws • Losses
2003
Rating 2688 (29 i n the world)
Karpov's first event of the year was Wijk aan Zee. In the first round he equalized against Krasenkow,
then sacrificed an exchange in the middlegame. The Russian-born Polish grandmaster went wrong
and soon had to resign. In the next two games he drew with Shirov and Grischuk; in both games
White had slight pressure but never any serious advantage.
In Round 4 Karpov made some mistakes against Judit Polgar and allowed his king to become
stuck in the centre. He never recovered, and despite stubborn defence he eventually succumbed.
In the next game Karpov sacrificed a piece for three pawns on the black side of a Caro-Kann
against Anand, but his activity slowly disappeared and he eventually went down. He was also
in trouble in the next round against Radjabov, but managed to simplifY to an endgame with
rook versus rook and bishop, which he managed to hold, thus avoiding a third successive
defeat.
In Round 7 Karpov held a draw against Topalov; he was always a bit worse but never in real
trouble. Then he defeated Van Wely, in a game referenced in the note to White's sixth move in
Game 56. He drew his next three games fairly solidly against Timman, Bareev and Ivanchuk. In
Round 12 he made a mistake in a worse position against Ponomariov, which allowed the recently­
crowned FIDE World Champion to win a pawn and later the game. In the final round Karpov
took a quick draw with Kramnik to finish on 6/ 1 3, which left him in equal eleventh place.
***
Karpov's next event was a rapid match against the future European Champion Bartlomiej Macieja.
He defeated the Polish grandmaster by a convincing score of 6-2, with five wins, two draws and
one loss.
Karpov's next significant event was the Spanish Team Championship, where he played six games.
He started with a good win over Korneev, then took a quick draw with Campos Moreno.
In his third game Karpov faced Mihail Marin. This was the Romanian grandmaster's only
encounter with Karpov, but he drew his other three games against world champions.
I Game �9 1
Mihail Marin
-
Anatoly Karpov
Lanzarore 2003
l.c4 ltJf6 2.ltJc3 e5 3.ltJf3 ltJc6 4.g3 �c5
The Prime Years
478
In the first volume of his Grandmaster
Repertoire series on the English Opening,
Marin refers to Black's set-up as the Karpov
Variation. The twelfth World Champion has
certainly been one of the most prominent
exponents of this system, although he has
chosen 4 . . . .tb4 more frequently.
Another possibility is 1 4.d4, after which
Black should play 1 4 . . J'hd8 .
14 lLld8
•.•
Karpov improves the knight and clears the
way for the c7 -pawn.
1 5.llJge2
5 ..ig2 d6 6.0-0 0-0 7.<13 h6 8.a3 a5
In his two previous games Karpov pushed
the a-pawn only one square.
9.e3 �e8
Karpov remains flexible and delays
committing his light-squared bishop for
another move.
10.b3 .if5 1 l .ib2 ia7
Now that White's bishop has gone to b2,
Karpov no longer needs his bishop on c5 to
prevent b4. Therefore he retreats it so that d3d4 will not come with gain of tempo.
a
b
c
d
e
f
g
h
15 c6!
12.h3 V;Vd7 13.cj;1h2 ih7
.•.
Karpov makes a precautionary retreat of the
other bishop as well.
8
Karpov blocks the long diagonal, thus
enabling his knight to move, while also
preparing to gain space in the centre with
. . . d5.
16.e4
7
Preventing the central expansion.
6
5
16 lLle6 17.£4
.•.
4
8
3
7
2
6
1
a
14.lLlgl!
b
c
d
e
f
g
h
Marin brings his knight to a more flexible
location and prepares to push his e- and
f-pawns. In a previous game Ehlvest preferred
1 4. ltl h4 and drew against Renet.
5
4
3
2
1
a
b
c
d
e
f
g
h
2003
Mihail Marin - Anatoly Karp ov
17 b5!?
•••
Since . . . d5 has been prevented, Karpov
chooses another way of gaining space, which is
often important in positions where all or most
of the pawns remain on the board.
The decision to play the text move required
considerable bravery, as Black is inviting his
opponent to make a dangerous exchange
sacrifice for very real compensation. In order
to judge it correctly, Karpov had to evaluate
several complex variations far in advance.
Black had a couple of reasonable alternatives
in 1 7 . . . ttJ d4 1 B .ttJxd4 i.xd4 and 1 7 . . . exf4
1 B .ttJxf4 ttJxf4 1 9.1'!xf4 i.d4, both of which
aim to make use of the recently-weakened d4square.
18.cxb5 cxb5 19.fxe5
1 9 .Wd2 b4 20.ttJa4 Wb5 is unpleasant for
White.
In Chapter 4 of the aforementioned book,
Marin recommends 1 9.b4! as the right way for
White to proceed. The Romanian grandmaster
gives some nice analysis, which is summarized
here: 1 9 . . . ttJ d4?! This allows White to
showcase his idea. Black should probably
look for an improvement here, although he
must constantly be on the lookout for the
l'!xf6 sacrifice. 20.fxe5! dxe5 2 1 .l'!xf6! This
is an improved version of the sacrifice seen
in the game. 2 1 . . .ltJxe2 (2 1 . . .gxf6 22.ltJd5)
22.'Wxe2 gxf6 23.ltJd5 'WdB (23 ... l'!e6 24.l'!f1
'Wd6 25 .'Wh5) 24.l'!f1 l'!e6 25 .'Wg4t (25 .h4!?)
25 . . . i.g6 (25 . . . �fB? 26.i.xe5!+-) 26.'Wh4
�g7 27.i.c l h5 2B.g4 hxg4 29.i.h6t �gB
30.hxg4± Black has serious problems in all
these lines, as pointed out in Marin's book. It
would have been interesting to see how Karpov
would have reacted had 1 9.b4 appeared on the
board.
19 dxe5
•..
479
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
20Jhf6
Marin cannot resist the sacrifice. In return
for the material investment he fractures Black's
pawn structure, while obtaining the d5-square
for his knight and attacking chances on the
kingside. Furthermore, he has chances to
bury the h7-bishop, j ust as in the classic game
Winter - Capablanca, Hastings 1 9 1 9 .
20 ... gxf6 21.tiJd5 Wid8
Black must defend the f6-pawn.
22.'Wdl
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
22 ... tiJc7
Karpov exchanges the opponent's strong
knight and makes room for his rook to defend
the f6-pawn from the adjacent square.
The Prime Years
480
22 . . . lLl c5 might have led to a repetition
after 23 .Wxh6 Ele6 24.Eln f5 25 .Wh5 .ig6
26.Wh6 .ih7. Black cannot play for a win
with 26 . . . lLlxb3? on account of 27.lLl ef4 WfB
2B .Wh4 when White's attack is too strong.
23.llJef4
25 . . . ElcB 26.Eln Elc2? (26 . . . Elcc6 is better,
when 27.Wd2 reaches an unclear but roughly
balanced position.) 27.Elxf6 WcB (27 . . . Elxb2
2B.Elxe6+-) 2B .Wg5 t 'i:t>fB 29.h4 Elxb2
30.Wh6t 'i:t>eB 3 1 .Wxh7 White is winning.
26.Eln Elaa6
23.lLlxc7 Wxc7 24.Wxh6? allows 24 . . . Wc2
winning a piece.
23 ... llJxd5 24.llJxd5 ge6
8
7
6
5
a
4
3
2
1
a
b
c
d
e
f
g
h
25.g4
Marin buries the bishop a la Capablanca,
but the cage is less secure than it was in the
aforementioned game.
25 .Eln ! ? sets a small trap, as 25 ... 'i:t>g7?? loses to
26.Elxf6!. However, the position would remain
balanced after 25 . . . .ic5 , 2 5 . . ..ig6 or 25 . . . ElcB .
25 .Wxh6!
Taking the pawn was the most challenging
continuation.
25 . . . .ic5 !
For the moment Black should organize his
defence. Attempts to wrest the initiative
could easily backfire.
25 . . .f5 ? ! 26.Wd2 ElcB 27.Eln .ibB 2B .Elxf5
.ixf5 29.exf5 Elec6 30.f6 Elxf6 3 1 .Wg5t Elg6
32.lLle7t \t>g7 33 . .ixe5t White is somewhat
better.
b
c
d
e
f
g
h
27.h4
27.Wd2 WfB and 27.Wh5 .ig6 2B .We2 �h8
are both okay for Black.
27 . . . b4
27 . . . f5 ?! Paradoxically, by exchanging one of
his doubled pawns, Black weakens his pawn
structure. 2B .Wh5 Elh6 29.Wf3 White has
nice play after 29 . . .f4 30 . .ixe5 or 29 . . . 1d6
30 . .ih3 .
2B.a4 Elac6 29 . .ih3 .ifB
29 . . . Eled6 is also playable.
30.Wd2 Eled6
The position is balanced.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
Mihail Marin - Anatoly Karpov
2003
25 .. Jks
48 1
There are no immediate threats, so Karpov
sensibly improves his rook.
After defending for a while Karpov makes
an attacking move and forces the opening of
another queens ide file for his rooks.
26.gfl
30.b4?
26.i.c3 can be met by 26 . . . a4.
26 ...i.c5 27.i.c3!
Marin finds a good way to combine play on
both flanks.
27.. J3aS!
Karpov finds the right reply, even if it means
losing time. 27 . . . a4 was possible, although
after 28.bxa4 bxa4 29.i.a5 Black must give up
the f6-pawn.
Marin chooses the wrong pawn to advance.
30.a4! bxa4 3 1 .bxa4 would have minimised
White's disadvantage. The pawn on a4 has less
protection than the b4-pawn in the game, but
more important is the fact that Black will have
a harder time utilizing the open b-file, as the
b6-square is not available. Furthermore, in this
line Black must keep an eye on the vulnerable
a5-pawn.
30 ... axb4 3 1 .axb4 i.d4 32.We2
27 . . . i.xa3 ? would have been a serious mistake,
and after 28 .i.xa5 Wd6 29.b4 the bishop is
in danger. 29 . . . Wc6 30.Wxh6 i.g6 3 1 .:Sxf6
In a few short moves White has obtained two
pawns and good attacking chances.
6
2s.'IWb2?!
5
It was not too late to play 28.Wxh6 when the
position would have remained balanced.
8
7
4
3
2
1
a
b
c
d
e
f
g
h
32 ... ged6!
a
b
c
d
e
f
g
h
2S ... gaa6!
Karpov overprotects the f6-pawn, freeing his
queen for other duties.
29.i.el WflU
Karpov's strategy is to threaten to sacrifice
back the exchange on d5 under favourable
circumstances.
It was also possible to begin with 32 . . . Wd8!,
which is likely to lead to the same position
as the game continuation in slightly fewer
moves.
33.tlJc7 gab6 34.tlJd5 gbc6 35.i.h4 WdS
36.Wa2 <i>g7
Karpov uses the king to defend the f6pawn, in order to give his heavy pieces more
freedom.
The Prime Years
482
37.�e2
39 . . . '.Wc8 40.'.Wf3 '.Wc2? squanders Black's
advantage: 4 1 ..ixf6t Wg8 42.'Llelt cj;lh7
43.g5 White is in the game.
39 .. J%dc6 is also strong, but allows White
more options: 40.'.Wf3 '.Wb8! 4 1 ..ixf6t
cj;lh7 42.g5 ! ? (42.'.Wg3 :ga2 43 .Whl :gcc2
transposes to the main line below, beginning
with 39 . . . '.Wb8!.) 42 . . . h5 43 . .ie7 :ge2
44. 'Ll f6t Wh8 45.'Llxh5
8
7
6
5
4
3
8
2
7
1
6
a
b
c
d
f
e
g
h
37.. J�a6!
Karpov improves the prospects of his rooks.
In certain lines he may wish to force an
exchange on the a I -square, while in others he
may be able to invade with one or even both
rooks on the second rank.
38.�f3?
This turns out to be an unfortunate choice.
38.'\Wd2!
White's best chance is to wait and challenge
his opponent to make progress.
38 . . . .ig6 39.'\We2
5
4
3
2
a
b
c
d
e
f
g
h
45 . . . '.Wc7!!
This
subtle
move
wins
convincingly. (45 . . . :gaa2?? allows mate in
two.) 46 ..if6t White saves his bishop, but
by blocking the f6-square he also saves
Black's king. (46 . .ic5 .ixc5 47.bxc5 ixh5
48 .'.Wxh5t cj;lg8-+) 46 . . . Wg8 Black will soon
play . . . :gaal with decisive effect.
40.Wh l
40.'.Wf3 ? :gxd5 wins, as does 40.'.Wd2?! Wia7
4 1 .'.We2 :ga2.
40 ... :gdc6 4 1 .'.Wf3 :gc2 42 . .ixf6t cj;lh7
8
7
6
5
4
a
b
c
d
e
f
g
h
39 . . . '.Wb8!
This seems to be the most precise
continuation. Black wants to play . . . '.Wa7
and invade on the second rank.
3
2
a
b
c
d
e
f
g
h
Mihail Marin
2003
43 .Wg3
This seems to be the only way to prolong the
game.
43 . . .�aa2 44.if3 Wd6 45.h4!
Instead of waiting for the axe to fall, White
plays to trap the g6-bishop.
45 . . .�f2! 46.�xf2 �xf2 47.h5
-
Anatoly Karpov
483
38 ... gxd5!
Eighteen moves after Marin sacrificed the
exchange, Karpov returns the material with
decisive effect. The knight was a powerful piece
which not only attacked f6, but also defended
the b4-pawn.
Although Karpov's move is the most effective
route to victory, it is worth mentioning that
38 . . . �dc6 was playable, with the eventual
idea of penetrating to the second rank, for
instance: 39.lt>h l ig6 40.lt>h2 It>h7 4 1 .ixf6
(4 1 .ttJxf6t? Wh8 42.Wg3 Wc8-+) 4 1 . . .Wb8
42.Wg3 �a2 43.Wh l �cc2 44.if3 �d2 Black
has decent winning chances.
39.exd5 �d6!
a
b
c
d
e
f
g
h
47 . . . Wa6!
Threatening a deadly check on a l .
48 .ig2
The immediate 48.hxg6t fxg6 changes
nothing.
48 . . . �xf6 49.ttJxf6t Wxf6 50.hxg6t
White can delay this move, but it is unlikely
to make much difference.
50 . . . fxg6
Despite the opposite-coloured bishops
and (temporary) material equality, Black's
positional advantages add up to serious
winning chances.
This is part of the plan which culminates in
Black's next beautiful move.
40.c;!;>hl
40.Wg3 was no better: 40 ... Wxb4 4 1 .ie4
ixe4 42.dxe4 Wd2t 43 .Wg2 (43 .lt>h l We2
44.�e l �al-+) 43 . . . Wd3 44.Wf3 Wxf3 4 5 .�xf3
b4 Black is winning.
8
7
6
5
4
8
3
7
2
6
1
5
a
4
b
c
d
e
f
g
h
40 ....ig6!!
3
With this move Karpov not only defends f7
and creates an escape square for his king on
h7 j ust in case, but also, quite incredibly, puts
his opponent in zugzwang! It is a remarkable
2
1
a
b
c
d
e
f
g
h
The Prime Years
484
achievement with so many pieces on the
board.
4U�cl
Any other reasonable move releases the
pressure on the f6-pawn, allowing Black to take
on b4. If White plays 4 1 .�b l then 4 L .�a3
wins the d3-pawn.
45 ...�c5 46..1£1 b4 47 ..1d3
Marin places his bishop on a square where it
controls the promotion square of the b-pawn,
while also defending the e4-pawn, which helps
to keep the g6-bishop in its cage.
47...�xc7 48.<i!lg2
41 ...�xb4 42.�c6
Marin sensibly tries to stir up problems.
42 .. J:�xc6
Karpov plays in his usual pragmatic style
and simplifies to a winning endgame. He
could also have won with an attack on the
enemy king: 42 . . . �a l t 43.i.fl (43.'kt>h2 i.g l t
44.'kt>h l i.f2t-+) 43 . . .Wb l 44.'kt>g2 (44.�xf6
e4 45 .Wf4 Wxd3-+) 44 . . . e4 4 5 .i.xf6t
a
b
c
d
e
f
g
h
48 ... b3?!
Pushing the pawn is natural, but it allows
White an opportunity to organize his pieces in
a more resilient defensive formation.
a
b
c
d
e
f
g
h
45 . . . ..t>h7! This calm move seals Black's victory.
(45 . . . i.xf6?? 46.Wxf6t ..t>h7 47.�c8 even loses
for Black.) 46.Wf4 �a2t 47.'kt>h l �f2 White
can resign.
43.dxc6 e4!
Karpov buries his own bishop, but the
benefits make it worthwhile. Now his dark­
squared bishop will find a great home on e5,
and the g2-bishop will be confined to a purely
defensive role.
44.dxe4 i.e5 45.c7
45 .Wd3 Wc5 wins.
48 . . . Wc 1 !
Marin demonstrated that the immediate
queen invasion was more accurate.
49 .i.fl
49.Wfl We3! (49 . . . Wd2t also looks good
enough: 50 . ..t>f3 [50.i.f2 Wf4-+] 50 . . h5
5 1 .i.g3 b3 52.i.b l b2 53 .i.d3 Wc 1-+)
50.Wf3 Wd2t Black wins, as pointed out by
Marin.
49 . . . b3 50.Wxb3 i.xe4t
The attack is irresistible.
5 1 ...t>gl
5 1 ...t>f2 Wd2t 52.i.e2 i.d4t 53.'kt>fl i.g2t!
leads to mate.
5 L .Wc5t 52.i.f2 i.h2t!
Black checkmates, as Marin pointed out in
his book.
.
Mihail Marin - Anatoly Karpov
2003
48 5
The point of White's idea is that the bishop
and queen provide a firm blockade of the
b-pawn, while also securely defending the e4pawn. Ultimately I suspect that Black is still
winning, but even so, it is a pity that Marin did
not choose this path, which would have tested
his opponent's legendary endgame technique.
Black has two main options:
8
7
6
5
4
3
2
1
a
b
49.i.fl?
c
d
f
e
h
g
White misses his opportunity. 49.Wd l ? is
also feeble: 49 . . . b2 50.j,b l Wd 5 l .Wd3 h5
52.j,f2 hxg4 53.hxg4 Wf4-+
When I asked Mihail about this game, he
mentioned that he could have obtained
reasonable drawing chances with the help of a
wonderful fortress idea. He told me he saw the
concept during the game but for some reason
decide� not to try it. He was kind enough to
send me his analysis, which I have reproduced
below, with some additions of my own:
49 .j,b l !
This is the reason why the black queen should
have gone to d on the previous move.
49 . . . b2
49 . . .1Mi'c2t is beautiful and imaginative, but
not sufficient: 50.j,xc2 bxc2 5 l .1Mi'a3 j,xe4t
(5 l . . .j,f4? 52.j,xf6t! wins.) 52.mf2 j,f4
53.1Mi'e7 c 1 =1Mi' 54.1Mi'xe4 White holds the draw.
50.Wd3 Wd 5 l .j,f2
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
a) 5 1 . . .Wf4
Marin did not consider this move, but it
looks reasonable. Black is playing with his
pieces before executing a pawn break.
52.Wf3 Wh2t 5 3 .'tt> fl j,d6 54.j,d4
White could consider 54.h4!?, although his
position is difficult here too.
54 . . . j,e7
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
5 5 .j,c3
Another line is: 5 5 .'tt> e l mg8 56.'tt> fl (56.j,c3
Wg l t 57.Wfl We3t-+; 56.'tt> d l j,b4 57.j,f2
Wc7 5 8 .We3 Wc4 5 9.Wd3 Wd t 60.me2
h5 Black should be winning.) 56 . . . Wd2
57.Wd3 Wd t 5 8 .'tt> e2 h5 It is hard to give
an absolute verdict, but I doubt that White
can hold this position.
55 . . . j,d8 56.j,d4
56.j,d3!? 'tt> g8 57.j,b l j,e7 5 8 .j,d4 (58 .j,d3
j,c5) 58 . . . Wd2 5 9.Wd3 Wd t 60.mg2 j,c5
6 l .j,xc5 Wxc5 White's position remains
critical, but he may be able to survive.
56 . . . Wd2 57.Wd3 Wd t 5 8 .me2 j,c7 59.Wd l
5 9 .j,f2 h5! creates fresh problems for the
defence.
486
The Prime Years
59 .i.e3 'Mfh l 60.i.d4 i.e5 6 1 .i.f2 h5! should
also be winning for Black.
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
5 9 . . . h5!
Black has manoeuvred for long enough, and
now it is time to break down the enemy
position.
60.i.e3 i.f4!
It is hard for White to keep his position
together, for instance:
6 1 .i.d4 i.g5 62.'Mfd3 hxg4 63.hxg4 'Mff4
64.�e1 'Mfxg4 65 .i.xb2 'Mfg l t
Black is winning.
b) 5 1 . . . h5
This is the most natural move, and indeed
it is the only one that Marin analyses. Once
again White can resist for a while, but Black
keeps excellent winning chances as we will
see.
8
This is White's strongest defence.
52.i.g l hxg4 53.hxg4 'Mff4 54.'Mff3 'Mfd2t
5 5 .i.f2 i.d4 56.c;t>g3 i.xf2t 57.'Mfxf2 Wfc l
5 8 .'Mfc2 i.h7 (58 . . . �h6 5 9.'Mfh2t! sends
the king back.) 5 9.�d3 c;t>h6 60.'Mfd4 c;t>g5
6 1 .'Mfd5t c;t>g6 62.'Mfd3 c;t>g7 63 .'Mfc2 i.g6
64.�d3 �h6 Black should be winning this
endgame.
52.i.c2 'Mff4 5 3 .'Mff3 'Mfh2t 54.�f1 h4! Black
can target the h3-pawn. 5 5 .i.b 1 i.g3 56.Wfg2
(56.i.xg3 hxg3! 57.'Mfg2 �h6! 5 8 .'Mfxh2 gxh2
5 9.�g2 �g5 60.�xh2 �f4-+) 56 . . . 'Mfxg2t
57.�xg2 i.xf2 5 8 .c;t>xf2 �h6 59.c;t>e3 c;t>g5
60.�f3
8
7
6
5
4
3
2
a
b
c
d
e
f
g
h
60 . . . i.h7! It turns out that Black's bishop is
not completely buried. 6 1 . �e3 i.g8 62. c;t>f3
f5 ! 63.gxf5 (After 63.exf5 f6 64. c;t>e3 id5
White is in zugzwang.) 63 . . . f6 64.�e3 if7!
(64 . . .i.c4 65.c;t>d4 i.f1 66.e5 i.xh3 67.e6
i.xf5 68.i.xf5 �xf5 69 .e7 b l ='Mf also wins.)
65.�f3 i.c4 Black is winning.
7
6
5
4
3
2
a
52.i.e3!
b
c
d
e
f
g
h
a
b
c
d
e
f
g
h
Mihail Marin
2003
52 . . . hxg4
This time it is better to create a weakness on
g4 than on h3.
52 . . . h4? 53 . .tf2! (53 . .tc2 .tb8! 54 . .tb l Wic7
is unpleasant.) 53 . . . Wif4 54 . .txh4 Wih2t
5 5 .i>fl .txe4 56.Wixe4 Wixh3t 57. We2 Wixh4
Black's winning chances are minimal.
53.hxg4 .tf4!
53 . . . .tb8 54 . .tc2 defends.
54 . .td4
-
Anatoly Karp ov
Wig 1 t 60.Wd2 Wig5t 6 1 .Wc3 .te5 t 62.i>b3
.txb2 63.Wxb2 Wie5t 64.i>cl f5 Black is
winning.
5 5 . . . Wif4
After 5 5 . . . Wixb l ?? 56 . .txf6t Black is mated.
56.Wixb2 .txe4t 57 ..txe4 Wixe4t
a
a
b
c
d
e
f
g
h
54 . . . .tb8!!
The point of this move is to prepare a queen
and bishop battery.
5 5 .Wic3
55 . .te3 Wic7! 56 . .tgl Wif4 is winning.
55 ..tc2 Wif4! Black can afford to give up his
advanced passed pawn to go after the enemy
king. 56 . .txb2 Wixg4t 57.i>fl Wif4t
487
b
c
d
e
f
g
h
5 8 .Wfl
5 8 .i>f2 Wif4t does not help White.
58 . . . .te5
58 . . . Wif4t should also be good enough.
5 9 ..txe5 fxe5
The queen ending looks to be winning.
a
b
c
d
e
f
g
h
49 VNc1!
...
a
b
c
d
e
f
g
h
58.i>el (58.Wg2 Wih2t 59.i>fl .th5 White's
king is in serious trouble.) 58 . . . Wih2 59 .Wie2
I t would b e interesting to know i f Karpov
noticed the aforementioned defensive idea and
deliberately prevented it, or if he merely chose
this move based on general principles. Either
way, it is good enough to win the game.
488
The Prime Years
57.�hl i.g3t 58.�gl hflt 59.�xf2 Wlh2t
60.�el Wlglt 61.Wlfl Wlxg4
50.Wle3 b2 5 1 .Wlel
Simplest, although 6 1 . . .Wfe3t 62.Wfe2 '!Wc l t
63.Wfd l cj;>h6 wins as well.
62.Wld3 Wlglt 63.�d2 Wlclt 64.�e2 .ih5t
65.�f2 �h6!
Karpov improves his king before he exchanges
queens.
66.�g2
a
b
c
d
f
e
g
h
5 1 ... h5!
Karpov starts his campaign to free the
bishop.
52.i.bl hxg4 53.hxg4
The cage is still standing, but its foundations
have been weakened.
53 ... Wlf4 54.Wle2 Wlh2t 55.�fl Wlh3t
56.�gl
56. cj;>e 1 allows a cute finish: 56 . . . Wfh 1 t
57.WffI
8
With a two-pawn deficit against arguably the
greatest endgame player of all time, White has
virtually no chance of escaping. In his excellent
Learn from the Legends book, the Romanian
grandmaster devotes a chapter to Karpov. He
fully understands how great a player Karpov
is, and perhaps he was simply enjoying the
experience of playing him and wanted to
prolong the game as long as possible.
8
7
6
5
4
3
7
2
6
1
5
a
4
3
b
c
d
e
f
g
h
66 ...Wlel
2
Karpov tightens the screw on the enemy
king.
a
b
c
d
e
f
g
h
57 . . . Wfxe4t! 5 8 . .ixe4 .ixe4-+
56 ...i.h2t
Karpov begins a forcing sequence which
culminates in winning a second pawn.
67.i.c2 �g5 68.i.bl Wle2t
With two extra pawns the bishop ending is
an easy win.
69.Wlxe2 he2 70.�f2 i.g4 71 .�e3 f5
0-1
2003
489
It has been fifty one moves since Black incurred a structural weakening in the form of doubled
f-pawns. Having lived with this handicap for so long, it is rather fitting that when Karpov finally
rids himself of the- weakness by exchanging it, his opponent immediately resigns.
In the remaining three games Karpov defeated Illescas, then drew with Kolev and Khalifman,
for a decent final score of 4Yz16.
Karpov's one remaining classical tournament of the year was the Essent Crown, a double-round­
robin event in Hoogeveen. In the first round he suffered an unpleasant defeat at the hands of
Judit Polgar, who won a fine attacking game against the Petroff. In the second round Karpov
had the black pieces again, and managed to draw against Ivan Sokolov after being under some
pressure. In the next two games he was pressing against Aronian and Polgar, but had to settle for
draws in both games. In Round 5 he was beaten comprehensively by Aronian, and in the last
round against Sokolov he took a quick draw. Karpov's disappointing final score of 2/6 left him in
last place. It would be more than three years before he next took part in an individual tournament
with a classical time control.
***
For the rest of the year Karpov played in a number of rapid and exhibition events. At the
prestigious Cap d'Agde tournament he failed to qualify for the elimination stage and actually
finished in last place in his preliminary group. However, he did manage to win a six-player round­
robin tournament in Lindsborg, after defeating Shulman and Charbonneau, and drawing with
Donaldson, Onischuk and Morovic Fernandez.
2003 Summary
Wijk aan Zee ( l I th- 1 2th place) : 6/ 1 3 (+2 =8 -3)
Spanish Team Championship: 4Yzl6 (+3 =3 -0)
Essent Crown, Hilversum (4th place) : 2/6 (+0 =4 -2)
Total 50.0% (+5 = 1 5 -5)
III Wins • Draws • Losses
2004
Rating 2684 (2 1 in the world)
Karpov started the year with several rapid events. Then he played a mixed format match in
Lindsborg, USA against Zsuzsa Polgar. There were no regular time-control games, j ust rapid and
blitz. The first rapid game was a draw, but we shall look at the second game.
I Game 60 I
Anatoly Karpov - uuzsa Polgar
Lindsborg (rapid
-
2) 2004
l.d4 lLlf6 2.c4 g6 3.ttJc3 i.g7 4.e4 d6 5.h3
Karpov had never played this move before against a decent opponent, but later he played it
regularly. Probably he did not feel like allowing his opponents to attack his king, and in this line
White rarely faces a kingside attack.
5 ... 0-0 6.i.e3
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
6... c5
Zsuzsa would like to play a Benoni-type position, but there were also a few King's Indian style
options which Karpov faced in later years:
The Prime Years
492
6 . . .lLl a6 7.lLl f3 e5
In 2006 Judit Polgar tried 7 . . . We8 in a blitz
game against Karpov, but he won that game
as well.
8.dxe5 dxe5 9.c5 lLl b4 1 O.Wa4 a5 1 1 .l'!d1 id7
1 2 .Wb3 We8 1 3 .ic4 lLlxe4?
Golubev, playing Black, is a very imaginative
player who has introduced at least a couple
of great novelties; however this move is
simply bad.
a
b
c
d
e
f
g
7 . . . Wa5 Black does not have to sacrifice a pawn.
A fairly recent game between strong players
continued: 8.id3 dxc5 9.e5 lLl fd7 1 0.f4 l'!d8
1 1 .We2 lLl c6 1 2.lLlf3 lLld4 1 3 .lLlxd4 lLlxe5
14.fxe5 cxd4 1 5 .id2 dxc3 A draw was agreed
in Miroshnichenko - Markos, Plovdiv 2008.
8.Wxd8 gxd8
a
h
1 4.l'!xd7! lLlxc3 1 5 .l'!xf7
1 5 .l'!xc7!?
1 5 . . . lLl d3t?
This is another mistake.
1 6.ixd3 l'!xf7 1 7.bxc3 a4 1 8.Wc4 <j{h8 1 9.ie4
c6 20.h4
White was winning in Karpov - Golubev,
Odessa 2008.
For 6 ... e5, see Karpov - Miroshnichenko in
the next chapter (page 505).
7.dxc5
7.d5, with the idea of taking back with the
e-pawn, is slightly more popular than Karpov's
move.
b
c
d
e
f
g
h
9 ..hc5
Karpov grabs the pawn. How does this
compare with the Samisch variation? His e4pawn is weaker but on the other hand the f3square is available for his knight, therefore Black
has less control over d4. By the way, Karpov
faced the similar Samisch pawn sacrifice twice
and scored 1 Yz points.
9 lLlc6 lo.lLlf3
•.•
Karpov of course develops his knight to the
most natural square.
lO ... b6 1 1 .i.e3 lLld7?
7... dxc5?!
Zsuzsa directs her forces against the c4pawn and bases her hopes on the speed of her
play, but also allows Karpov to accelerate his
development.
Polgar plays a rarely-employed move;
probably she wanted to copy a similar pawn
sacrifice in the Samisch variation. The usual
move is:
1 1 . . .ib7!?
Just developing looks more attractive, as it
may also put pressure on the e4-pawn.
Anatoly Karp ov - Zsuzsa Polgar
2004
493
Zsuzsa is looking for counterplay rather too
impatiently. It was not too late to change plans
and look for compensation with:
1 2 . . . .ib7!?
This makes it trickier for White to keep an
advantage. Let's see:
a
b
c
d
e
f
g
h
1 2.2"kl
After 1 2.c5? tLl b4! White's position falls
apart.
1 2 . . . tLl b4
If 1 2 . . . tLl d7 1 3 . .ie2 2"i:ac8 1 4.0-0 tLl c5 then
Black has compensation for the pawn.
1 3 . tLl d2 e6
1 3 . . . tLl e8 could be met by 1 4.f3 .
1 4.a3 tLl d3t 1 5 . .ixd3 2"i:xd3 1 6 .We2 2"i:ad8
Black's better-placed pieces give sufficient
play for the pawn.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
12.0-0-0!
Karpov takes advantage of the fact that
Zsuzsa allowed him to castle long. Now he
can develop more easily and the king will
be great on the queenside in the ensuing
endgame.
12 ...hc3?!
a
b
c
d
e
f
g
h
1 3 . .ie2!
1 3 . tLl d4 tLlxd4 14 . .ixd4 tLl f6 1 5 .e5 ( 1 5 . f3 ?
tLlxe4!) 1 5 . . . tLlh5 1 6.2"i:gl tLl f4 Black i s too
active; sooner or later White must give back
the pawn.
1 3 . tLl d5 tLl c5 1 4 . .ixc5 (or 1 4.e5 i>f8 1 5 .2"i:e1
[ 1 5 . .ig5 tLl e6] 15 ... 2"i:ac8 when the position
is balanced) 14 . . . bxc5 1 5 . .id3 2"i:ab8 Black's
domination on the dark squares gives enough
play.
1 3 . . . .ixc3
1 3 . . . 2"i:ac8?! is too slow as White can
consolidate his advantage. 1 4 . tLl d4! tLlxd4
1 5 . .ixd4 tLl f6 1 6.f3 Black has very little for
the pawn.
1 3 . . . tLl c5 1 4.2"i:xd8t 2"i:xd8 1 5 . .ixc5 bxc5
1 6.2"i:dl tLl d4 1 7.tLlxd4 .ixd4 1 8 .Wc2 It
is hard to tell whether Black can hold the
position despite the pawn deficit.
1 4.bxc3 2"i:ac8
1 4 . . . tLl a5 1 5 .2"i:d4 tLl c5 ( 1 5 . . . tLl c6! ?) 1 6.2"i:xd8t
2"i:xd8 17 ..ixc5 bxc5 1 8 . tLl d2 White should
be better with the extra pawn, even if it is a
doubled pawn.
The Prime Years
494
14.c5!
Karpov returns the pawn and in addition
gives up the two bishops, but he also manages
to exchange a lot of pieces; by doing so, he
makes sure his king will be superior to his
opponent's. He also creates a weakness in Black's
camp and, thanks to the limited number of
pieces on the board, he can put pressure on it
more easily.
a
b
c
d
e
f
g
14...i.xfl
h
1 5 .Ei:d2!
White can keep the initiative on the d-file.
After 1 5 .c5 ltJxc5 16 ..ixc5 bxc5 1 7.Ei:d5
ltJ a5 Black is safe as there are a lot of pieces
on the board, so White cannot get at the
weak c5-pawn.
1 5 . . .'i!:rfS
1 5 . . . ltJa5?? loses to 1 6.Ei:hd l .
1 5 . . . ltJ f6 1 6.Ei:xdSt ltJxdS 1 7.e5 ltJ d7 I S .Ei:dl
.ixf3 1 9 .9xf3 ltJxe5 20.c5 bxc5 2 1 ..ia6
White's bishops are unpleasant to live with.
1 6.Ei:hd l i>eS 1 7.c5! ltJxc5 I S . .ixc5 bxc5
1 9 . .ib5
Though Black may hold, White still has an
edge.
After 1 4 . . . ltJxc5 1 5 . .ixc5 Ei:xd l t 1 6.mxd l
.ixfl 1 7.Ei:xfl bxc5 l S .i>c2 White's advantage
is similar to the game.
15J�hxfl llJxc5
If 1 5 . . . bxc5 1 6.Ei:d5 ltJ f6 1 7.Ei:xdSt ltJxd8
I S . ltJ d2 ltJ e6 1 9.i>c2 White has several plans
to increase his advantage, including f2-f3
and ltJ c4 or ltJ b3 with Ei:d l . White could also
consider g2-g4, or even f2-f4 and f5 .
16 ..ixc5 bxc5
13.bxc3 .ia6
The eldest Polgar sister immediately targets
the weak pawn.
a
b
c
d
e
f
g
h
17.i>c2!
After creating the weakness on c5, Karpov
approaches it with his king.
17...e6?
a
b
c
d
e
f
g
h
Zsuzsa probably thought that Ei:d5 was
White's main threat, and perhaps this idea led
2004
Anatoly Karpov - Zsuzsa Polgar
to this clear mistake. We shall look at her other
options:
1 7 . . . Wf8 1 8 .�xd8t
If 1 8 .�d5 �xd5 1 9.exd5 tt:J a5 20.�b l tt:J c4
Black is clearly worse but has some chances
to escape.
1 8 . . . tt:Jxd8
1 8 . . . �xd8 1 9 .�b l a6 Black cannot exchange
rooks, but she can still put up resistance.
20.�b7 (if 20.�b6 �d6 2 1 .Wb2 tt:J a5 then
Black is worse but still alive) 20 . . . �d6 2 1 .�c7
White has excellent chances to squeeze a
win.
1 9.�b l �c8
White has many promlsmg tries, such as
20.tt:Je5 f6, 20.'it>d3, or 20.�b5 but Black's
chances here are much better than in the
game.
1 7 . . . �db8!
Black's best chance is to try to keep both
rooks on the board, so that the white king
will find it more difficult to get near Black's
weak point.
495
tt:J e6) 20 ... tt:J e6 2 1 .�c4 (2 1 .�b5 �c6)
2 1 . . .�ab8 Black finds some counterplay, or
at least White has difficulty in winning.
1 8 . . . �b5
After 18 ... �b6 1 9.�xc5 �ab8 20.a4 tt:J d8
2 1 . tt:J d4 �b2t 22.'it>d3 White has excellent
winning chances.
1 9 .a4 �a5 20.�al �c8 2 1 . tt:J d2
Black is facing a long tough defence.
8
7
6
5
4
3
2
1
a
b
c
d
f
e
g
h
18J�xd8t!
With this exchange Karpov opens a path to
invade; this strategy works so well because his
king is superior to Black's.
18 ... gxd8 19.9bl c4
Zsuzsa uses the c-pawn as a sacrificial
lamb. Instead, trying to stop the invasion by
exchanging rooks is hopeless:
1 9 . . . �b8
a
b
c
d
e
f
g
h
1 8 .�d5!?
The strategy of exchanging a pair of rooks,
which works superbly in the game, is not
so effective here: 1 8 .�b l ?! �b6! 1 9.�fd l
�a6 20.�b2 �a4 Black gains reasonable
counterplay.
1 8 .a4 �b6 1 9.�d5 tt:J d8 20.�xc5 (20.tt:Je5
a
b
c
d
e
f
g
h
The Prime Years
496
20.Elxb8t ltJxb8 2 1 .ltJe5
White also wins after 2 1 .<j{d3 c;f;>g7 22.<j{c4
ltJ d7 23.c;f;>b5 c;f;>f6 24.a4.
2 1 . . .f6
Or 2 1 . . . c;f;>f8 22.<j{b3 <j{e7 23.<j{c4 f6 24. ltJ d3
and the c5-pawn falls.
22. ltJ d3 ltJ d7 23.<j{b3
White wins.
25 ... <j{g7 26J3c5 c!tJe7
Zsuzsa gives up a pawn hoping to postpone
the inevitable. If 26 . . . e5 then 27. ltJ d6 wins a
piece.
a
b
c
d
e
f
g
h
27.:axa5 c!tJd5
a
b
c
d
e
f
g
h
20J3b7!
Karpov goes after the correct target; he wants
to destabilize the knight with Elc7. Instead
after 20.ltJd2?! ltJe5 Black would gain some
counterplay.
Black's position is hopeless. Probably
Karpov was short of time, as otherwise such
a strong player as the former Women's World
Champion would have resigned.
28.:aa7t @g8 29.ga6 ge8 30.a5 g5 31.gc6
Karpov prepares to push the pawn.
31. .. ga8 32.a6 @f7 33.lLlc5 c!tJe7 34.gxe6
20 :ac8 21 .g3
..•
Zsuzsa managed to stop Elc7, but now her
pieces are almost unable to move.
21...a5 22.a4
Karpov fixes the pawn; now Black has one
less soldier. . .
22 ... h6 23.:ab5!
Black loses a second pawn.
34 ... gc8 35.a7 ga8 36.:aa6 lLlc8
8
7
6
5
With her last move Zsuzsa enabled her king
to improve with . . . <j{g7; that is why Karpov
starts harvesting pawns.
4
23 ... £5 24.lLld2 fxe4 25.lLlxe4
2
25 .Elc5 also wins.
3
1
a
b
c
d
e
f
g
h
2004
Glenn Flear - Anatoly Karp ov
37J:�c6!
I Game 61 I
Of course everything wins here but Karpov
finds the most purposeful move.
Glenn Flear
37... llJxa7
Zsuzsa steps into a lethal pin, but
approaching with the king loses a piece by
force. For example, 37 . . . 'it>e7 3S.l'l:xcS! l'l:xcS
39.tLla4 or 37 . . . 'it>eS 3S.l'l:xcSt l'l:xcS 39.tLla6
and White wins after either 39 . . . l'l:aS 40.tLl c7t
or 39 . . . �dS 40.tLlbS.
497
-
Anatoly Karpov
Aix-en-Provence (rapid
-
1 .2) 2004
l.d4 llJf6 2.c4 e6 3.llJc3 .ib4 4.\Wc2 0-0
Karpov rarely plays the 4 . . . c5 line.
5.e4
38J�a6 �e7 39.llJa4 �f8 40.llJb2 �e7?!
Flear, a strong theoretician, opts for a sharp
line; Karpov has played this variation with
both colours.
This hastens the end but Black was desperately
lost anyway.
5 ... d6
41 .llJxc4
1-0
Karpov played a match against Sadvakasov in
Astana, the capital of Kazakhstan. The first fout
games were at a regular time control and the
last four were rapids. Karpov started as White
and drew. In the second game, close to the
time control, he blundered a pawn and lost.
The third game was drawn and in the last one
Karpov was in trouble, but managed to draw.
Thus he lost the match 2Y2- 1 Y2. The rapid part
ended 2-2.
Karpov also played a few more rapid events.
In the Aix-en-Provence knockout tournament
he started against Glenn Flear. The first game
ended in a draw, so we shall look at the second.
The English grandmaster has faced three world
champions at regular time controls, and drew
all four games against them.
Karpov sticks to the move he had already
played twice as Black.
6.a3 .ixc3t
Flear chooses the main line. When I played
a tournament with grandmaster Balashov,
who worked with Karpov for a long time, he
told me that he had played hundreds of blitz
games with Karpov in the Samisch variation
of the Nimzo-Indian. So Karpov had plenty of
experience in the resulting structure.
7.bxc3 llJc6
This knight move had been played only once
before. Earlier Karpov had twice played 7 . . . e5,
losing one and drawing the other.
8 ..id3
S.�g5 ! ? was worth considering.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
The Prime Years
498
8 ... h6!?
A novelty, preventing the bishop pinning his
knight. Of course it is more unpleasant to face
a novelty in a rapid than in a regular game, as
there is less time to find the answer.
9.lLle2?!
Since this game White has usually played
9.f4 and developed the knight to f3; White has
scored well with this plan.
Karpov puts more pressure on the c4-pawn,
but this is rather slow. Black could instead
hold back the attack first, and only then turn
his attention to the queenside:
1 3 . . . f5 !? 1 4 .exf6 ct'lxf6 1 5 .lMrxe6t (after 1 5 .0-0
lMrd7 Black is not worse) 1 5 . . . 'it>h8 1 6.0-0 We8
Black wins back the pawn.
9 ... b6!
The e2-knight is no help in defending the
c4-pawn, so Karpov prepares to attack it.
10.f4 lLlaS
8
7
6
a
5
b
c
d
f
e
g
h
14.cxdS?
4
Flear saves the pawn, but brings Black's
queen into play.
3
2
a
b
c
d
e
f
g
h
1 4 . f5 !
White should sacrifice the c4-pawn and go
after Black's king at once.
1 1 .eS
8
If 1 1 . ct'l g3 then Black has a pleasant choice:
1 1 . . .ia6 (or 1 1 . . .ct'l g4!?, planning . . . lMrh4, is
also rather annoying for White) 1 2 .lMre2 Black
can choose between playing against c4 with
1 2 . . . c5, or exchanging the dark-squared bishop
with 1 2 . . . ct'lb3; both would give a playable
position.
7
1l ... lLle8 12.<!Llg3 .ta6 13.�e2
It would be too risky to sacrifice the c4pawn.
13 ... dS?!
<-=0///////
6
5
4
3
2
a
b
c
d
e
f
g
h
1 4 . . . ixc4
1 4 . . . ct'lxc4? 1 5 . 0-0 c5 ( 1 5 . . . lMrh4 loses to
1 6.lMrf3) 1 6.f6 White is winning.
2004
Glenn Flear - Anatoly Karpov
14 . . . dxc4 1 5 .t.e4 t.b7 1 6.t.xb7 It:lxb7
1 7.0-0 White has an extremely dangerous
initiative on the kingside.
14 . . . lt:l b3 1 5 J'&b l It:lxc l 1 6.:1'1xc l t.xc4
1 7.t.xc4 dxc4 I B .Wxc4 Wd5 1 9.Wxd5 exd5
White's space advantage gives him good
chances.
1 5 .t.xc4 dxc4
After 1 5 . . . lt:lxc4 1 6.0-0 f6 1 7.fxe6 fxe5
I B .:1'1xfBt �xfB 1 9.dxe5 c6 20.Wf3t �gB
2 1 .Wf7t �h7 22.t.f4 White is almost
winning.
1 6.0-0 It:l b3
1 6 . . . c5 loses to 1 7.f6 cxd4 I B .fxg7.
499
14....bd3 1 5.W1xd3 W1xd5 16.0-0
a
b
c
d
f
e
g
h
16 ... f5!?
Karpov stops White's kingside attack. Karpov
twice beat Yusupov with a similar . . . f5 move. It
is worth having a look at both of these games:
l .d4 It:l f6 2.c4 e6 3 . lt:l c3 t.b4 4.e3 c5 5 .t.d3
It:l c6 6.a3 t.xc3t 7.bxc3 0-0 B . lt:l e2 b6 9.e4
It:l eB 1 0.0-0 t.a6 1 l .f4
a
b
c
d
e
f
g
h
Now White must resist the temptation to
move the a I -rook, and choose between two
vicious attacking moves:
1 7.Wg4!!
Also strong is 1 7.Wxc4!. For example:
17 . . . lt:lxal I B .fxe6 We7 1 9 .1t:lf5 Wxe6
20.Wxe6 fxe6 2 1 .lt:l e7t �h7 22.:1'1xfB It:l b3
23 .t.e3 Black is in deep trouble.
After the text move Black has a few tries, but
no way to stop the attack. For example:
1 7 . . . lt:lxa l
1 7 . . . exf5 I B .lt:lxf5 �h7 1 9 .1t:lxh6! gxh6
20.Wf5 t �g7 2 1 .t.xh6t �xh6 22.:1'1f3+­
Also 1 7 . . . �hB I B .f6 is devastating.
I B .t.xh6 It:l c2 1 9 .fxe6 We7 20.lt:l f5 Wxe6
2 1 .t.g5 !
The threat of It:l e7t is deadly, especially as
2 1 . . .f6 loses the queen to 22.lt:lh6t.
a
b
c
d
e
f
g
h
1 1 . . . f5
As mentioned above, Karpov played this
variation twice against Yusupov, beating him
on both occasions. The first game was a vital
one as it was towards the end of their World
Championship Candidates match.
1 2.lt:lg3 g6 1 3 .t.e3
Now Karpov deviated from their first game
in this variation and played:
The Prime Years
500
1 3 . . . cxd4
When Karpov reached this position against
Yusupov for the first time, he played:
1 3 . . . tLl d6 1 4 .exf5 tLlxc4! 1 5 .ixc4 ixc4
1 6.fXg6 ixfl 1 7.1Wh5! 1We7 1 B .l'hfl hxg6
1 9.1Wxg6t 1Wg7 20.1Wd3 (or 20.1Wxg7t <j;lxg7
2 1 .dxc5 bxc5 22.ixc5 and the position is
unclear) 20 . . . cxd4 2 1 .cxd4 1Wh7 22.1Wb5 a6!
23 .1Wb3 tLl e7
23 . . . tLl c4!
Karpov again manages to create dominating
knights.
24.ig5 tLl xd4 25.h5 gxh5 26.Ei:fl t me8
27. tLl xh5 tLl xa3
In Yusupov - Karpov, Linares 1 993, Karpov
had a large advantage and went on to win
convincingly.
Back to the game.
8
7
6
5
4
a
b
c
d
3
e
f
g
h
Karpov obtained an edge and went on to
win, Yusupov - Karpov, London (6) 1 9BB.
It would be interesting to know what
Yusupov had prepared if Karpov repeated
the first game, as it looks alright for Black.
1 4.cxd4 d5 1 5 .cxd5 ixd3 1 6.1Wxd3 fXe4
1 7.1Wxe4 1Wxd5
Karpov takes the sting out of White's
attack.
1 B .1Wxd5 exd5 1 9 .Ei:ac 1 Ei:cB 20.f5 tLl d6 2 1 .fXg6
hxg6 22.Ei:xfBt <;t>xfB 23.h4?
After 2 3 . tLl e2!? tLl c4 24.Ei:c3 White is slightly
better.
2
1
a
b
c
d
e
f
g
h
17.exf6
A reasonable decision; White at least keeps
Black busy with the e6-pawn. Black has the
upper hand on the queens ide because of his
control of the c4-square.
17 lLlxf6 18.i.d2?
•.•
Flear wants to put pressure on the e6-pawn,
but this plan is too slow.
1 B .1Wg6 1Wd7
After 1 B . . . 1Wc6 1 9.Ei:f3 tLl c4 20.Ei:a2 the
position is equal.
1 9.Ei:a2
Not 1 9 .f5 ?! as 1 9 . . . 1WeB! exchanges queens.
1 9 . . . tLl b3 20.Ei:e2 tLl xc 1 2 1 .Ei:xc 1 Ei:aeB 22.tLle4
1Wa4 23.Ei:ce 1 tLl d5 24.g3
White should not be worse.
1 B .Ei:e l
a
b
c
d
e
f
g
h
2004
501
Glenn Flear - Anatoly Karpov
22.E!dl h5 23.�g6 �d7
Karpov retreats the queen to neutralize
White's kingside attack.
a
b
c
d
e
f
g
h
l S . . . E!aeS!
l S . . . ttJ c4?! allows the rapid transfer of the
rook to the e-file. 1 9 .E!a2! c5 20.E!ae2 E!feS
2 1 . ttJ e4 ttJxe4 22.E!xe4 White has enough
pressure on the e-file to compensate for the
strong c4-knight.
l S . . . E!feS 1 9 .ttJ e4 ttJxe4 20.E!xe4 ttJ c4
2 1 .E!a2 White can set up pressure against
the e6-pawn.
1 9.E!b l !
1 9 . ttJ e4 c5
1 9 . . . ttJ c4
After 1 9 . . . c5 20.E!b2! cxd4 2 1 .cxd4 E!dS
22.E!be2 ttJ b3 23 .ib2 ttJxd4 24.E!e5 Wd7
25 .Wc4 White has compensation for the
pawn.
20J''1 b 4 b5 2 1 .a4 c6 22.Wc2
Black's advantage is rather small.
18 ... llJc4
Karpov naturally puts his knight on a great
square.
a
b
c
d
e
f
g
h
24.S?
Allowing a queen exchange prevents White
from creating counter-chances. White already
had problems, but he could have tried:
24.E!fe l
Exerting pressure o n the e6-pawn would tie
Black's hands at least a little.
24 . . . a5 2 5 . ttJ e4
If 25 .Wd3 ttJ d5 26.E!fl Wc6 Black has nice
prospects on the queenside.
25 . . . ttJxe4 26.E!xe4 E!f6 27.Wg4 Wd5
White will have big problems when the
b-pawn advances.
19.E!ael c5!
Karpov starts attacking the centre.
20 ..icl cxd4 21 .cxd4
2 1 .Wxd4 would not avoid all the trouble as,
for example, 2 1 . . .Wc6!? is quite strong.
21..J�ad8!
Karpov gives his opponent no time to exert
pressure on the e-file.
a
b
c
d
e
f
g
h
The Prime Years
502
24 ...�e8!
Exchanging queens leads to a superior, quite
possibly winning, endgame.
then 33 . . . �a7 is very strong or if 33.�dd3 �a6
34.'tt> e l (34.�g3 ga l ) 34 . . . gal and White is
lost.
25.�xe8
33 ... gf7!
2 5 .�fe l is also tough for White: 25 . . . �xg6
26.fxg6 �d6 27.�e2 ltJd5 28.ltJe4 �c6 29.h4
as Perhaps White's problems are smaller than
in the game.
Karpov ties up the fS-knight.
34.i.cl
After 34.ltJ g3 �xf3t 3 S .�xf3 b3 Black wins
as the b-pawn is so strong.
25 .. JUxe8 26.fxe6 gxe6 27.�f5 �d5
Karpov blocks the d-pawn and the knight
will help the b-pawn. Black's superb knight
will dominate the rest of the game.
28.ga �h7
Karpov takes the time to improve his pieces
before he pushes his b-pawn.
29.h3 gd7! 30.g4 a5 31.�fl
34... ga7
It is impressive and instructive how Karpov
uses the whole board to crack his opponent's
resistance; now he occupies the a-file, a few
moves earlier he placed his king on h7.
Incidentally, 34 . . . ltJ c3 was also winning:
3 S . ge l (3 S .gfl ge2t) 3S . . . gxe 1 36.<;:t>xe l g6
37.ltJh4 �e7t 38.<;:t>fl �a7
35.g5 hxg5 36.i.xg5
8
7
6
5
4
3
a
b
c
d
e
f
g
h
31...b4!
The b-pawn
soon it will be
not only pushes
to promotion, it
rook.
becomes threatening and
irresistible. Karpov's move
the pawn one square closer
also opens the a-file for his
32.axb4 axb4 33.i.d2
IfWhite concentrates on the third rank, then
Black can use the a-file effectively. If 33 .�b3
2
1
a
b
c
d
e
f
g
h
36 ... �c3!
After outplaying his opponent strategically,
Karpov uses a few little tactics to finish him
off.
37.gcl �e4t 38.�gl �xg5
Karpov wins the bishop; the game is over.
2004
Glenn Flear - Anatoly Karp ov
503
39.�g3 tLld2
The knights are working their magic for Karpov in this game.
40.d5 �e5 41.tLle3 tLlgf3t 42.�hl
If 42.'tt> f2 then 42 b3 wins.
...
42 ... �xe3
0-1
In the semi-final Karpov eliminated Kazhgaleyev and then beat Istratescu in the final. Both
matches ended 1 Y2-Y2.
2004 Summary
Match versus Sadvakasov, Astana: Lost 2 Y2- 1 Yz (+0 =3 - 1 )
Total 37. 5% ( + 0 =3 - 1 )
iii! Wins • Draws • Losses
2005
Rating 2674 (29 i n the world)
In 2005 Karpov played another mixed-format match - he played four regular and four rapid
games against Andrei Istratescu in Bucharest. The first game ended a draw, then Karpov won a
fine ending in the second. However Istratescu managed to strike back with a win; the Romanian
grandmaster attacked, sacrificed a pawn, and after a mistake by Karpov he gained a winning
position. In the last regular game Karpov had a small edge due to his bishop pair; he won a pawn
and reached a winning position. A slip allowed his opponent to escape to a queen versus rook and
pawn theoretical draw, but Istratescu was unable to hold it. Thus Karpov won the regular part of
the match 2Y2- 1 Y2. The rapid match also ended in favour of the ex-World Champion, this time
by 3Y2-Y2.
Karpov also played two games in the Russian team championship. We will look at his game
against the Ukrainian GM Miroshnichenko.
\ Game 6l \
Anatoly Karpov - Evgeny Miroshnichenko
Sochi 2005
l.d4 tiJf6 2.e4 g6 3.tiJc3 i.g7 4.e4 0-0 5.i.e3 d6 6.h3 e5
White players often develop the bishop to g5 instead of e3; one major difference is that
after 5 .i.g5 d6 6.h3 if Black plays 6 . . . e5 it costs him material after 7.dxe5 dxe5 8.'lWxd8 E!:axd8
9.tLld5.
7.d5
In this variation White is virtually forced to push the d-pawn in reply to . . . e7 -e5 , as otherwise
Black would take on d4 when h2-h3 would become a significant weakness. For example, after
f2-f3 to defend the e4-pawn, a hole would appear on g3 .
7... a5 8.e5!?
Karpov plays a novelty, but probably he j ust wanted to get out of theory. It is, of course, hard
to tell how much time he invested analysing the variation.
Another difference between 5 .i.g5 and 5 .i.e3 is that with 5 .i.g5 h6 6.i.e3 inserted, Black
rarely sacrifices a pawn with the sequence . . . tLl d7-c5, a2-a3 . . . a5-a4, i.xc5 . . . dxc5 and tLlxa4,
because Black no longer has the option of . . . i.h6.
The Prime Years
506
Karpov played differently against Kasim­
dzhanov at the 2007 blitz world championship:
8.g4 tLla6 9. �e2 tLlc5 1 O. Wc2 �d7 1 1 .0-0-0
Wh8? 1 2.g5 tLlg8 1 3 .h4 f6 1 4 . tLlf3 �g4 1 5 . tLld2
�xe2 1 6. tLlxe2 tLld7 1 7.h5 fxg5 1 8 .hxg6 h6
1 9 .1"1dg1 1"1f4
Black accepts a backward pawn on c7, but on
the other hand he develops quickly. Instead
9 . . . cxd6 1 O. tLlf3 b6!? is interesting, with the
idea of continuing with . . . tLlc5 .
8
7
6
5
4
3
a
b
c
d
e
f
g
h
Here Karpov missed a clear win with
20. �xf4 exf4 2 1 . tLlf3 �f6 22. tLled4. Even so
he had a very promising position, but he did
not manage to win in Karpov - Kasimdzhanov,
Moscow (blitz) 2007.
2
1
a
b
c
d
f
e
g
h
lO...c6?
At some point Black has to push the c-pawn
as he no longer has the option of defending
the b5-square with . . . a7 -a6 and playing instead
on the kingside would not be fast enough.
However, the text move, trying to solve the
problem of the backward c-pawn with no
preparation, is obviously premature, so we
should consider the alternatives:
1 O . . . 1"1d8
8
',=Fmn
7
6
5
4
8 . . . tLle8, planning an early . . . £7 -f5 , is
interesting and absolutely in the spirit of the
King's Indian.
3
2
a
9.cxd6 'lWxd6?!
This is an unusual way of recapturing on d6.
1 1 . �e2!
b
c
d
e
f
g
h
2005
507
Anatoly Karpov - Evgeny Miroshnichenko
White must be a little careful. For example:
1 1 . �c4?! ltJc5 1 2.'1W c2? This natural move
leads to trouble (instead after 1 2. ltJd2
c6 l3.dxc6 'iffixc6 Black is doing fine) .
1 2 . . . ltJfxe4! 1 3 . ltJxe4 ltJxe4 1 4 . 'iffixe4 �f5
1 5 . 'iffih4 'iffib 4t 1 6. �d2 'iffixb2 1 7.E&dl (after
1 7.0-0 b5 Black is taking over) 1 7 . . . b5
1 8 . � b3 a4 1 9 . �c 1 'iffic 3t 20. �d2 'iffid3 Black
can avoid the repetition and win.
1 1 . . .c6 1 2.dxc6 'iffic7 l3. 'iffib 3 ltJ b4 1 4 . 0-0 �e6
1 5 . 'iffia4 ltJxc6
After 1 5 . . . 'iffixc6 1 6. �b5 'iffic7 1 7.E&ac 1 Black
is lagging well behind in development.
1 6.E&fdl E&ac8 1 7. 'iffib 5
White has the initiative on the queenside.
1 O . . . � d7!?
This is probably Black's best choice; apart
from preparing the natural . . . c7-c6 this
move also gives the option of playing . . . b7b 5 . Now White has to play accurately to
gain a small advantage.
1 1 .E&c 1
After 1 1 . ltJ d2 'iffib 4 1 2. 'iffic2 c6 l3.a3 'iffie7
Black is very much in the game.
1 1 . . .c6 1 2.dxc6 'iffixdlt l3.E&xdl �xc6
1 4. �d3 (l4. ltJxe5 ltJxe4!) 1 4 . . . E&fe8 1 5 .0-0
(after 1 5 . �c4, trying to exploit the newly­
weakened V-pawn, 1 5 . . . E&ac8 1 6.0-0 h6 it
is hard to do anything with White's small
advantage) 1 5 . . . �f8 1 6. �c4 h6 1 7. ltJ h2
White is just fractionally better.
1 2. �d3 c6
Black is j ust a fraction worse after 1 2 . . . ltJc5
1 3 .0-0 ltJxd3 1 4 . 'iffixd3 c6 (l4 . . . E&fb8
1 5 . ltJ e2) 1 5 .E&fdl E&fc8 1 6.dxc6 'iffixd3
1 7.E&xd3 �xc6 1 8. ltJxe5 �xe4 1 9 . 1tJxe4
E&xc 1 t 20. �xc 1 ltJxe4.
l3 .dxc6 �xc6 1 4 . 0-0 E&fd8 1 5 . �xb5 'iffixdl
1 6.E&fxdl �xb5 1 7. ltJxb5 ltJxe4
If White has an edge then it is very small.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
1l.dxc6! '\We7?!
a
b
c
d
e
f
g
h
1 1 . . . b5
The text move is the boldest try, but we
should also consider other options:
1 1 . . .E&fd8 1 2. ltJd2 c6 (if 1 2 . . . ltJ c5 then
l3 . 'iffif3!? defends e4 well, and White will
soon exert pressure on the c-file) l3. ltJc4
'iffif8 1 4 . ltJ b6 White is somewhat better.
Black plays boldly, but this is not best.
After 1 1 . . . 'iffixc6 1 2. ltJxe5 'iffic7 l3. ltJ c4 �e6
1 4 . 'iffid6 'iffixd6 1 5 . ltJxd6 ltJ b4 1 6.E&c 1 ltJxa2
1 7. ltJxa2 �xa2 1 8 . �b5 a4 1 9.0-0 Black still
has problems; he can probably exchange the
b2-pawn and try to hold a draw even if he loses
the b7-pawn.
12.i.c4!
Karpov develops the bishop to a fine diagonal
rather than taking a pawn. Black sooner or
later will have to take on c6, which will give
Karpov a lovely target to build pressure against.
The Prime Years
508
1 2.cxb7? �xb7 was Black's intention, which
would give him fluent piece-play and he would
also be likely to win the e4-pawn.
12...bxc6
Winning a tempo with 1 2 . . . EidS would
not change much. 1 3 . Wa4 (also good for
White is 1 3 . Wb3, making sure that Black will
have an isolated pawn on c6 after 1 3 . . . bxc6
1 4.0-0) 1 3 . . . lUb4 1 4.0-0 lUxc6 1 5 . �b6 EieS
1 6.Eifd 1 Black has managed to avoid structural
weaknesses, but his position is still worse as
White is much more active.
13.0-0 �c5 14.'1Wc2
This is natural and strong; Karpov defends
the e4-pawn and starts lining up his heavy
pieces against the c6-pawn.
1 4 . . . �b7!?
This passive-looking move offers the toughest
resistance.
1 5 .Eiael h6
1 5 . . . EifdS loses to 1 6. lUg5 EifS 1 7. �xf7t
as 1 7 . . . Eixf7 ls. lUxf7 g;xf7 1 9 . �xc5 '.Wxc5
20. Wb3t wins back the piece.
1 6.Eifd 1 EiabS
Black j ust holds his position together.
It would have been interesting to see
how Karpov would have increased his
advantage.
1 7. �xc5
1 7.Eid2 is also very strong, as Black is
struggling after 1 7 . . . � aS lS.Eicdl Eife8
1 9.Eid6 lUcd7 20. lU a4.
1 7 . . . Wxc5 1 S. lUb5
Or lS. lU a4 We7 1 9. 1Uc5 when Black is very
passive.
lS . . . We7 1 9 . 1Ud6 �aS 20. �b3
Black can hardly move.
15.�a4!
A great strategic decision; Karpov tries to
exchange around the weakness in order to be
able to get closer to it.
15... .!Lle6?!
a
b
c
d
e
f
g
h
14... .!Llfd7?
Black wants to have as many pieces as
possible around c6, but Karpov will soon
reveal that Black has more than one weakness.
Black's move also has the drawback of making
the cS-bishop very passive. The following
alternatives are stronger, but cannot solve all
Black's problems:
After 1 4 . . . �a6 1 5 . �xa6 lUxa6 1 6. lUa4 Wb4
1 7. lUb6 EiabS ls. lUc4 Black has a difficult
position.
Exchanging would have given Black more
air:
1 5 . . . lUxa4 1 6. Wxa4 lUf6 1 7. Wc2
The text move creates the nasty threat of
lS. � c5 .
After 1 7. �d3 � b7 lS.Eiael EifdS 1 9.Eifdl
Black's position is hard to defend as White
has many squares for his pieces on the
queens ide and the c6-pawn to attack, yet
here Black's prospects are less gloomy than
in the game.
1 7 . . . �e6
Of course 1 7 . . . �b7? would allow lS. �c5.
lS.Eifel
I think White's chances of winning are better
than Black's of drawing.
Anatoly Karpov - Evgeny Miroshnichenko
2005
16JUdl
509
18..J�fxe6 19J�adl
One might think Karpov was simply
stopping Black playing . . . tt.'l d4, but there is
more to it . . .
The rook threatens to invade on d6.
19...i.e7
16....if6
Black stops ixe6 and tt.'lg5 which indeed
would lead to a very difficult position. For
example, after 1 6 . . . Ei:eB 1 7. ixe6 �xe6 1 B. tt.'lg 5
�f6 1 9 .Ei:ac 1 h6 20. tt.'l f3 Ei:a6 2 1 .Ei:d3 Black is
almost lost.
8
7
6
5
a
b
c
d
e
f
g
h
4
20.tL'lg5!
3
Karpov forces Miroshnichenko to give up
his dark-squared bishop, which covers the d6square.
2
1
a
b
c
d
e
f
g
h
17J:�d2!
Karpov prepares to double rooks on the
d-file, which would create more pressure than
Black could withstand.
17..J3e8
20....ixg5
After 20 . . . �xa2 2 1 . �xc6 Ei:bB 22. tt.'lc3 �b3
23.Ei:xd7 White wins material.
21.i.xg5 f6 22.i.e3 tL'lfS
8
Or 1 7 . . . tt.'lg5 IB. ie2 tt.'lxf3t 1 9 . ixf3 ib7
2o. ig4 and Black's position falls apart.
7
18.,ixe6!
5
Karpov simplifies to a posItIon that is
winning due to his total control of the d-file.
Occupying the d-file at once with IB.Ei:adl! is
also close to winning. IB . . . tt.'l dfB (if IB . . . tt.'l d4
then 1 9 . tt.'lxd4 exd4 2o. ixd4 �xe4 2 1 . � b3
and Black's position is desperate) 1 9. tt.'l b6 Ei:bB
20. ixe6 (or 20. i b3 � c7 2 1 . tt.'l c4 i a6 22. � c3
when Black is in big trouble) 20 . . . ixe6
2 1 . �xc6 ixa2 22.b4 White is winning.
4
6
3
2
1
a
23.Ei:d6!
b
c
d
e
f
g
h
The Prime Years
510
Karpov's strategy prevails! On the 1 7th move he played :Bd2 to gain control of the d-file; now
on the 23rd he invades on the file.
23...WI'xa2
The queen takes a pawn but abandons the defence of the king.
24J�xf6
Karpov takes a pawn and creates yet another black pawn island, while also weakening Black's
king. White had an alternate win with 24. Wxc6.te6 2 5 . tt:lb6 :Bab8 26 .t
. h6.tfl 27.:Bxf6.
24 .ie6 2S.tt:lcS geb8
•.•
After 25 . . .t
. fl 26.:Bxc6 :Beb8 27.:Bd2 White is winning, as Black is a pawn down with badly
placed pieces.
26.gd2
Karpov defends the pawn before he attacks the rather deserted black king.
26...ctt>g7 27.gB .if7
Or 27 . . . Wc4 28. Wdl .tfl 29. tt:ld7 :Bb5 30. tt:lxf8 :Bxf8 3 1 .:Bd7 and White invades and wins.
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
28.liJd7!
The knight is on its way to the kingside.
28...gd8
If 28 . . . tt:lxd7 then 29.:Bxd7 :Bf8 30. Wxc6 wins instantly.
29.tlJf6
1-0
Karpov's knight has completed its journey; Miroshnichenko was not interested in watching
Karpov kill his king, so he resigned.
In the second game Karpov lost to Shomoev because of a fatal blunder before the time control.
2005
511
Later Karpov played in the Bordeaux Grand Prix knockout rapid event. In the quarterfinal he beat
Marie Sebag, then in the semi-final he eliminated Chabanon. In the final against Hamdouchi he
won 3-2, with all the games being decisive.
Karpov also played a rwo-game match against David Navara, the Czech number one. Both games
were hard fought, but both ended in draws.
2005 Summary
Match versus Istratescu, Bucharest: Won 2 Y2- 1 Y2 (+2 = 1 - 1 )
Russian Team Championship: 1 12 (+ 1 =0 - 1 )
Match versus Navara, Prague: Drew 1 - 1 (+0 = 2 -0)
Total 56.2% (+3 =3 -2)
i!il Wins • Draws • Losses
2006
Rating 2672 (29 in the world)
Karpov started the year well, finishing first equal with Ivanchuk and Kasimdzhanov in the strong
Tallinn rapid tournament with a score of 7/9. Karpov continued with a rapid match against
Portisch, which was organized to celebrate Gligoric's birthday. Karpov won the first game and
drew the last three to win 2\t2- 1 Y2.
Karpov played only one regular time control event in 2006 - the Russian team championship.
He played only three games, drawing quickly with Gelfand and Zvjaginsev, but was nicely beaten
by Akopian.
Karpov lost a rapid match against Spassky by the score of 1 Y2-\t2, but won a four-player blitz
tournament which included Kasparov.
Karpov then played a rapid match against Leko in Miskolc, Hungary. In the first two games
Karpov was pressing but failed to win. In the third he equalized after having a few problems in
the opening, but blundered in the end. The fourth was a fair draw, but from then on Karpov's
level dropped, even though he still managed to draw the last four games.
In the Cap D'Agde rapid tournament he made 50% in the prelims and failed to qualifY for the
four-player play-off.
Karpov finished the year with a rapid match against Giovanni Vescovi in Guarulhos, Brazil. The
Brazilian grandmaster won the first game with a powerful kingside attack, but we shall look at the
second game. Incidentally Vescovi has played three regular games against two world champions,
Kramnik and Topalov; he drew one and lost two.
I Game 63 I
Anatoly Karpov - Giovanni Vescovi
Guarulhos (rapid
-
2) 2006
1.<:4 tlJf6 V!tJO e6 3.d4
One benefit of the move order Karpov chose is that it avoids both the Nimzo-Indian Defence
and the Benko Gambit - two of his opponent's favourite lines.
The Prime Years
S14
3...d5 4.tlJc3 .ib4
Vescovi had played the Ragozin against
Leitao not so long before this game, so Karpov
must have considered it during his preparation.
Vescovi told me that he had played many blitz
games with Karpov and in such cases both
players know each other's strong and weak
points.
ig4 l S. i e2 as 1 6.0-0 ttJb4 1 7. Wb2 l"le8
1 B.h3 ihS 1 9 .a3 ttJdS
8
7
6
a
5
b
c
d
e
f
g
h
20.l"lfc 1 Karpov had cleverly prepared his pieces
to put pressure on the Cuban grandmaster's
queens ide and went on to win, Karpov - Vera
Gonzalez, Benidorm (rapid) 2002.
4
3
2
1
a
b
c
d
e
f
g
h
5.VfYa4t
Karpov had played this check once
before, intriguingly at the Poikovsky Karpov
tournament. The Poikovsky event was founded
to celebrate the 2Sth anniversary of Karpov
becoming the 1 2th World Champion, and has
grown into a strong grandmaster tournament.
It is a rare honour for a player to have a
tournament named after him while he is still
alive.
After the obligatory S . . . ttJc6 the knight blocks
the c-pawn so Black cannot undermine the
centre with . . . c7-cS. Since Vescovi won the
first game of the match with an attack, possibly
Karpov wanted to create a calmer situation in
this game.
It is worth looking at another game of Karpov's
in the Ragozin: S. ig5 h6 6. ixf6 Wxf6 7. Wb3
ixc3t B.bxc3 WdB 9.e3 0-0 1 O. ie2 ttJc6
l 1 .cxdS exd5 1 2. Wa3 b6 1 3 .c4! dxc4 1 4. ixc4
He also had an example in the Manhattan
Variation: S.cxdS exdS 6. ig5 ttJbd7 7Y.flc2
0-0 B.e3 cS 9. id3 Wa5 1 0.0-0 c4 1 1 . 1fS
l"leB 1 2 . ttJd2 g6 1 3 . ih3 ixc3 1 4 . Wxc3 'lMrxc3
lS.bxc3 ttJe4 1 6. ttJxe4 dxe4 1 7. ig4! ttJb6
1 B. ie2 ie6 1 9 .1"lab 1 l"lacB 20.l"lfcl l"lc7
2 1 .f3 idS 22 . .;t>f2 l"lc6 23.h4 'it>g7 24.a4
ttJxa4
a
b
c
d
e
f
g
h
25 .!'lxb7 Karpov obtained a clear advantage
and went on to win, Karpov - Milov, Cap
d'Agde (rapid) 2002.
5...tlJc6
2006
Anatoly Karpov - Giovanni Vescovi
515
23.g4!
Karpov decides to keep his king in the centre:
a very interesting decision.
23 . . l"1:a2
.
24. ii.c3l"1:bB 2 5 l"1:
. xbBt ttJxbB 26. Wd3
hxg4
26 . . l"1:a3!
.
ii.
27. b3l"1:a3 2B. ii. b2l"1:a5 29.hxg4 Wxg4?!
29 . . . ttJa6!
30. ii.c3 Wg2 3 1 l"1:
. fll"1:a3 32. Wb5 ttJd7
8
7
6
5
4
3
2
1
a
b
c
d
f
e
h
g
6.e3
Karpov chooses a line which keeps rwo of the
bishops passive - his own on c 1 and Black's on
cB - rather than make both bishops active with
cxd5. In my opinion it is risky to pay attention
only to the c 1 -bishop as if 6. ii.g5 then . . . dxc4
is possible. This is not considered dangerous,
but more importantly such a move is not in
Karpov's style.
Karpov's only regular game with this line
resulted in a tremendously interesting battle:
6. ttJe5 ii.d7 7. ttJxd7 Wxd7 B.e3 ttJe4 9. '.Wc2
ttJxc3 1 O.bxc3 ii.e7 1 1 .cxd5 exd5 1 2. ii.d3 h6
1 3 . Wa4 a6 1 4 l"1:
. b 1 l"1:a7 1 5 .c4?!
1 5 .e4! looks even stronger.
1 5 . . . b5! 1 6.cxb5 axb5 1 7. Wxb5 0-0 1 B. ii.c2?!
l"1:xa2 1 9. '.Wd3 g6 20. ii.d2 l"1:a3 2 1 . We2 We6
22.h3 h5?!
After 22 ... l"1:eB Black's position is preferable.
a
b
c
d
e
f
g
h
33. ii.xd5!
The position is still very exciting and Karpov
managed to win it, Karpov - Lautier, Baden­
Baden 1 992.
6...0-0 7..id2 a6 S.a3 .ie7
It is a matter of taste whether in such a
position one prefers to simplifY with B . . . ii.xc3!?
9 . ii.xc3 ttJe4 or choose the game continuation.
On the other hand B . . . ii.d6?! is less reliable as
White scores well after 9.c5 .
9.'?9c2 dxc4
Vescovi plays a move which was played
before only once; he is aiming to free his pieces
with . . . e6-e5.
10.i.xc4 .id6 11.0-0 e5 12.d5!?
Taking on e5 would be inappropriate as it
would give Black freedom to develop, but 1 2.h3
comes into consideration. However Karpov
preferred to make the c7 -pawn backward rather
than accept an isolated pawn on d4.
a
b
c
d
e
f
g
h
The Prime Years
516
12...tlJe7 13.h3
Karpov stops . . . �g4 or perhaps Black might
also have been considering . . . It:lg4.
a
b
c
d
f
e
h
g
13....ifS
This was an exhibition match, so perhaps
Vescovi wanted to make the game more
interesting than if it was a play-off game in the
world championship. His move provokes e3e4, which loses control of the f4-square and
offers Black a chance to hit back at White's
e-pawn with . . . 0-fS. However, giving up a
tempo to achieve this is too high a price.
1 3 . . . bS! ?
Developing the bishop to b7 gives hopes of
equalizing.
1 4 . � a2 � b7 IS.e4
White defends the dS-pawn; now it is not easy
for Black to choose between several somewhat
worse, but close to equalizing continuations.
1 5 . . . h6
Black stops �g5 .
After 1 5 . . . c5 1 6. lt:ldl c4 1 7. lt:l e3 Elc8 the
position is complex, but perhaps White's
chances are somewhat favourable.
1 5 . . . c6 1 6. �g5! ( 1 6.dxc6 lt:lxc6 1 7. �e3 Ele8
is alright for Black) 1 6 . . . cxd5 (if 1 6 . . . lDg6
1 7.dxc6 �xc6 1 8 .Elfdl then White's initiative
on the c- and d-files is unpleasant) 1 7. �xf6
gxf6 1 8 . lt:lxd5 �xd5 1 9. �xd5 Elc8 2o. 1Wd3
It would not be not fun for Black to play
with doubled pawns.
1 5 . . . lt:lg6!? Black quickly prepares room
for the queen. For example, 1 6. �e3 Ele8
1 7.Elfdl We7 1 8 .Elac 1 ( 1 8.g3 c6) 1 8 . . . Elfd8
looks okay for Black.
1 6.Elfdl It:lg6
1 6 . . . c6?! looks premature because of 1 7 .dxe6
It:lxc6 1 8 . �e3.
1 7. �e3 Elc8
1 7 . . . lt:lh5!?
1 8 .Elac 1 We7 1 9.93 Elfd8
Black has caught up in development and is
ready to carry out . . . c7 -c6.
14.e4
Karpov was probably happy to play this
move, which strengthens his centre.
14....id7
a
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
2006
Anatoly Karpov - Giovanni Vescovi
517
lSJ3fdl
21.i.e2
Karpov makes room for his bishop and takes
into consideration that Black is playing for
. . . c7-c6.
Karpov quickly adjusts to the current
situation; after Black's last move White no
longer needs to worry about a sacrifice on
h3. Karpov is not threatening to take on e5
immediately, but at some point the loose
knight on h5 may be significant.
lS...llJg6 16.i.f1!
The bishop defends its king and clears the
c-file to prepare the attack on the backward
pawn.
16...�hS
Vescovi is confident of his kingside play and
rather burns his bridges; after this move Black
virtually says goodbye to any chance of getting
rid of the backward c-pawn by pushing it.
21...�h4 22.�a2
Karpov's last move defended the f3-knight,
so now he can start targeting the c7 -pawn.
Karpov wants to exchange the bishop on b4
without doubling his pawns. He could also
have played 22. ctJa4 b6 23 . .ib4 ctJxf3t 24 . .ixf3
.ieB 25. ctJ c3 when White is a bit better.
22...�xf3t
17.i.gS
Karpov spreads a little confusion among
Black's pieces.
At some point Black has to exchange on f3 .
23.i.xf3 Wl'gS?!
Vescovi now thinks that 23 ctJf6 was
stronger and if 24 . .ib4 :gf7 the Brazilian
grandmaster evaluates the position as unclear.
•••
17...Wl'c8
1 7 . . . f6!? IB . .i e3 f5 looks playable for Black,
as he would have a tempo more compared
with the game.
18.i.e3 f5
Black had another option: IB . . . ctJhf4!?
1 9 .:gacl 'lMfdB 20. ctJa4 b6 Black's play on the
kingside seems enough to keep the position
balanced with . . . 'lMff6 or . . . f7-f5 .
8
7
6
5
4
3
19J3acl f4 20.i.d2 Wl'd8
2
8
1
7
a
6
24.�f1!
5
Karpov keeps open the option of running
away to the centre to escape the attack. Black's
King's Indian style pawn storm will be far less
effective if the white king is not in the firing
line.
4
3
2
24...aS
1
a
b
c
d
e
f
g
h
b
c
d
e
f
g
h
The Prime Years
S18
Black stops i. b4. If instead 24 . . . i. bSt then
2S. �el when any check would kill the king,
but all Black's pieces are far away.
25.\Wd3!
Karpov again reacts instantly to his
opponent's last move. Now he takes advantage
of the fact that the a-pawn no longer controls
the bS-square and he wants to play lilc3-b 5 .
a
b
c
d
f
e
g
h
27.:9:c6!!
Karpov brilliantly sacrifices an exchange; in
return he wins the bS-pawn and the dS-pawn
transfers to c6. Such positional sacrifices work
when the opponent's rooks do not have open
files. The Brazilian grandmaster told me he
completely missed this move and I think it
damaged him psychologically.
a
b
c
d
e
f
g
h
25 h5?!
..•
Vescovi stops lilbS by pushing his b-pawn,
which is certainly stronger than 2S . . . :9:f6?
losing a piece to 26.g4.
Let me show you an example from my own
praxis, which illustrates that such sacrifices can
work even in positions without queens.
8
However, best of all was covering the bS-square
with the help of his rook by 2S . . . :9:a6!. Next
the rook can be employed in an original way:
26. lilc3 :9:b6 27.:9:c2 lilf6 2s. Wfc4 a4 29. �el
:9:b3 The position is unclear.
7
26.b4!
2
6
5
4
'-=;J"UU'__ =, /,N/," ', ;;,,'-;,
3
Karpov fixes the b5-pawn.
26...a4?!
Correct was 26 . . . axb4! . Even though
exchanging on b4 does not solve all Black's
problems, it should have been played because
it creates play for the black rooks. 27. i.xb4
lilf6 2S.:9:c3 :9:fcs 29. lilcl! White has time to
improve the knight. 29 . . . :9:a4 30. Wf c2 White
retains an edge.
a
b
c
d
e
f
g
h
23.:9:el
White hopes to win back the e4-pawn when
the position would be so closed that Black
might be unable to invade. White would
probably take with the bishop, but sacrificing
the exchange might also work.
23 . . . :9:f3!
2006
Anatoly Karpov - Giovanni Vescovi
My strong opponent was utterly surprised
when I played this exchange sacrifice. The
new f3-pawn secures plenty of space and
creates excellent winning chances.
24. ixf3 exf3 2 5 .tLl e3
25 .h4! ? was worth a try.
25 . . . e4
Black threatens . . . tt'ld3 and . . . ixe3 .
26. tt'l c2?!
After 26. tt'lg4 ig7 Black's advantage is
smaller than in the game.
26 . . . id2 27.:§:dl ixc3 2B. tt'ld4 ixd4
29.:§:xd4
519
Back to the main game.
a
b
c
d
f
e
g
h
27 J�ab8?
..
The local grandmaster prepares to sacrifice
an exchange on bS. This reminds me of other
sports, where players take an idea from the
opponent - his sacrifice generates the idea that
I can do it. However, in this case Black achieves
nothing and ruins his position.
a
b
c
d
e
f
g
h
29 . . . g5 !
Fixing the h2-pawn.
30.h3 �f7
Centralizing the king is important and
White has no remedy against it.
3 1 . W fl Wf6 32.:§:a3 We5 33 .:§:dl :§:fB 34. �el
:§:f6 3 5 . W d2 :§:h6 36. We3
A witty idea from Skembris who was
already short of time, but it can be tactically
refuted.
36 . . . :§:xh3 37.:§:aal tt'l d7 3B.:§:hl :§:xhl 39.:§:xhl
tt'lf6
The knight arrives in time.
40. Wd2 tt'lg4 4 1 .:§:xh7 tt'lxf2
The connected passed pawns give Black an
easy win.
42. �c3 tt'ldlt 43. �d2 tt'l b2 44. �e3 tt'lxc4t
4 5 . �f2 g4 46.:§:xc7 e3t 47. �el tt'lb2
0- 1 Skembris - Karolyi, Geneva 1 9B9 .
Regardless of how hard it would be to defend
Black's passive position, Black should take the
exchange this move or the next:
27 . . . ixc6 2B.dxc6
8
7
6
5
4
V///"//'''�='/'
"'-',%"''''%
3
2
a
b
c
d
e
f
g
h
Now there are two moves to consider:
2B . . . :§:fdB and 2B . . . Wg6!?
2B ... :§:fdB 29. WxbS tt'l f6 30. Wc4t WhB 3 1 . tt'lc3
ifB 32. ie 1 !
The Prime Years
520
Inaccurate is: 32. Wff7 �g6! (32 . . . .id6 33. tLlb5)
33. Wfxg6 hxg6 34 . .icl (34. We2 :ga6!)
34 . . . :gxdl t 35 . .ixdl :ga6 36.b5 :ga5 37 . .ib2
g5! Black has managed to tie White up.
32 . . . :gxdl 33 . .ixdl .i d6 34. tLlb5
34 . .ixa4 f3!
34 . . . tLle8 3 5 . .i c3
White has lovely compensation for the pawn
as the e5-pawn is vulnerable.
28 . . . Wfg6!?
This may be the best chance.
29. tLlc3 Wff7 30. tLlxb5 :gfd8 3 1 . Wfc3
After 3 1 . .i cl .i f8 32. Wfc2 Black is struggling
to avoid losing more material.
3 1 . . . tLlf6 32 . .i cl .i f8 33.:gxd8 :gxd8 34. Wgl
White will soon win a second pawn for the
exchange. Having collected enough material he
will start pushing his pawns, with reasonable
winning chances. Nevertheless Black still has
chances to hold the position.
28.ttlc3 �h4?
29..J3xb5
The exchange sacrifice hardly even slows
down Karpov.
30.�xb5 g5 3 1.�c4!
Karpov wants to take on c7.
31...g4
To illustrate the comment above, if 3 1 . . . tLlf6
then 32.:gxc7 wins.
32.hxg4Axg4
8
7
6
5
4
3
2
Black's attack contains little poison as he
has only a few pieces against a well-protected
king. Taking on c6 was weaker than before
27 . . . :gab8, but Black should still have tried to
take the exchange.
8
1
a
b
c
d
e
f
g
h
33.<;!;e2!
Karpov simple walks away with his king; he
has enough pawns to use as a shield.
33...ttlg3t
7
A desperate attempt to create chances; Black
had no other move that causes any problems.
6
5
34.fxg3
4
It is safe to take more material.
3
34...fxg3 35.i.e3 �hf3
2
1
a
b
c
d
e
f
g
h
Vescovi sacrifices the second exchange. It can
be unpleasant to receive such a "gift" if one is
short of time, but in this case it is far from
creating enough play.
29.ttlxb5
Black's queens ide falls apart.
36.gxf3 �h2t 37.<;!;d3 i.xf3 38J3el �h4
2006
Anatoly Karpov - Giovanni Vescovi
52 1
8
7
6
5
4
3
2
1
a
b
c
d
e
f
g
h
39J�x d6!
The simplest way to win is to go after the king. Time and again Karpov's games exhibit the
following pattern - he starts playing on one wing, his opponent starts on the other, and yet
Karpov finally wins by switching to the side where his opponent had the initiative.
39 ... cx d6 40JW c8t
Transferring the queen to the kingside will end all resistance.
40 ... � g7 41.� d7t � h8 42.� f5
1-0
Karpov won the match 2Yz-l Y2.
2006 Summary
Russian team championship: 1 /3 (+0 2
=
Total 33.3% (+0 2
=
-
-
1)
1)
I]l Wins • Draws • Losses
2007
Rating 2668 (37 in the world)
Karpov played three games in the Russian team championship. In the first he agreed a draw after
nine moves against Grischuk; the second was drawn in fifteen moves against Miton. In the third
he tried somewhat longer, nevertheless that game also ended in a draw against Timofeev.
After what had been a long break, Karpov entered a proper, regular tournament in Gorenje
Valjevo. In the first round Karpov drew a long game against Predrag Nikolic; strangely they
played an additional thirty moves after exchanging all the queenside pawns to reach a symmetrical
ending with four pawns versus four and two minor pieces each. In the next two rounds he drew
two largely uneventful games against Ivanisevic and Atalik, before facing Kiril Georgiev.
I Game 64 I
Anatoly Karpov
-
Kiril Georgiev
Valjevo 2007
l .d4 llJf6 2.c4 e6 3.llJ c3 � b4 4.� c2 d5 5.a3
Karpov obtains the advantage of two bishops. This was only the third time he had faced the
4 . . . d5 line; in the other two games Karpov took on d5 and eventually drew.
5 ...�x c3t 6.�x c3 llJ e4
8
7
6
5
4
3
2
1
The Prime Years
524
Georgiev bases his plan on fast piece-play; he
had already played this line once and that was
against none other than Kasparov.
7.WI e2 e5 S .m e5 tiJ e6 9.cx d5 ex d5 10.tiJa
Karpov plays the main line. A year later,
against the young Chinese star Hou Yifan,
Karpov deviated with:
20. ltJ f4 Wg5
20 . . . ltJ e5!? 2 1 . Wd4 ltJc6
2 1 .l"i:h 1 g6 22.l"i:d 1 We??
If 22 . . . a6 then 23. Wc3! , but Black could
have tried 22 . . . d4!?
23.b5 ltJe5 24. Wd4 f6 2 5 . ltJxd5 Wfl 26.e4
Karpov obtained a winning position and
made no mistake in Karpov - Hou Yifan, Cap
d'Agde (rapid) 200S.
lO.e3
He followed up with a very interesting and
exciting exchange sacrifice.
lO . . . � f5 1 1 . �d3 WIg5
a
b
c
d
e
f
g
h
10 ...WIf6
1 2.f3!?
Karpov sacrifices an exchange for a pawn
and a strong bishop on the long diagonal.
12 ... Wh4t 1 3 . c;t>fl ltJg3t 1 4.hxg3 �xd3t
1 5 . Wxd3 Wxh 1 1 6.b4 0-0 1 7. �b2 l"i:adS
Georgiev avoids the most common move,
which is 1 0 . . . � f5 .
Against Kasparov the Bulgarian grandmaster
played lO . . . Wa5t . The game continued:
1 1 . �d2 Wxc5 1 2. Wxc5 ltJxc5 1 3 . �e3 ltJe4
1 4 . ltJ d4 0-0 1 5 .l"i:d 1 �e6 1 6.f3 ltJ d6 1 7. ltJxe6
fXe6 lS. �c5 l"i:adS 1 9.e4 l"i:feS 20.exd5 exd5t
2 1 . <j;>f2 Kasparov had a slight edge and went on
to win, Kasparov - Ki. Georgiev, Thessaloniki
(01) 1 9S5.
1 l .e3.i g4 12.i
. e2
On 1 2. ltJd4 Black plays the excItIng
1 2 . . . � dl! of which there are four examples in
the database; Black scored fifty percent.
a
b
c
d
e
lS. c;t>f2 Wh6 1 9. 1tJe2 l"i:feS
1 9 . . . a6!?
f
g
h
12 ... 0-0
2007
Anatoly Karpov - Kiril Georgiev
525
Karpov attacks the b-pawn at once. It would
seem logical to play 1 7. <j;Jd2 then bring the
other rook to b 1 , but this would give Black
time to consolidate the position with 1 7 . . . tLla5
1 8 .�hbl �xc5 .
17... tLl d8
If 1 7 . . . tLla5 then 1 8 .�b5 wins.
18.i> d2!
a
b
c
d
e
f
g
h
13.i.d2
Karpov plays a novelty, which will generally
lead to a queenless middlegame.
1 3 .. .l:Ue 8
This is the most common move, but I would
prefer to go after the c-pawn at once with
1 3 . . . �fc8 or 1 3 . . . �ac8 .
14.i. c3
Karpov will temporarily have an extra
pawn. Instead after 1 4.0-0 tLle5 1 5 . tLld4 Axe2
1 6. tLlxe2 'i!f1g6 1 7.f3 tLlxd2 18. 'i!f1xd2 tLlc4
1 9 . 'i!f1d4 tLlxe3 20. tLlf4 the position is equal.
The king clears the first rank for the rook
and defends the c3-pawn j ust in case. If some
pieces are exchanged it may prove important
that the white king is already in the centre. The
alternatives were less convincing:
After 1 8 .c6 �xc6 1 9. tLld4 �g6 20. Axg4 White
is j ust a fraction better.
And certainly not 1 8 .�b5 ? a6 1 9 .�a5 Axf3
20. Axf3 tLle6 2 1 . Axd5 tLlxc5 when Black traps
the rook on a5 .
18 ...� e7 19.� b4 i. e6?!
If after 1 9 . . . Ad7!? the play continued as in
the game, then Black would be a tempo up.
Black could take advantage of this after 20. tLld4
�xc5 2 1 .a4 �a5 when the position is equal.
14 tLlx c3 15.'i!f1x c3 'i!f1x c3t 16.bx c3 � ac8
..•
8
8
7
7
6
6
5
5
4
4
3
3
2
2
1
a
17.� bl!
b
c
d
e
f
g
h
1
a
b
c
d
e
f
g
h
The Prime Years
526
2 1 .a4!
This endgame is objectively balanced, but
Karpov can still exert pressure. The more pieces
are exchanged, the more one should be careful
about placing pawns on the same colour as the
opponent's bishop. However, principles can
clash with reality, and with 2 1 .a4 Karpov gains
space and starts to push back his opponent on
the queenside.
2 1 ...i. d7?!
At present, it is hard to spot the drawback
of this move, but as we shall see, blocking the
seventh rank is significant. Black had numerous
alternatives:
2 1 . . .a5 ? Black tries to fix the a4-pawn as a
target, but in fact the a5-pawn is the real
weakness. 22.Ei:b6 i.d7 23 .Ei:d6! 'it>f8 24.Ei:bl
me8 2 5 . 4J b3 Black loses a pawn.
2 1 . . .Ei:ec7 Going after the c-pawn at once
also fails to solve Black's problems. 22. 4Jb5
a5 23. 4Jxc7 axb4 24. 4Jb5 White keeps some
advantage.
2 1 . . .b6 22. 4Jb3 Ei:c8 23.a5 4Jc6 24.Ei:b5 bxa5
2 5 . 4Jxa5 4Jxa5 26.Ei:xa5
the pawn. 27.Ei:bl! m f8 (27 . . . Ei:xc3 loses to
28. i.a6) 28. i. a6 Ei:c5 29.Ei:a3 The endgame
is very difficult for Black.
27. i.d3 Ei:b2t 28. i. c2 i.f5 29.Ei:cl
Black loses a pawn; he is active but a pawn
is a pawn.
Best of all was: 2 1 . . .g6! 22.Ei:al mg7 23.a5
4Jc6 24. 4Jxc6 (if 24.Ei:b5 Ei:xb5 2 5 . i.xb5 4Jxd4
26.cxd4 Ei:c7 Black has nothing to worry
about) 24 . . . Ei:xc6 25 .Ei:abl Ei:cc7 I do not see
how White can make any progress.
22J� al!
Karpov of course knows that rooks belong
on open files, and yet he puts the rook behind
his pawn. Of course, the reason is that he wants
to push is further.
22, ..g6
Georgiev puts his pawn on the same colour
as White's bishop, but also takes away the
important f5-square from White. There are
still many pieces on the board and White is
far from being able to fix the pawn. If more
pieces are swapped then Black still has time
to reshape his pawn structure by placing his
pawns on dark squares.
23.a5!
Karpov not only gains space but creates a
threat.
8
7
6
5
a
b
c
d
e
f
g
h
26 . . . Ei:b7
Black might be tempted to demolish the
c-pawn with 26 . . . Ei:ec7 and hope to hold
the position, but he has no time to take
4
3
2
1
a
b
c
d
e
f
g
h
2007
Anatoly Karpov - Kiril Georgiev
23 a6
•••
Georgiev faced a difficult decision between
several