Transpo Tricks in Chess
Finesse your Chess Moves and Win
Andrew Soltis
BATSFORD
First published in the United Kingdom in 2007 by
Batsford
10 Southcombe Street
London Wl4 ORA
An imprint of Anova Books Company Ltd
Copyright © Batsford 2007
Text copyright © Andrew Soltis
The moral right of the author has been asserted.
All rights reserved. No part ofthis publication may be reproduced, stored in a
retrieval system, or transmitted in any form or by any means electronic,
mechanical, photocopying, recording or otherwise, without the prior written
permission of the copyright owner.
ISBN 9780713490510
A CIP catalogue record for this book is available from the British Library.
15 1413 1211 10 09 08 07
10987654321
Reproduction by Spectrum Colour Ltd, Ipswich
Printed and bound by Creative Print & Design, Ebbw Vale, Wales
This book can be ordered direct from the publisher at the website
www.anovabooks.com. or try your local bookshop.
Distributed in the United States and Canada by Sterling Publishing Co., 387 Park
Avenue South, New York, NY 10016, USA
Contents
Page
Introduction
5
Chapter One:
Double KP Openings
14
Chapter Two:
Ruy Lopez
34
Chapter Three: Sicilian Defense
54
Chapter Four: Semi-Open Games
94
Chapter Five:
Double QP Openings
120
Chapter Six:
Indians
153
Chapter Seven: Dutch Defense and Flank Openings
191
Index of Opening Variations
213
3
4
Introduction
Openings have become so
complex and convoluted that we've
forgotten our basic goal in them.
The reason we pick, say, 1 e4 over
1 d4 - or 11 h3 rather than 11 .te3,
for that matter - is simply to reach a
middlegame we want to play.
Unfortunately, our opponents are
making choices too, in order to
reach the middlegame they want
to play. It's rarely the same
In 99 of 100 games Black
continues 5... ttJd6 or 5... .te7 and
life goes on.
middlegame.
But lurking in the warren of
footnotes of 'book' is 5... a6!? It's
How do you get the one you usually followed by the comment
want? Unfair as it may be, you can't 'If 6 .ta4, then 6... b5 7 .tb3 d5
rely on your legendary positional transposes to the Open Defense. '
skill, your better-than-Fritz calculating ability or your iridescent
personal charm. What you can use
is trickery - the trickery of
transposition.
Since it's a footnote we take little
notice. After all, 5 ... a6 'just
transposes.' That tells us it doesn't
really matter because it only leads
to some other opening, something
irrelevant on that page.
Consider this opening: 1 e4 e5
2 ttJO ttJc6 3 .tb5 ttJf6 4 0-0 ttJxe4
5 d4, the main line of the Ruy
Lopez's Berlin Defense.
But 5... a6 has been a valuable
weapon for players from Paul
Morphy to Vasily Ivanchuk. Its
power lies in how it gets Black
5
Introduction
where he wants to go - to the Open
Defense - and avoids what he wants
to avoid - the Exchange Variation
(3 ... a6 4 .txc6).
We don't judge transpositions by
the same standard as we do other
moves. An original opening idea, a
TN as they're called, is evaluated
by the new position it creates. But a
Finesses like that are rarely transposition by definition reaches
appreciated except when they make an old position, as 5 a3! brilliantly
new 'book'. That was the case in did. Instead, you should judge it by
Botvinnik-Capablanca, A.Y.R.O. its effects, especially:
1938, an instantly famous game that
How it degrades your opponent s
began 1 d4 ttJf6 2 c4 e6 3 ttJc3 .tb4
choices or improves your own
4 e3 dS.
A basic strategy in any opening is
to increase the options at your
disposal. Take this familiar
position.
White's S a3 .txc3+ 6 bxc3
seems obvious today. But in 1938 it
was a masterstroke. 'The idea of the
move is typically modem - to
transpose into a favorable variation
which would not be reached in any
normal manner,' Reuben Fine
wrote.
This is a 'tabia,' that is, a standard
starting point in a major opening. In
this case it's a tabia that has served
as the launching pad for thousands
of Dragon Sicilians, which often
continue 10 ... ttJe5 and .. J:tc8/
... ttJc4.
In truth, there is a normal
manner, the discredited 4 a3 .txc3+
But in a 1997 game, Anand5 bxc3 d5?!, which allows White to
rid himself of a doubled pawn. Ki.Georgiev, Black tried 10...
Botvinnik used 5 a3! to trick ttJaS!? White quickly appreciated
Capablanca into that favorable the difference: After the natural
11 .tb3 Black could transpose into
version of 4 a3.
6
Introduction
more familiar lines with 11..J1c8 slow moves often prove fatal, as in
and 12 ...tbc4 13 i.xc4 l:hc4. But this case:
10 ... tba5 gives Black an extra
option, ... tbxb3+, that he may
employ depending on White's next
few moves. In other words, Black
gets to choose whether he wants to
transpose with ...tbc4 or not.
Instead of trying to figure out
how dangerous ... tbxb3+ would be
White made a practical choice,
11 i.e2!. This is a counter-finesse.
It takes away Black's extra option
13 g4! bS 14 h4 e6 1S a3 hS
and leaves the knight with nothing 16 i.gS hxg417 hS! gxhS 18l:1xhS
better to do than go to c4.
l:1cS 19 fxg4 l:1xc3 20 i.xf6 'iixf6
This had a bonus effect because a 21 'ifxc3 tbc4 22 i.xc4 bxc4
move order finesse can also be 23 'ifxc4 and White won.
judged by:
This shows how transpositions
How it unnerves or confuses your play tricks not just with move order
opponent
but with your opponent's equanimObjectively, 11 i.e2 is no better ity. When he realizes he is being
than 11 i.b3. But psychologically dragged into your middle-game, he
it was a potent blow - and may lose the nonnal composure that
transpositions typically have players enjoy in the opening, when
greater psychological power than they confidently rattle off the first
15 moves. Players who lose their
objective strength.
confidence make mistakes.
After 11 .. J1c8 12 ~b1 Black
couldn't bring himself to play the
Let's go back to 1 e4 eS 2 tbfJ
best move, 12 ... tbc4, because it tbc63 i.bS tbf6 4 0-0 tbxe4 S d4
would create the middlegame White a6. Giovanni Vescovi was rated No.
wanted to play. Black had more or 60 in the world when he first saw
less decided, when he passed up that position from the White side, in
10 ... ~e5, that he didn't want that 2005. He decided not to be tricked
middlegame.
into the Open Defense.
So he chose a very different
policy, 12... a6? In the Dragon such
But that meant choosing
6 i.xc6?!, which turned out to be a
7
Introduction
prelude to a worse idea, 6... dxc6
7 'Yi'e2 .ltf5 8 g4? .ltg6 9 h4.
Anand
began
calculating
furiously, trying to find out what
was wrong with White's move. But
there's nothing wrong with it. It's
Black could have refuted that just unfamiliar. After spending two
with 9...'Yi'd7! 10 ~xe5 'Yi'xd4, as of his precious five minutes, he
Johannes Zukertort played way played 4... d6 5 ~f3 d5 and reached
back at London 1883 (!).
a book position.
Vescovi spent 40 minutes to find
This was a case of a move whose
that dubious line. That leads to a major benefit was simply to give
another criterion of a transposition. the other player something to think
It can be measured by:
about. Typically, these moves do
not reduce his options. Rather they
increase them, giving him more to
consider.
How it gets your opponent to
think
The real battle of the opening
begins when you can force your
opponent out of his book knowledge. Only then does he risk
making errors and spending costly
minutes.
There really isn't much value, for
example, to 1 d4 c6 and then
2 c4 d5 compared with the normal
route, 1 d4 d5 2 c4 c6. But this
order, employed by Anatoly Karpov
among others, gets White debating
with himself over whether he
knows more about the Caro-Kann
(2 e4) than the Slav. Or what he
would do about 2 c4 b5!? A strange
move order can do that.
That was case when Vishy Anand
was Black in a speed playoff game
at a big-bucks event in 1994. In the
Petroff Defense, 1 e4 e5 2 ~f3 ~f6
3 d4 ~xe4, his opponent failed to
play 4 .ltd3, the move considered
virtually automatic. He played
4 ~xe5!? instead.
A 'something to think about'
move can cost your opponent more
8
Introduction
And there's a fourth way of
evaluating a crafty transposition,
by:
than minutes. It can prompt a bad
decision. The most drastic recent
example of that befell Vladimir
Kramnik in what was then the most
important game of his life, the final
game of a 1994 Candidates match.
How it preserves your mental
health
We all have to deal with an everHis opponent, Boris Gelfand, expanding amount of book analysis.
opened with 1 c4 and there Almost as bad as trying to
followed 1..• c5 2 liJc3 liJf6 3 g3. memorize all that at home is trying
This pOSItIOn had occured to remember it at the board. This
gazillions of times before - but can be maddening.
never to Kramnik. As routine as
We'd love to cut down the
3 g3 was, it confused him. He amount of book we need to know
replied 3... d5 4 cxd5 liJxd5 and and still reach the middlegames we
then on 5 .i.g2
like. The best ways to save our
midnight oil - and our sanity - is
through transpositions. Consider
the main line of the Winawer
French, 1 e4 e6 2 d4 d5 3 liJc3 .i.b4
4 e5 .
... he played 5... e6?? 'almost
without thinking,' according to his
opponent. Kramnik was assuming
he would transpose, after 6 liJf3
liJc6, to another tabia that is quite
good for Black.
The familiar path is 4 ... c5 and
then 5 a3 .i.xc3+ 6 bxc3 liJe7. But
But 6 liJxd5! cxd5 7 'ii'b3 won a White has numerous sidelines such
pawn (7 ... c4 8 'iib5+), the game and as 5 dxc5, 5 .i.d2, 5 'iVg4 and 5liJf3.
the match. The confusing effect of Theory regards these as not quite as
3 g3 set back Kramnik's world good as 5 a3. But in practice they
championship aspirations for are dangerous to an ill-prepared
Black.
several years.
9
Introduction
Rather than spend hours studying
them. Black can play 4...liJe7!. This
used to be purely a prelude to 5 a3
i.xc3+ 6 bxc3 b6. But Wolfgang
Uhlmann showed that 6... c5! was a
simple transposition to the 4 ... c5
main line. Black gets the benefits of
the normal move order without
having to agonize over the 4 ... c5
sidelines. It's the 'mental hygiene'
move order.
Every transpositional trick has to
be judged by a balance sheet: What
are the benefits? What are the
drawbacks? Which order counts
more? Some, like 4 ... liJe7, may
have more plusses than minuses.
Other's like Kramnik's 5... e6?? are
disastrous.
This is another book POSItIon
from the Keres Attack - but White
has lost a move because he spent
two tempi to get his h-pawn to h4.
A lost tempo should make a huge
difference.
But what mattered most was
Black's unfamiliarity with Kereslike positions. After 9...liJde5?!
10 i.e3 b5?! she was worse and
after 11 liJxc6 liJxc6 12 'ifd2 i.b7
13 f4 ~c7 14 'ifflliJb8 15 a3liJd7
16 f5 liJe5? 17 fxe6 fxe6 18 i.h3
she was lost. White's objectively
bad 8 g5? worked brilliantly as a
transposition. It was really 8 g5!.
But the vast majority are
somewhere in between. The bottom
line is not whether they give you a
superior position but whether it's a
position you want to play. Ideally,
it's also one your opponent wants to
avoid.
'IT JUST TRANSPOSES'
Consider 1 e4 c5 2 liJf3 liJc6 3 d4
cxd4 4 liJxd4 liJf6 5 liJc3 d6 and
now 6 h3 e6 7 g4. This transposes
into a conservative version of the
Keres Attack, in which White relies
on a slow buildup with i.g2, rather
than g4-g5.
Some opening positions can
come about from two, three or more
logical orders. Books - and even
experienced players - tend to
dismiss an alternative route because
'it just transposes'. But each route
is likely to have benefits and
liabilities that a shrewd transposer
knows to evaluate. Take this
However, Andreikin-Kosteniuk,
Tomsk 2006 went 7... a6 8 g5liJd7
9 h4.
10
Introduction
position from the Taimanov
Sicilian. It's White's move and
when it was analyzed, by Mark
Taimanov, in the 1972 edition of
ECO he claimed Black wins.
realizes Black isn't going to allow
10 e5. So he'll be aware that Black
is planning to transpose, with
9....tc5! to the diagram, and may
look instead for alternatives such as
9 h3.
The third route, S....te7!?, is the
most deceptive. An experienced
Sicilian player in White's chair
would recognize that the thematic
move is 9 f4. He would see that the
naturaI9 ... d6 transposes into a book
Scheveningen. So he'll play 9 f4 allowing Black to reply 9 ... .tc5 and
get where he wanted to go. Of the
three, 8... .te7 may be Black's best
route to the diagram.
In fact, White has a good reply,
10 .!Df5!. The critical line is
10....!De7!, which forces 11 .!Dxg7+
~f8 12 .txc5 'i'xc5+ 13 ~h1
~xg714 e5.
THE LURE OF THE
FAMILIAR
Whether this is sound is in
There are players, like Uhlmann,
dispute and that means both White Lajos Portisch and Mihai Suba,
and Black may be interested in who used finesses of move order
reaching the diagram. There are solely to reach the middlegame they
three quite reasonable routes to it.
want. There are others, like Bent
One is 1 e4 c5 2 .!Df3 e6 3 d4 Larsen, who also use them as
cxd4 4 .!Dxd4 .!Dc6 5 .!Dc3 'i'c7 'confuse-moves' to pose puzzles or
6 .te3 a6 7 .td3 .!Df6 S 0-0 and to get their opponents to think. And
now S....tb4 9 f4 .tc5. But if Black then there are tricksters.
seeks the diagram this order is
The trickster looks for the crafty
seriously flawed: White has a way to reach the middlegame he
strong alternative in 9 .!Dxc6! and wants. He knows, for example, that
10.!Da4.
when opponents are confronted
with an unfamiliar move, they are
strongly, even irrationally, tempted
to look for a way to reach a
recognizable position.
The second route is forcing.
Black attacks the h-pawn with
S....td6. White's instinct is to meet
the threat with a threat, 9 f4. But he
11
Introduction
Laszlo Szabo fell victim to that
temptation in the 1953 candidates
tournament when his game with
Paul Keres began 1 d4 d5 2 ttJf3
ttJf6 3 c4 dxc4 and then 4 ttJc3 and
4••• a6.
2 ttJf3. He knows that books
recommend 2 d4!. But about half of
all 1 e4 ttJc6 games continue 2 ttJf3
because White wants to transpose to
the familiar (2 ... e5).
Similarly, when a 1 d4 player
faces 1...d5 2 ttJf3 ttJc6 3 c4 e5?!,
he is likely to transpose into the
Albin Counter Gambit, 4 dxe5?!
even though he suspects 4 ttJxe5! is
better.
The trickster exploits that foible
of human nature. The rest of us
have to be willing to punish him for
taking such liberties. For example,
1 e4 c5 2 ttJf3 ttJc6 3 d4 cxd4
4 ttJxd4 ttJf6 5 ttJc3 g6.
That was unfamiliar to him. But
he remembered getting a very good
game once with 4 'ifa4+ ttJbd7
5 ttJc3 a6. After studying the
position for a while he played
5 'ifa4+, seeing that 5... ttJbd7
would transpose.
A stunned Keres took 15 minutes
to make sure he wasn't dreaming.
Then he played 5... b5! and won
easily (6 ttJxb5? Jid7).
A much more common, and less
painful, error occurs when a good
player takes the easy way out of an
unfamiliar position and transposes
to a recognizable one. He may do
this even if he suspects - or knows
for certain - that he is playing an
inferior move.
This order gives Black all the
tactical benefits of an Accelerated
Modem Dragon without allowing
the Maroczy Bind as the Accelerated does. But there is a big minus 6 ttJxc6 sentences Black to either a
dubious middlegame (6 ... bxc6 7 e5)
or a no-fun endgame (6 ... dxc6
7 'ifxd8+).
You've seen this happen when a
1 e4 player meets 1...ttJc6 with
12
Introduction
Nevertheless the trickster will
weigh the risk of being squeezed to
death in a Maroczy Bind against the
risk of White playing 6 tiJxc6!. The
odds are on his side. A database
check finds White played 6 tiJxc6!
less than a third (327 of 1,000
games) of the time. Even superGMs like Bobby Fischer and
Mikhail Tal played the inferior
6 i.e3 to return to a position they
knew.
Examples like that should tell
you to beware those words 'it just
transposes.' There is usually a plus,
a minus, an extra option to consider.
If you don't, you are allowing your
opponent to control the direction of
the opening. And the price may be
ending up in the middlegame he
wants.
13
Chapter One:
Double KP Openings
Trickery begins with the oldest
openings, like the King's Gambit
Accepted. Today's theory says the
KGA (1 e4 e5 2 f4 exf4 3ltJf3) is no
longer dangerous because of 3... d5
and then 4 exd5 ltJf6 with equal
chances.
Breyer Gambit, 1 e4 e5 2 f4 exf4
3 'fif3 d5 4 exd5. But the Breyer is
harmless and Black can even try the
sacrificial 4 ... ltJf6 5 J.b5+ c6!?,
recommended by Yacov Estrin.
This order can upset a Bishop's
Gambiteer. There are few 1 e4 e5
2 f4 exf4 3 .tc4 players who are
comfortable in KGA positions,
which they would reach by playing
4 ltJf3 in the diagram.
If that makes Black happy, he
should consider the 'something to
think about' order of 1 e4 e5 2 f4 d5
3 exd5 exf4!?, as played by Anatoly
Karpov and Artur Yusupov.
So they'll look at 4 J.c4 and
realize it transposes into an ugly
Bishop's Gambit - 1 e4 e5 2 f4 exf4
3 J.c4 and then 3... d5 4 exd5 rather
than the recommended 4 J.xd5.
That's a psychological trap
because White doesn't really stand
badly after 4 exd5!? ltJf6 5 lLlc3
and then 5... c6 6 d4. He has simply
transposed into the main line of the
normal Bishop's Gambit, that is
1 e4 e5 2 f4 exf4 3 J.c4 ltJf6 4 ltJc3
c6 5 d4 d5 6 exd5. But it takes a lot
of clock time to figure that out over
the board.
What is there to think about?
Well, for starters, White wonders
whether he should head into the
book line with 4 ltJt3 ltJf6 or try
4 'ift3, which transposes into the
14
Double KP Openings
If Black doesn't like the ... d5
KGA or wants more than equality
he can aim for this:
The natural 4... g5 falls into a
trap because 5 h4 transposes into
a better-than-usual version of
Allgaier Gambit (5 ... g4 6 liJg5 h6
It's a tabia that's been studied 7 liJxf7 'l;xf7, rather than 6... f6
since the 1860s and today is 7 liJh3 gxh3 8 'ii'h5+).
regarded as excellent for Black
In contrast with the normal
(8 g3 i.h3! 9 :f2 liJf6! or 9 gxf4
Allgaier,
1 e4 e5 2 f4 exf4 3 liJf3 g5
'i'd7! 10 :f2liJf6).
4 h4 g4 5 liJg5?! h6 6 liJxf7 'l;xf7
But Black can't force his way to 7 i.c4+ d5!, White has an extra
it from the traditional 1 e4 e5 2 f4 move, liJc3, which is better than
exf4 3 liJf3 order. He needs to use Black's extra ... d6?!.
misdirection.
Viktor Korchnoi claimed that
4... h6! favors Black in the diagram.
It may be his best but after 5 d4 g5
6 g3! we've reached a position that
often occurs in KGA lines and the
evaluation is at best fuzzy.
Bobby Fischer's try was 3... d6
and then 4 i.c4 h6! 5 d4 g5 6 0-0
i.g7 7 c3 liJc6 reaches the tabia.
The problem with 3... d6 is that
4 d4 threatens 5 i.xf4 before Black
For example, 6... fxg3 7 hxg3
can put his kings ide house in order i.g7 8 i.c4 and 8... i.g4 9:f1 'i'd7
with both ... g5 and ... h6. His best is 10 'iVd3, Gallagher-Jurgens, Bad
4... g5 5 h4 g4 after which 6 liJg5? Worishofen 1994. Remember the
f6! is unsound but 6 liJgl leads to 6 g3 position. We'll see it again.
double-edged play.
Black can try to improve on the
There is no objectively better Fischer defense with the neglected
move than 4 d4. But there is a more 3.•. h6, again seeking that good-forconfusing one, 4liJc3!?
Black tabia. This can confuse White
15
Double KP Openings
unless he knows that 4 d4 gS and
S ttJc3! d6 6 g3! reaches the fuzzy
line.
BISHOP'S GAMBIT
If your opponent is a Bishop's
Gambiteer, you have good reason to
fear he's better prepared than you, a
lot better. The careful transposer
will answer 1 e4 e5 2 f4 exf4 3 .tc4
with the forgotten favorite of
Mikhail Tchigorin, 3... ttJc6.
That leaves us with 3•.• g5. The
Romantic-era continuation was
4 .tc4 g4 but 4••.ttJc6! clears a path
to the tabia.
Black invites S 0-0 .tg7 6 d4 h6
and so on. Once White plays 4 .tc4
he can't insert h2-h4 effectively
because S h4? g4! 6 ttJgS
boomerangs badly after 6... ttJeS!
7 .tb3 h6 8 d4 hxgS 9 dxeS .tg7.
Black can then meet the natural
4 ttJf3 with 4•.. g5!, reaching that
good form of the KGA that we
considered in 3 ttJO gS 4 .tc4
ttJc6!. This is a psychological plus
because, as noted before, Bishop's
Gambiteers are often uneasy in a
KGA.
The best answer to 3 ... gS is
supposed to be 4 h4, which rules
out the tabia after 4 ... g4 S ttJeS.
However, the analysis of that
position is immense and White may
be tempted by the confuse-move
4 d4. Then 4 ... g4 S .txf4!? gxf3
6 'iVxf3 is an unusual Muzio-like
line that served Alexander
Morozevich well.
The bonus is that White has some
natural but bad responses, such as
4 ttJc3? which allows a strong
4 ... 'iVh4+ S ~f1 .tcS. Whether
White has an edge after 4 d4 ttJf6
S eS dS! is murky (6 .tb3 ttJe4).
VIENNA GAME
The safe response is 4 d4 h6
since S h4 .tg7 6 g3 g4! is
promising for Black. But S ttJc3!
and S... d6 6 g3 puts us in the fuzzy
position again.
There are two types of Vienna
players. The first don't want to
memorize volumes of Ruy Lopez
theory. The second is booked up
16
Double KP Openings
and prepared to inflict volumes of
Vienna theory on you, such as with
1 e4 eS 2 lbc3 lbf6 3 ii.c4 lbxe4
4 'iih5.
defend - any- cramped -posi tion
reputation but it's based on White
expanding with c2-c3 and d2/d4 or
'iib3.
Against either kind of opponent
it may pay to be devious as early as
2... d6!?
Expansion like that isn't possible
here and it's doubtful he has more
than an optical edge after 5 0-0 lbf6
6 d4 ii.g4! or 5 d4 lbf6 6 h3 0-0
7 0-0 lbd7 8 d5 lbcb8 (MovsesianEfimov, Pula 1997).
This looks prohibitively passive.
But White has to be careful since
3 f4 exf4 and then 4 lbf3 h6! S d4
g5 is that fuzzy KGA again.
There's much more experience
with the other 'beginner's move,'
2 ••.ii.c5, which has been put to good
use by Karpov, Bent Larsen and
Vasily Smyslov. Books used to
claim Black is already worse in
view of 3 lbf3 lbc6 4 lbxe5! lbxe5
5 d4.
White's best is probably 4 d4 or
4 ii.c4. But few Vienna players are
familiar or comfortable with KGA
lines in which Black gets to play
... 'iih4+.
Well, what about 3 ii.c4 ? Then
3 ... lbc6 leaves White to decide
whether to allow 4 d3 lbaS!? or
plunge into another unfamiliar
KGA (4 f4 exf4 5 lbf3 h6 or 5 ... g5
6 h4 g4 7lbgS lbeS).
But Black has a good alternative
in 3 ..• d6! since 4 d4 exd4 S lbxd4
lbe7 or 5 ... lbc6 heads toward a
reasonable version of the Scotch
Game after 6 ~e3, 6 lbb3 or
6lbxc6 'tWf6.
Instead, he may assume that
simple development, 4 lbf3, will
punish 2 ... d6. That transposes into a
Hungarian Defense after 4...ii.e7!?
The Hungarian has a stodgy, I-can-
For example, 5 ... lbc6 6 ~e3
ii.b6 7 lbdS lbf6 8 lbxb6 axb6 9 f3
0-0 10 ii.c4 d5 !. Or 10 c4 l:te8
lllbc2lbd7 12 'i'd2lbde5 13 ~e2
~e6 14 b3 f5 with good play for
17
Double KP Openings
Black in Zarnicki-Bianchi, Buenos
Aires 1989.
White can stay in Vienna mode
with 3 f4. Then 3... d6 is a King's
Gambit Declined in which White
has forsaken c2-c3, his most
ambitious plan in the KGD, and
books disagree about his chances
for advantage.
There is also 3 g3, which
transposes into a main line of Louis
Paulsen's variation. Black's only
minus is relinquishing the bookrecommended defense (I e4 e5
2 ttJc3 ttJf6 3 g3 d5!). So 2... iLc5
has no major drawbacks or benefits
and is more of a 'something to think
about' move.
CENTER GAME
Most players who meet 1 e4 with
1... e5 know that 2 d4 exd4 3 'ir'xd4
ttJc6 4 "iVe3 is neutralized by
4•..ttJf6 5 ttJc3 ii.b4 6 iLd2 0-0.
That's about all that anyone
remembers. But White has a
confuse-move in 5 iLd2.
This often has the effect of
scaring Black into another defense,
such as 5 ... iLe7 6 ttJc3 0-070-0-0that is, transposing into the less
ambitious 5 ttJc3 iLe7 6 iLd2 0-0.
But this is primarily a bluff
because if Black meets 5 iLd2 with
5... iLb4! White has nothing better
than transposing into the book line
with 6 ttJc3!.
The other finesses in the Center
Game arise if Black is afraid of a
line that hardly anyone knows
about. This is 5 e5!?, rather than
5 ttJc3 or 5 iLd2. That seems to just
lose a pawn but 5... ttJg4 6 'ife4
ttJgxe5? allows 7 f4.
To play 4 ... ttJf6 with confidence
Black should know a little about
crazy lines like 6... d5 7 exd6+ iLe6
and then 8 iLa6 or 8 iLe2! ttJf6
9 dxc7 'ir'xc7.
But he can save himself a lot of
worry and/or study by picking
another fourth move. One is
4 ..•iLb4+, with the idea of reaching
the equalizing book line after 5 ttJc3
ttJf6. The drawback is Black may
have to face White's extra options,
such as 5 c3 iLa5 6 iLc4 or 6 'ifg3.
Black's other alternative in
4 •••iLe7.
Then 5 ttJc3 ttJf6 transposes into
that 4 ... ttJf6 5 ttJc3 iLe7 line,
as Boris Spassky and Alexander
Alekhine did.
18
Double KP Openings
This succeeds after 3... c5 4 c3
dxc3? 5 tbxc3 when he's transposed
into 1 e4 e5 2 d4 exd4 3 c3 dxc3
4 tbxc3 c5? 5 tbf3, a ludicrous
defense to the Goring Gambit. (But
3... c5 4 c3 d5 or 4 ... ife7 in the
diagram are virtually untested.)
Another version of the Goring
arises after 3... i.b4+ and then 4 c3
dxc3 5 tbxc3. This forgotten
variation is quite playable after
5...tbc6 6 i.c4 d6 7 tbg5 .Jl.xc3+
8 bxc3 tbe5 and now 9 ..tb3 h6
10 f4! hxg5 11 fxe5 'iie7 12 ifd5
dxe5 13 ..txg5!.
But there's another wild card for
Black to worry about, 5 'iVg3 and
then 5...tbf6 6 e5, which Paul Keres
said favors White, or 5... iu6 6 tbc3
tbge7 7 .td2 as in ZiemackiFatyga, correspondence 1992
(7 ... d6 8 0-0-0 .te6 9 f4).
The main benefits to 1 e4 e5 2 d4
So Black has to decide whether
exd4
3 lbf3 are forcing Black to
the drawbacks of 4 ... .tb4+ or
4 ... .te7 outweigh the merit of start thinking at move three and
preventing him from playing a
avoiding the rare 4 ... tbf6 5 e5!?
Petroff, e.g. 3...lbc6 4 lbxd4 or
3... .tc5 4 tbxd4 tbc6.
DANISH GAMBIT
The Danish is a blood relative of
the Goring and Scotch Gambits.
The Danish has the poorer
reputation of the three but White
can try to improve with 1 e4 e5 2 d4
exd4 and now 3 tbf3!?, rather than
the Danish 3 c3.
Black's chief alternative is
3 ...lbf6. Then 4 e5 is another
Petroff line, I e4 e5 2 tbf3 lbf6 3 d4
exd4 4 e5, that Black has been
avoiding for years (by playing
3... tbxe4!). It's not a bad line but
few Black players will know
the theory and that makes 3 tbf3
a worthy weapon against a
Petrofessional.
BISHOP'S OPENING
Independent lines in the Bishop's
Opening are somewhat rare today.
Instead, 2 ..tc4 is used primarily as
a route to the Vienna or Giuoco
19
Double KP Openings
Piano that allows White to seek or published theory. The primary
avoid specific Vienna and Giuoco virtue of using 2 .i.c4 to reach those
positions (2 ... 4Jf6 3 d3 4Jc6 44Jf3)
positions.
is that White avoids the Petroff.
He benefits from 2 .i.c4 if he
likes 2... 4Jc6 3 4Jc3 4Jf6 4 d3 or
PHILIDOR'S DEFENSE
2... 4Jf63 d3 4Jc6 4 4Jc3 but doesn't
want to try to get there via the
Books used to ridicule the
Vienna because of the chaotic 2 Philidor because of White's space
4Jc3 4Jf6 3 .i.c4 4Jxe4!?
advantage. But young masters are
Note that after 1 e4 e5 2 .i.c4 challenging that view. To them the
key issue is how to reach this tabia.
4Jf6 3 d3 and the natural 3....i.c5:
Then 4Jc3 gets White into a
Vienna without having to face
2 4Jc3 4Jf6 3 .i.c4 4Jc6 4 d3 .i.b4 or
4... 4Ja5. Black retains those options
with the more precise 3•••4Jc6!.
There's a reason they say 'Knights
before bishops.'
The traditional order, 1 e4 e5
2 4Jf3 d6 3 d4 4Jd7, named after
the U.S. Civil War major James
Hanham, allows White to win the
two bishops, 4 .i.c4 c6 5 0-0 .i.e7
6 dxe5! dxe5 7 4Jg5 .i.xg5 8 'ii'h5.
It has many of the benefits of a
If Black is concerned about
Aron Nimzovich helped popularThe other direction White can go ize another sequence, 1 e4 e5
after 2 .i.c4 is towards the Pseudo- 2 4Jf3 d6 3 d4 4Jf6. This is more
Lopez. That is a Ruy Lopez-like likely to get Black to the tabia
system that can begin 1 e4 e5 2 4Jf3 (4 4Jc3 4Jbd7 5 .i.c4 .i.e7) because
4Jc63 .i.c4 followed by 4 d3, 5 c3 few opponents are willing to wade
and 4Jbd2, 0-0, .i.b3 and eventually into the 4 dxe5 4Jxe4 5 'ifd5
d3-d4.
complications.
regular Lopez, minus the 30 tons of 4 dxe5 he should consider offering
20
Double KP Openings
an endgame, because even fewer
Whites will trade queens after 1 e4
d62 d4 e5!? Emanuel Lasker used
this version in a world championship match against David ('I detest
endgames') Janowsky. Today's
players, like Michael Adams, are
more confident on the Black side of
1 e4 d6 2 d4liJf6 3liJc3 e5!?
gives up the center, such as in 1 e4
e5 2 liJf3 d6 3 d4 exd4 and then
4liJxd4 g6 or 4 ... liJf6 5liJc3 iJ.e7.
White can try to exploit this with
Henry Bird's 4 iJ.c4!?
A key point is that 4 ... liJc6
5 liJxd4 is a fine Scotch Game. And
5 c3 dxc3 6 liJxc3 transposes into a
good Scotch Gambit. It's as if Black
met 1 e4 e5 2 liJf3 liJc6 3 d4 exd4
4 c3 dxc3 5 liJxc3 with 5 ... d6
6 iLc4.
Now 4 liJf3 liJbd7 transposes to
Nimzovich's line and 4 dxe5 dxe5
5 'ii'xd8+ ~xd8 is a slightly worse
endgame, e.g. Vaisser-Rontaine,
French Championship 2006 went
6 iLc4 i.b4! 7 iJ.d2 ~e7 8 f3 c6.
The test of 4 iJ.c4 is 4...liJf6. In
the first ECO Lev Polugayevsky
gave 5 liJg5 d5 as favoring White
(6 exd5 h6 7 liJf3 iJ.b4+ 8 c3 dxc3
9 'ii'a4+ and wins).
Black was soon equal (9 liJge2
liJbd7 10 liJc 1 b5! 11 iJ.e2 iJ.d6
12 a4 b4 13 liJdl a5), then better
(14 liJe3 liJb6 15 liJd3 g6 16 b3
liJfd7 17 c3 bxc3 18 iLxc3 f6
19 liJb2 liJc5) and eventually won.
On the same page he considered
1 e4 e5 2 liJf3 d6 3 d4 liJf6 4 liJg5!?
exd4 5 iJ.c4 d5 and called it equal.
It's the same position.
Note that the 1...d6 orders have
Better, after 4 iLc4liJf6 5liJg5, is
5...iJ.e6. This dates back to Paul
the added benefit of cutting down
White's 1 e4 e5 options, such as
2 f4, 2 iLc4, 2 liJc3 and so on.
Rudolf von Bilguer, who thought
6 liJxe6 fxe6 7 iJ.xe6 'ike7 (8 iLc8!
'ikxe4+ 9 'ike2) was even. To
modem eyes this looks like a slight
There's a separate chapter in
Philidor theory in which Black
21
Double KP Openings
White edge. Therefore 4 iL.c4!? is
more than just a confuse-move.
brothers popularized 1 e4 eS 2 ttJf3
ttJf6 3 ttJc3 to reach a Four Knights
Game, for example.
Observe that in traditional (1 e4
eS 2 ttJf3 d6) orders, White can try
3 iL.c4 with the idea of 3... ttJd7
4 0-0 iL.e7 S d4. In this way he
avoids 3 d4 exd4 and traps Black in
the inferior Hanham line.
Their 19th Century colleagues
tried 1 e4 eS 2 ttJf3 ttJf6 3 d4 exd4
and 4 iLc4 in an effort to get into a
Two Knights (4 ... ttJc6), which is
terra incognita to a Petroff player.
But the critical line is 4 ... ttJxe4!, the
dubious Prince Yurosov Gambit.
Better is 3...iLe7! 4 0-0 ttJf6
which leaves White with a tiny edge
after S 1::te I 0-0 6 c3 ttJbd7 7 d4.
More ambitious is 3... ttJf6!? 4 ttJgS
dS S exdS, which reaches a Two
Knights Defense in which Black is
a tempo behind.
These days 3 d4 ttJxe4 is more
common, and theory says the
chances are roughly balanced after
4 iL.d3 dS S ttJxeS. Yaacov Murey's
discovery, 4 ... ttJc6, enables Black
to avoid this. It's based on S .txe4
dS 6 .td3 e4, which appears to be
sound.
If White wants to avoid Murey
and seek a main line he can do it
with 4 ttJxeS!?
But the missing tempo is ... ttJc6,
which may benefit Black since the
knight is not attacked here. He can
create an unclear position with
S... h6 6 ttJf3 iL.d6 or 6... e4 7 'i'e2
..tb4 8 c3?! O-O! (Miguel Najdorf).
Then 4... dS S .td3 transposes
into the main line as if 4 .td3 dS
had been played.
PETROFF DEFENSE
The Petroff is so annoyingly
solid that a frustrated White will be
sorely tempted to transpose into
another opening. The Paulsen
This order gives Black an extra
option, 4... d6!? S ttJf3 dS. He's
transposed into the 'other' Petroff,
22
Double KP Openings
the 3 ltJxeS line, which White
White reaches a Smyslov after
presumably wanted to avoid when 7... 0-0 8 c3 dS 9 i.c2 lite8 10 d4.
he chose 3 d4. This skirmishing - It's a more conservative form of it
Murey's 4 ... ltJc6, the preemptive because White has given up on the
4 ltJxeS and the counter-finesse useful c2-c4. But it's also a much
4... d6 - has nothing to do with more modest position for Black
finding the objectively best move than many Petroff players can
but rather seeking the position in tolerate.
which your opponent will be most
The best challenge to S i.d3 is
uncomfortable.
5... d5 and then 6 'iVe2 'ile7. A
The 'other' Petroff is the most typical case was Romero Holmespopular today and it runs 3 ltJxe5 Garcia Padron, Las Palmas 1991:
d6 4 ltJo ltJxe4 S d4 dS. The 7 0-0 ltJcs 8 lite 1 'ilVxe2 9 i.xe2
Smyslov alternative, S... i.e7 6 i.d3 ltJe6 10 d4 c6 11 c4! i.b4 12 ltJc3
ltJf6, has a stodgy reputation 0-0 13 a3! i.xc3 14 bxc3 b6
because 7 h3 leaves Black's QB IS ltJeS l:d8 16 a4! f6 17 ltJg4 and
without a good sqaure. Then 7... 0-0 White won.
8 0-0 dS transposes to a favorable
A later Petroff tabia begins with
Exchange French in which h2-h3 is
3 ltJxe5 d6 4 ltJo ltJxe4 5 d4 d5
an extra move.
6 i.d3 followed by some Black
White can try to trick his mixture of ... i.g4, ... i.e7 and
opponent into the Smyslov line via ... ltJc6. But the order is in dispute.
an order popularized by Alfonso In the definitive 19th Century
Romero Holmes, starting with authority, the Handbuch, Emil
5 .id3!? Then S... ltJcs 6 i.e2 gives Schallop recommended 6 ... i.e7
him 7 d4 with tempo. So 5.•.ltJf6 7 0-0 ltJc6 followed by ... i.g4. But
60-0 .ie7 7 h3! is more common. others argued for 6... ltJc6 7 0-0
i.e7, or even 6... i.g4 7 0-0 ltJc6
and ... i.e7 as Carl Schlecter
recommended in his version of the
Handbuch.
We can dispense with the last
order, since 6... i.g4 allows 7 'iVe2!
with an edge after 7 ... 'iVe7 8 0-0
ltJc6 9 i.bS! and i.xc6+, or 7... fS
8 h3 i.hS 9 g4!.
23
Double KP Openings
Garry Kasparov said 6.•. tbc6
7 0-0 i.e7 8 :tel i.g4 was fine for
Black.
chooses between undermining the
e4-knight with 8 c4 or attacking it
with 8 .:tel. Isaac Kashdan had
success in the 1930s with a line that
runs 8 c4 tbb4! 9 cxd5 tbxd3
10 'iVxd3 'iVxd5 11 .:tel i.f5.
The Kashdan line still holds up
well today, and if Black likes it he
can try to reach it even after 8 .:tel.
He does that with 8...i.f5, rather
than the popular 8... i.g4.
On 9 c4 Black has 9 ... tbf6
10 cxd5 i.xf3! with a good game.
But he added that White gets an
initiative if he undermines the
knight a move earlier with 8 c4!.
That's why Karpov preferred
6... tbc6 7 0-0 i.g4 in their 1985
match, so Black could meet 8 c4
Then 9 c4 tbb4 10 cxd5 tbxd3
with 8... tbf6!? Then 9 tbc3 is
considered best (9 ... tbb4 10 i.g5! transposes to what Black wants or 9... dxc4 10 i.xc4 i.e7 11 d5!) which White probably doesn't as he
and the results of 9 ... i.xf3 10 'iVxf3 could have sought it directly with
tbxd4 11 'iih3 have been in White's 8 c4.
favor lately.
PONZIANI OPENING
The bottom line is that no order
The Ponziani (1 e4 e5 2 tbf3 tbc6
is universally endorsed. Jan 3 c3) gets no respect today. But if
Timman and Alexander Belyavsky Black is caught off guard by 3 c3 he
like 6...tbc617... i.e7 while Vishy can bail out with 3...tbf6 4 d4 d6.
Anand, Vladimir Kramnik and
There doesn't seem to be
Alexey Shirov preferred 6... i.e7!
anything
in 5 i.c4 tbxe4!, so
7... tbc6 - and Yusupov plays both.
5 i.b5, transposing to the old
Another tabia arises after, say, Steinitz Defense of the Ruy Lopez,
6...tbc6 7 0-0 i.e7. White usually is played most often.
24
Double KP Openings
best order because 4 ... i.h4?! is
dubious. In contrast, in the
traditional Scotch order, 1 e4 e5
2 ttJf3 ttJc6 3 d4 exd4 4 ttJxd4
Black can avoid this line via
4 ... i.c5, among others.
There are several finesses in the
Four Knights line after 5... i.b4
6 ttJxc6 bxc6 7 i.d3.
The Steinitz is regarded as a
clumsy antique but this is a Steinitz
in which White has played
conservatively (1 e4 e5 2 ttJf3 ttJc6
3 i.b5 d6 4 c3?! ttJf6 5 d4 instead
of 4 d4!).
This order was a favorite
Ponziani evasion for GMs like
Smyslov and Yefim Geller and it
has no major drawbacks, e.g.
5... i.d7 6 0-0 i.e7 7 'iWe2 0-0 8 :dl
'i'e8 (threat of ... ttJxd4!) 9 d5 ttJd8
11 c4 c6 11 i.a4 b5! as in ManikMokry, Olomouc 1998.
Now 7... d5 8 exd5 cxd5 9 0-0
0-0 10 i.g5 is played with almost
robotic uniformity and is very
slightly in White's favor or even,
depending on whom you read. But
there are several diversions starting
with 8...'iWe7+!?
But if, on the other hand, Black
knows his way in a sharp line of the
Scotch Gambit (1 e4 e5 2 ttJf3 ttJc6
3 d4 exd4 4 c3 ttJf6!? 5 e5 ttJe4) he
can transpose into it from the
Ponziani by means of 3... ttJf64 d4
exd4!.
Theory has gone back and forth
about the value of 9 'iVe2. In any
case, against an opponent who likes
to keep queens on the board,
8... 'iWe7+ has obvious merits.
SCOTCH GAME
But if White is happy in an
ending, he can try to force it by
meeting 8... cxd5 with 9 ~e2+!? so
that 9... 'i'e7 transposes. There are
extra benefits when an endgamehating Black plays 9 ... i.e6
White can get into a major
Scotch variation via the Four
Knights (1 e4 e5 2 ttJf3 ttJc6 3 ttJc3
ttJf6 4 d4 exd4 5 ttJxd4). If that's a
position he wants to play, this is the
25
Double KP Openings
(10 i.b5+! tbd7 11 i.c6 d4 12
'ife4) or 9 ... i.e7 (10 i.g5 0-0
11 O-O-O!? with better-than-usual
chances).
These days Black is more
concerned about the standard
Scotch order, 1 e4 eS 2 tbo tbc6
3 d4 exd4 4 tbxd4 and then 4 ...tbf6
5 tbxc6 or 4 ... i.c5 5 tbxc6. For that
reason, he can try to reach the Four
Knights version with 4..•i.b4+!?
Black can anticipate this and rule
out endgames through another route
to the tabia devised by Georg
Marco, 7..•0-0 and then 8 0-0 dS
9 exdS cxdS 10 i.gS. This appears
to be the most exact - and the most
deceptive since White will begin
thinking about 8 i.g5 and the threat
of 9 e5. Black is worse after 8... d6,
so 8... dS:
Now 5 tbc3 tbf6 transposes to it,
and considering the current view of
theory, that would be a small
victory for Black. The obvious
objection to 4 ... i.h4+ is S c3. Then
S...i.cs 6 i.e3 i.b6 reaches a book
position - from 4 ... i.c5 5 i.e3 - but
with an extra c2-c3 for White.
Black's point is 9 eS 'ife8! and
White is the one surprised, e.g.
10 f4? tbg4 11 'ife2 f6 (12 exf6
gxf6 13 h3 £Xg5 14 bxg4 i.xg4!
15 iLxh7+ <Jiig7 and he resigned
in Candela-Korneev, Ponferrada
1997).
Who benefits? White gets added
protection of d4. But he loses one of
the major assets of the Scotch, the
knight hop from c3 to d5. Recent
practice indicates Black is OK after
7 tbfS i.xe3 8 tbxe3 tbf6 because
White cannot defend the e-pawn
with tbc3 (and 9 tbd2 allows the
equalizing 9... d5). Or 7 i.c4 d6
8 0-0 tbf6 9 tbd2 tbe5! 10 i.e2 0-0
11 h3 :e8 as in GoloshchapovBrodsky, Hoogoveen 2006.
If White spots that and heads
instead towards the tabia, with
9 exdS, Black can use an extra
option, 9...'ifxdS!. This hits at g5
and g2 and after 10 i.xf6 "iVe6+ and
11..:iVxf6 he stands well.
26
Double KP Openings
And what about the standard
order, 4 llJxd4 .tcS S .te3,
threatening llJxc6 ? Much midnight
oil has been spent on 5.. :iVf6. The
mental hygiene alternative is
S.••.tb6!.
Now 6 llJc3 d6 and 7... llJf6 or
6... llJf6 transpose to a solid Four
Knights/Scotch line. For example,
Sveshnikov-Mi. Tseitlin, Budapest
1989 went 6 llJc3 d6 7 g3 llJf6
8 .tg2 0-0 9 llJxc6 bxc6 10 .txb6
axb6 with equality.
Another trendy line is 4 ... llJf6
5llJxc6 dxc6 5 e5. Black can avoid
that with 4••:i'f6. Then he is trying
to reach, not avoid, the 5 .te3 .tc5
position. This order has the added
benefit of dodging another difficult
line, 4 ....tc5 5 llJb3.
The best try at punishing 4 ... 'i'f6
is 5 llJb5 .tc5 6 'i'e2. But the
verdict is out on 6....tb6 or 6... 'i'd8
(and ... a6).
the queen move is even playable. Its
fate depends on how we evaluate
this tabia.
The final word on 7 ... ..txc3+
8 llJxc3 or 8 bxc3 is still to be
written. But equally important is
finding White's best way to the
diagram. There are four routes.
(a) S llJc3 ..tb4 6 llJdbS 'iYxe4+
7 ..te2 gets there. But Black does
much better with 6 .....ta5! and ... a6,
e.g. 7 .te2 a6 8 llJd4 ..txc3+ 9 bxc3
llJf6 10 llJxc6 dxc6 11 e5 llJg4
(Renteria-Miktov,
Minneapolis
2005). For that reason White should
prefer:
(b) SllJc3 ..tb4 6 ..te2! and then
6•.:i'xe4 7 llJdbS. If Black varies
with 6 ... llJf6 7 0-0 .txc3 then
8 llJf5! is potent, e.g. 8... 'iYxe4
9 .td3 'iVg4 10 f3! ~a4 11 bxc3
0-0 12llJxg7! with a winning attack
in Karjakin-Malinin, Sudak 2002.
But White may opt for 5 llJb5
because it is more forcing and seeks
something better than the diagram.
For example:
At one time the fate of the Scotch
seemed to hinge on the soundness
of 4•..'i'h4. Today there's doubt that
27
Double KP Openings
(c) S tbbS 'fixe4+ 6 .te2 i.b4+
7 tblc3 gets to the diagram. But
White can improve with 7 .td2!,
with elaborate lines following
7... .txd2+ 8 'iVxd2 'fie5 9 f4!, for
instance.
•.• tbd8-e6/ ..•cS sets up a darksquare blockade. This is Black's
best way to play for a win
in the Four Knights, since Akiba
Rubinstein's 4 ... tbd4 allows a
draw-minded White 5 tbxd4 exd4
6 e5.
(d) Black can try to improve with
S....tb4+ since 6 tblc3 'iVxe4+
7 .te2 transposes. But again the
bishop interpolation, 6 .td2!, is
better and reaches ( c) after
6... 'iVxe4+ 7 i.e2.
There are ways for White to
avoid the Metger - and still play to
win - and other ways for Black to
seek it. White's sidestep is 7 tbe2,
which has been revived by Daniel
Campora, who showed that White's
edge in a !1ymmetrical position
(7 ... tbe7) is something he can work
with.
The main problem with 5 tbb5 is
S.•..tcS and then 6 'iVe2 tbd4, when
White doesn't reach the diagram or
prove a clear edge. As it stands now,
(b) may be the most accurate.
For example, 8 c3 .ta5 9 tbg3
tbg6 10 d4 c6 11 .td3 .l:te8 12 h3
.tb6 13 .l:tel h6 14 .tc2 'fic7
15 i.e3 i.e6 16 'fic 1! ':ad8 17 a4
a5 18 tbf5 d5 19 tbxg7!, Campora-
FOUR KNIGHTS GAME
The main Four Knights line for
most of the last century was
Johannes Metger's 'unpin', after
1 e4 eS 2 tbf3 tbc6 3 tbc3 tbf6
4 .tbS .tb4 S 0-0 0-0 6 d3:
Baron Rodriguez, Lanzarote 2003.
Black can avoid this by capturing
on c3 earlier. Reuben Fine did it
with 6..•'fie7 7 i.gS .txc3 8 bxc3.
To avoid the drawish 8... d6 9 i.xc6
he played 8...tbd8!:
Black breaks the pin with 6... d6
7 .tgS i.xc3 8 bxc3 'iVe7 and
28
Double KP Openings
Now 9 d4 d6 is the Metger
position. If White starts the knight
maneuver with 7 lDe2, Black has
7... d5!?, e.g. 8 exd5 lDxd5 9 lDg3
lDf4 10 d4 lDg6 11 c3 .td6,
Levenfish-Alatortsev,
Moscow
1940.
Nevertheless the more popular
method is to delay .. .'ike7, as in
6....txc3 7 bxc3 d6 and 8 .tg5
Now 5... exd4 6 e5lands Black in
'iie7). This was a favorite order of the Lange he presumably wanted to
avoid. If he still feels that way,
Frank Marshall and Smyslov.
5....txd4! and 6 lDxd4lDxd4 7 ..tg5
d6 8 f4 'iie7 or 8... .te6 is the
GIUOCO PIANO!
antidote.
But Yakov Estrin argued
TWO KNIGHTS
that White had reasonable chances.
There's a recognized boundary What's certain is that this is much
after 1 e4 e5 2 lDD lDc6 3 .tc4 that better than 4 d4 .txd4 5 lDxd4
separates the Giuoco, 3....tc5, from lDxd4 6 f4? (6 0-0 transposes) d5!
the Two Knights, 3...lDf6. But the 7 exd5 'iih4+ 8 g3 'iih3, which
border is often blurred by trans- favors Black.
position.
Because 3... .tc5 makes no threat
White has more freedom than in the
Two Knights. He can head towards
the Pseudo-Lopez with 4 d3, 4 c3 or
4 0-0 while retaining other options.
With 3... .tc5 Black is avoiding
certain lines, such as 3 ... lDf6
4 lDg5, which have good
reputations, or even the questionable Max Lange, 3...lDf6 4 d4 exd4
Theory says 4 c3 can be met by
50-0 .tc5 6 e5!?
4 ... lDf6 or 4...'iie7 5 d4 .tb6. But if
holding e5 as a strong point is
White can try to push Black into Black's intent, it's more precise to
the Lange via 1 e4 e5 2 lDo lDc6 play 4....tb6. Then 5 d4 'iie7, as the
3 .tc4 i.c5 and now 4 0-0. Castling Berlin Pleiades played, transposes.
was once ridiculed but today it's
often seen as an alternate route to
the Pseudo-Lopez (4 0-0 lDf6 5 d3).
There is also 5 d4!?
The rationale for 4....tb6 is that
.. .'ike7 is not as useful as ... ..tb6 in
the Pseudo-Lopez. White can
change his mind after 4...'iie7 and
29
Double KP Openings
switch to S d3 with idea of hitting
the vulnerable queen later on with
ttJe3-dS or ttJfS. The minus to this
order is that after 4 ... ..tb6 S d3
Black will have lost a tempo if he
feels ... a6 and ... ..ta7 are necessary
later on.
Another move order issue in the
Two Knights concerns the periodically revived 4 d4 exd4 S eS dS
6 ..tbS ttJe4 7 ttJxd4.
Note that 3 .te4 ..teS is a route to
the Canal Variation (4 ttJe3 ttJf6
S d3 d6 6 ..tgS!?), one of the most
promising Giuoco lines. White
can't reach it via the Four Knights
because of 1 e4 eS 2 ttJf3 ttJc6
3 ttJc3 ttJf6 4 .tc4? ttJxe4! and he
needs Black's cooperation in the
Vienna order. However, he can get
to it in the Two Knights Defense,
1 e4 eS 2 ttJf3 ttJe6 3 .te4 ttJf6 ifhe
avoids 4 ttJc3? ttJxe4! and plays
4 d3! and then 4.....teS S ttJe3.
It was once assumed Black must
reply 7.....td7, Paul Morphy's
move, after which 8 ..txe6 bxe6
9 0-0 ..teS and 10 ..te3 is fairly
balanced. But recent experience
suggests 10 f3! gives White the
upper hand.
In a standard Two Knights order,
1 e4 eS 2 ttJf3 ttJe6 3 .te4 ttJf6
4 ttJgS, there are two offbeat but
related lines, 4... dS S exdS ttJd4 and
S... bS. They don't look at all alike.
Yet they transpose after S... ttJd4
6 c3 bS! 7 .tfl! and S... bS 6 .tfl!
ttJd4 7 c3.
To avoid that, Black can employ
Adolf Anderssen's 7.....teS!? Then
8 .te3 .td7 9 ..txc6 bxc6 10 f3
would be met by 10 .. .'ifh4+ 11 g3
ttJxg3 !. Instead, White usually
replies 10 0-0.
The point is he transposed into
the equal 7... ..td7 line in which
10 ..te3 is played and lost the
promising 10 f3 option.
Black's choice should depend on
which line is likely to provoke a bad
response by White, such as S... bS
6 dxc6?! bxc4 7 ~e2 h6 8 'ilxeS+
JLe7 which favors Black. Hans
Berliner revived S... bS in his world
postal championship run, explaining that S... ttJd4 'is less clear' after
6 ttJc3.
The risk of this finesse is
allowing 7 ... .tcS 8 ttJxc6!? and
then 8 ... ..txf2+ 9 ~fl or
9 ~e2!? ..tg4+ 10 ~fl. White also
gets an extra option after 8 ..te3
JLd7 9 ..txe6 bxe6 when he tries
30
Double KP Openings
10 liJd2!, which has been doing
well lately, rather than transpose
with 10 0-0.
The prime benefit is that 6 'tib3,
which is a valid move against
5... i.a5, is poor here since 6 .. :i!le7
threatens 7... liJa5 !.
But that benefit is trumped by
6 d4!. Black can't reach the safety
of the Lasker Defense, 6... i.b6
(7 0-07 d6), in view of 7 dxe5!.
For example, Sveshnikov-Zaja,
Bled 2001 went 10 ... liJxd2
11 'i'xd2 0-0 12 liJb3! i.b6
13 0-0-0 'i'e7 14 l::the1 a5 15 a3 a4
16 ..tg5! 'i'e6 17 liJd4 'i'g6 18 f4
and White eventually won.
Therefore he has to enter one of
the main Evans lines, such as
6... exd4 7 O-O! d6 8 cxd4 i.b6,
where White's claim to compensation has withstood the test of time a century and a half and counting.
On balance, 7 ... i.d7 appears
better than 7... i.c5. And Morphy
beats Anderssen once more.
Once Black decides on S...i.aS a
key question is whether White
EVANS GAMBIT
should play the immediate 6 'ib3 or
The first question to be asked in insert 6 d4 first. More attention is
the Evans (1 e4 eS 2 liJf3 liJc6 paid to 6 d4 d6 7 'iVb3 but 7.. :i!ld7
3 i.c4 ..tcS 4 b4 i.xb4 S c3) is seems to be adequate.
where Black should retreat the
Compare that with 6 'ib3:
bishop. There are two ways for it to
reach b6 and set up the solid Lasker
Now ... 'i'd7 is impossible. The
Defense - 5... i.c5 6 0-0 d6 7 d4 other queen moves, 6 .. :iVf6 or
..tb6 and 5... i.a5 6 0-0 d6 7 d4 6.. :iVe7, seem to transpose into bad
..tb6. The order matters because versions of 6 d4 d6 7 'iVb3. For
White has other options at move example, 6 'iVb3 'iff6 7 d4 d6?
walks into 8 d5! liJ-moves 9 Vi'a4+.
six. Consider S•.. i.cS:
31
Double KP Openings
Also on 6 'i!Vb3 'fie7 7 d4 exd4
8 0-0 .tb6 9 cxd4 or 6 ... 'fif6 7 d4
exd4 8 0-0 d6 9 e5! we're in
variations that often prove fatal to
Black before move 25.
He wants to save c2-c3 for later,
such as with 5... d6? 6 d4 exd4
7 c3!?, as suggested by Savielly
Tartakower. But Marshall showed
that 5 0-0 can be safely answered by
5...tDf6! 6 d4 exd4 7 c3 dxc3 8 e5
d5!.
The real challenge to 6 'ifb3 is
6...'fif6 7 d4 tDxd4!. Then 8 tDxd4
exd4 9 0-0 has been tested a bit in
correspondence (9 ... dxc3 10 e5
'fixe5! 11 .txf7+ ~f8) without a
conclusion. If White likes the looks
of this, 6 'tWb3 would be much better
for him than 6 0-0 and perhaps
superior to 6 d4, too.
SCOTCWGORING GAMBIT
These two gambits are closely
related and may reach Giuoco and
Two Knights positions. The Scotch
order (1 e4 e5 2 tDo tDc6 3 d4 exd4
4 .tc4) gives Black more choice.
Note that early in the Evans, each
side can give his opponent a
dubious 'something-to-think-about'
move. Black's is 4 ... tDxb4?!, rather
than 4 ... .txb4. But there's not that
much for White to consider and
after 5 c3 tDc6 all that's happened is
Black transposed to the questionable 4 ....txb4 5 c3 .tc5.
White's bid for cleverness
worse. It is 5 0-0 after 4 ... .txb4.
The simplest reply is 4...tDf6
which transposes into the Two
1S
32
Double KP Openings
Knights (1 e4 e5 2 tDf3 tDc6 3 ~c4
tDf6 4 d4 exd4). This is a good
choice for a Black who likes to face
the Max Lange (5 0-0 ~c5) or the
anti-Lange (5 0-0 tDxe4).
But if Black normally meets
3 ~c4 with 3... ~c5 this is a bad
order because he may land in
unfamiliar territory, That's why he
may prefer 4... ~b4+. His idea is
that 5 c3 dxc3 6 tDxc3 transposes
into what was long considered his
best Goring defense (1 e4 e5 2 tDf3
tDc6 3 d4 exd4 4 c3 dxc3 5 tDxc3
~b4! 6 ~c4).
His idea is not 6 c3 dxc3?! because that transposes into an
unattractive Goring.
Better is 6... ~g4 with unclear
play. For example, 7 'ii'b3 'ii'd7
8 cxd4 ~xf3 9 dxc5 tDd4 10 'ike3
'ii'g4 11 'iVg5 with a complex
endgame in Wist-Rausis, correspondence 1994.
That conclusion is in recent
dispute. However White doesn't
have to get involved in the debate
since in the Scotch order 6 bxc3!
favors him (6 ... ~e7? 7 'ii'd5).
One final point is that 4... ~c5
can also transpose, after 5 c3 tDf6 to
a Giuoco Piano. That's a safer
choice than 5... dxc3?! 6 tDxc3 d6,
another unclear Goring line (7 ~g5
tDge7 8 tDd5 f6 9 ~xf6!?).
Well, then, what about the third
option in the last diagram, 4...~c5 ?
That is a not-so-veiled invitation to
the Lange (5 0-0 tDf6 6 e5 d5).
Compared with 4... tDf6, Black loses
the anti-Lange possibility. But he
benefits by not having to face
the 5 e5!? line, which Yevgeny
Sveshnikov has proven to be
dangerous.
So the bottom line of the Scotch
Gambit, 4 ~c4, is that 4... ~b4+?!
is to be avoided, 4 ... tDf6 is the safe
choice and 4 ... ~c5 is the most
double-edged. And that's one of the
more definitive conclusions we can
make about those trap-happy 1 e4
e5 openings.
Another virtue is that after
4.•.~c5 5 0-0 Black has an extra
5... d6 option.
33
Chapter Two:
RuyLopez
There are far more tricks in the
move order of the Ruy Lopez than
in any other I e4 e5 opening and
they often start as early as move
four, as in the Berlin Defense. The
bona fide Berliner is content to head
into the endgame of 1 e4 eS 2 ttJf3
ttJc6 3 .i.bS ttJf6 4 0-0 ttJxe4 5 d4
ttJd6 6 ..txc6 dxc6 7 dxe5 ttJf5
8 'iixd8+. He might be upset by
4 'iWe2!?
Black should know the counterfinesse, 4 ... a6!. Then 5 .i.xc6 dxc6
6 ttJxe5 'iWd4 regains the pawn.
White's best is 5 .i.a4!, but he
has to know the WorralllWormald
Variation, (usually reached by 3 ... a6
4 .i.a4 ttJf6 5 'iWe2), into which he's
transposed.
CLASSICAL DEFENSE
In the Classical Defense (1 e4 eS
2 ttJf3 ttJc6 3 .i.bS .i.cS) there are
two early tabias and various ways
of reaching them. One tabia, with
... ttJge7, occurs after 4 c3 ttJge7
S 0-0 .i.b6 or 4 0-0 ttJge7 S c3 .i.b6.
He is unlikely to be happy
defending the Steinitz-Defense-like
4....i.e75 c3 d6 6 d4. Nor will he be
eager to gambit a pawn, 4 .....tcS!?
5 .i.xc6, e.g. 5... dxc6 6 ttJxe5 'iVd4
7 ttJd3!. That's probably not in his
DNA.
There's a consensus that White is
at least slightly better after 6 d4 and
34
RuyLopez
11 exd6 cxd6 12 l:.d1 as in
Matulovic-Ilincic, Sombor 2004.
his main concern is what he wants
to avoid: 4 0-0 will sidestep 4 c3
f5!?, for example. But 4 c3 will
avoid 4 0-0 ttJd4.
There's psychology behind 4 0-0
and then 4.•.ttJf6 5 ttJc3!?
What would really turn matters
in favor of 4 c3 is if someone found
a way to get an advantage from
White's extra option, 4... ttJge7 5 d4.
But 5... exd4 6 cxd4 ..tb4+! has
been equalizing for decades.
If Black intends ... ttJf6 instead of
... ttJge 7, there are three orders
worth considering. The first and
most effective is 3.•.ttJf6! and then
4 0-0 ..tc5. This exploits White's
preference for 4 0-0 when he
expects Black to continue with a
normal Berlin. In this way Black
avoids an early c2-c3/d2-d4 in view
of3 ... ttJf6 4 c3? ttJxe4!.
What's happened is we've landed
in a Four Knights Game (1 e4 e5
2 ttJf3 ttJc6 3 ttJc3 ttJf6 4 ..tb5) in
which Black has played a rare
sideline, 4 ... ..tc5!? and White
replied 5 0-0.
The other two orders start with
3••...tc5 - 4 c3 ttJf6 5 d4 ..tb6 6 0-0
0-0 and 4 0-0 ttJf6 5 c3 0-0 6 d4
..tb6. Books give much more
attention to the first order because it
grants White an extra option,
6 'iWe2. But if he intends 6 0-0
anyway, and he doesn't remember
the complex 4 c3 f5 theory, the
second order is best.
That's fine for Black - if he
knows the theory (5 ..• 0-0! 6 ttJxe5
ttJxe5 7 d4 ..td6!).
Alexander Be1yavsky was once
surprised by Ljubomir Ljubojevic
in this move order and was
outplayed after 6.. J~e8?! 7 ttJf3
ttJxe4 8 d4 ttJxc3 9 bxc3 ..tf8
10 ttJg5! h6 11 ttJh3 d6 12 'iif3!
..td7 13 l:.b1 l:.b8 14 'iWg3 ttJe7
It has the added benefit of 15 ..txd7 'iixd7 16 ..txh6!.
tempting Black into 4 ... ttJf6 5 c3
ttJxe4? That looks like the 3... ttJf6
FIANCHETTO/COZIO
4 c3? ttJxe4! mentioned above. But
DEFENSES
this one turns out to be bad after
When Black develops his KB on
6 'iWe2!, e.g. 6... f5 7 d3 ttJf6 8 d4
..te7 9 dxe5 ttJe4 10 ..tc4! d5 g7 he is playing a reversed Vienna,
35
RuyLopez
a mirror image of 1 e4 e5 2 ttJc3 version of the Giuoco Piano in
ttJf6 3 g3 Jtb4. But White's extra which Black has an extra ... a6, e.g.
tempo in the Lopez version (1 e4 e5 5... d6 6 d4 'ilie7 7 0-0 Jtg7.
2 ttJf3 ttJc6 3 Jtb5 g6) means he
There's a back-door route to a
can push his d-pawn faster.
Black fianchetto and it avoids the
He should do that quickly 3... g6 4 c3 issue as well as the more
because after 4 0-0 Black has time troublesome 3 ... g6 4 d4 exd4
for 4... Jtg7 5 c3 ttJge7 6 d4 and 5 Jtg5!. This is the Cozio Defense
then 6... exd4 7 cxd4 d5! with a (3 ...ttJge7), which transposes to a
double-edged game.
3 ... g6 4 0-0 position after 3... ttJge7
4
0-0 g6 5 c3 Jtg7 6 d4 exd4 7 cxd4
More dangerous is 4 c3:
d5.
If White replies 4 d4 exd4
5 ttJxd4 we get 5... g6 6 ttJc3 !i..g7
7 Jte3 0-0:
Black is in trouble after 4 •.•d6
5 d4 Jtd7 6 'ifb3! (6 ... Jtg7 7 dxe5
dxe5 8 Jtc4 hits f7 and b7). Vassily
Smyslov, the maestro of this
variation, said 4... a6! is best. Black
is pleased to transpose, after 5 i.a4,
to a Modem Steinitz Defense (3 ... a6
4 i.a4 d6 and ... g6).
This becomes an old Steinitz
Defense, in fact one of the better
versions of the old Steinitz, after
8... d6. But Black may do better with
Why would Black prefer 3 ... g6 to ... d5!? For example, 8 'ilid2 d5
the normal Modem Steinitz order? 9 exd5 ttJxd5 10 tLlxc6 bxc6
The answer is he avoids other 11 Jtxc6 ttJxe3! 12 'ilixe3 l:.b8 or
Modem Steinitz lines, such as 12 !i..xa8 'iixd2+ l3 Wxd2 ttJc4+
5 O-O! and 5 i.xc6+!?
14 WeI ttJxb2.
White might prefer to meet 4 ... a6
with 5 .tc4. That's a pretty solid
There must be a drawback to the
Cozio order or everybody would be
36
RuyLopez
playing it. It is 3...tbge7 4 tbc3!
which gets very complex after
4..•g6 S d4 exd4 6 tbdS!. Whether
White has an advantage after
6... i.g7 7 i.gS is in dispute. Yet
few Lopez players are willing to try
it and that makes the Cozio order a
good practical weapon.
After S 0-0 the most popular
move in recent years is S...i.cS.
BIRD'S DEFENSE
Henry Bird's defense (1 e4 eS
2 tbO tbc6 3 i.bS tbd4 4 tbxd4
Now 6 d3 c6 7 i.c4 dS! heads
exd4) looks absurd: Black moves
his only developed piece a second towards the good S d3 line. More
time in order to double his own accurate is 6 i.c4!, breaking the
pawns.
rule against moving a piece twice in
But those are doubled center the opening. Black must deal with
pawns and they do a good job of the threat of 7 i.xf7+ ~xf7 8 'iihS+
restricting White after S d3 c6 and doesn't have time for ... dS.
6 i.c4 dS!. For example, 7 exdS
cxdS 8 i.bS+ i.d7 9 i.xd7+ 'ii'xd7
100-0 i.cS! 11 l:.e1+ tbe7 is quite
nice for Black.
STEINITZIANA
Wilhelm Steinitz's solution to the
Lopez
problem was maintaining a
White's error was committing
himself too early (S d3?!). The best solid bridgehead on eS, after 1 e4 eS
waiting move in many if not most 2 tbo tbc6 3 i.bS, by means of
openings is castling. Here S O-O! 3 ... d6. His disciples tried 3... d6
allows him to decide between 6 d3, 4 d4 i.d7 S tbc3 tbf6 6 0-0 i.e7 or
6 c3, 6 i.c4 or even 6 'ii'hs after he its transpositional kin such as
3••• tbf6 4 0-0 i.e7 S tbc3 d6 6 d4
sees Black's fifth move.
i.d7.
He has excellent chances
following S... c6 6 i.c4 dS 7 exdS
A key position arises after White
cxdS 8 i.bS+ i.d7 9 l:.e1+ tbe7 protects his e-pawn with 7
and
10 i.xd7+ 'ii'xd7 and now l1"iihS threatens to win Black's. That leads
or 11 c4 (rather than transposing to a tabia, 7... exd4 8 tbxd4 0-0
with 11 d3?).
9 i.xc6 bxc6.
ne1
37
RuyLopez
White is better but Black has
play along the b-file and from ... c5!/
... iJ.c6, e.g. 10 iJ.f4 c5 11 ltJD iJ.c6
12 'it'd3 .!tJh5! 13 iJ.e3 l:te8 14 l:Iadl
'ifb8 15 iJ.c1 'iib7, Narciso DublanMizoev, Salou 2006.
And now 5... exd4! 6 ltJxd4 ltJf6
7 0-0 iJ.e7. Then 8 iJ.xc6 or 8 l:Ie 1
0-0 9 iJ.xc6 heads towards the
double-edged ... bxc6 tabia that
should make Black happy.
Unfortunately for him there's a
counter-counter-finesse. White can
move up the capture on c6, as
Jackson Showalter showed more
than a century ago, 3... d6 4 d4 .td7
and then 5 iJ.xc6 iJ.xc6 6 .!tJc3. This
transposes to the favorable line after
6 ...ltJf6 7 'it'd3 exd4 8.!tJxd4.
Experience shows that White
should try to force Black to retake
on c6 with the bishop, thereby
eliminating ... c5! counterplay. He
does that with 3... d6 4 d4 iJ.d7
5 .!tJc3 .!tJf6 6 iJ.xc6! since 6... bxc6?
7 dxe5 drops a pawn.
Of course, Black can accelerate
his plan, too, with 4 d4 exd4, as
Alexander Onischuk does. But
there's a difference because White
can retake with his queen, 5 'it'xd4
iJ.d7 6 iJ.xc6, which has been
known to favor White since
Morphy's day.
Instead 6... iJ.xc6 7 'it'd3 renews
the dxe5 threat and favors White
following 7 ... .!tJd7 8 iJ.e3 exd4
9 iJ.xd4! or 7... exd4 8 .!tJxd4 iJ.d7
9 iJ.g5 iJ.e7 10 0-0-0 and 8... g6
9 iJ.g5 iJ.g7 10 .!tJxc6 bxc6
11 e5 dxe5 12 'it'D 'it'd6 13 l:tdl
(Lupulescu-Pessi, Bucharest 2004).
To resuscitate the Steinitz Black
needs a new idea, perhaps the littleexplored 5....!tJge7!? in the last
order, or something that confuses
White about the timing of iJ.xc6. He
might try the Berlin move order 3... .!tJf6 4 0-0 and then 4 ... d6 or
Black's counter-finesse lies in
moving up ... exd4 before iJ.xc6 so
that he can retake with the b-pawn.
He accomplishes this after 3... d6
4 d4 iJ.d7 5 .!tJc3:
38
RuyLopez
4 ... i.e7 5 :tel d6. This has the
benefit of denying White his most
aggressive plans such as 'iVd3/0-0-0
and avoiding 4 d4.
But White should be aware after
3•••lbf6 4 0-0 d6 5 d4 i.d7 that he
has 6 i.xe6! i.xe6 7 lbe3.
The liquidation that follows
7••. lbxe4 8 lbxe4 i.xe4 9 :tel
favors him (9 ... f5 10 lbg5 or
9 ... i.f5 10 dxe5 d5 11 lbd4).
Instead, Black usually transposes
with 7... exd4 8 lbxd4 into a ... i.xc6
line, rather than ... bxc6!.
most obvious improvement is that
by inserting 3... a6 4 i.a4 before
4 ... d6, Black makes d2-d4 a trifle
risky. After 1 e4 e5 2 lbf3 lbe6
3 i.b5 a6 4 i.a4 d6 5 d4 Black can
reply 5...b5! 6 i.b3 lbxd4 7 lbxd4
exd4
Now 8 'ifxd4? is the Noah's Ark
Trap (8 ... c5 9 'i'd5 i.e6 10 'iVc6+
i.d7 11 'i'd5 c4 and wins). But
8 e3! and then 8... dxc3 9lbxc3lbf6
10 0-0 is a worthwhile gambit.
White's most flexible fifth move
is 5 O-O!, which gives him all sorts
Bottom line: The Steinitz of finesses. He can, for example,
deserves its reputation but only if meet 5... i.d7 with 6 c3. This
White times i.xc6 correctly. If usually transposes into the old
Black likes the ... bxc61ines he has a Modern Steinitz main line, 5 c3
much better chance of reaching i.d7 but without risking the Siesta
them via a delayed Steinitz, such as Variation (5 c3 f5!?).
3... a6 4 i.a4lbf6 5 0-0 i.e7 6 :tel
d6 when White's best is 7 i.xc6+
Another virtue concerns Oldrich
bxc6 8 d4 exd4 9lbxd4 i.d7.
Duras's idea, c2-c4, It's not
considered dangerous because 5 e4
MODERN STEINITZ
i.g4! exploits the d4-hole. But after
DEFENSE
5 0-0 i.d7 White can seek a
This is called the 'Improved superior Duras with 6 e4. Then
Steinitz Defense' in Russia and the 6... i.g4 would cost Black a tempo
39
RuyLopez
and it's not clear if that or 6... tbf6
7 tbc3 tbd4 equalize.
7 .txc6 and get a favorable old
Steinitz position. '
The most common continuation
after 5 0-0 .td7 is 6 d4.
He means 7....txc6 8 %:tel exd4
9 tbxd4 because 8... tbxe4? 9 d5 and
8... .txe4 9 tbc3 favor White. As
often happens with transpositions,
both players may be right. White
got the middlegame he wanted and
Black avoided the middlegame he
sought to avoid.
Lajos Portisch added another
finesse to the Modem Steinitz when
he met 5 0-0 with 5....te7!?
This is an improved gambit in
view of 6•.. b5 7 .tb3 tbxd4
8 tbxd4 exd4 9 c3 because ... .td7 is
not as useful as ... .tb7.
And 6 ... tbge7 7 c4! is an
excellent Duras, e.g. 7...tbg6 8 tbc3
.te7 9 .te3 0-0 10 %:tel .tg4
11 .txc6 bxc6 12 h3 as in SmaginBabu, London 1990.
This looks like an innocuous
route to quiet lines, such as 6 c3
tbf6 7 !:tel 0-08 h3 .td7. But there
are differences between 5... .te7 and
5....td7.
Theory prefers 6...tbf6!. Then
both players can claim they won the
battle of the opening. Black says,
'J tricked you into a Steinitz
Defense Doubly Deferred without
allowing you a chance to use your
best weapon.'
One is that Black is better
situated to play 6... .tg4 compared
with the somewhat dubious 5 0-0
.tg4.
Translation: This position can be
reached via 3... a6 4 .ta4 tbf6 5 0-0
d6 6 d4 .td7 but only if White
passes up 6 .txc6+!.
For example, 6 c3 .tg4 7 d4 b5
8 .tb3 tbf6 intending ....txf3. Also
'No,' replies White. 'J got what J 7 d3 tbf6 8 tbbd2 tbd7 9 h3 .th5
wanted because J can now play 10 .tc2 tbf8!? was good for Black
40
RuyLopez
(11 g4 .tg6 12 d4 h5) in BrownePortisch, Milan 1975.
followed by ...ti:Je7. The forgotten
5....te7 deserves to be remembered.
Another difference is Black can
try to upgrade the Siesta Variation
by meeting 6 c3 with 6... f5!? 7 exf5
.txf5 8 d4 e4.
Black can also mix the Modern
Steinitz with the Cozio if he plays
1 e4 e5 2lbf3lbc6 3 .tb5 a6 4 .ta4
and now 4...lbge7.
More of a test of 5... .te7 is 6 d4.
Then 6...b5 7 .tb3 lbxd4 8 lbxd4
exd4 9 c3 is another center gambit.
After 9 ... lbf6! 10 cxd4 i.b7 11 f3 0o we've transposed into a roughly
even line (5 d4 b5 6 i.b3 lbxd4
7lbxd4 exd4 8 c3 .tb7 9 cxd4lbf6
10 f3 i.e7).
Alexey Suetin recommended
'5 c3!' and cited one of his games
that led to a 5...b5 6 .tb3 d5 7 'ife2!
edge.
If there is a refutation of 5... .te7
it is 6 .txc6+ bxc6 7 d4. Then
7 ... f6?, which is thematic in
comparable .txc6+ lines of the old
Steinitz, makes the kingside pieces
stupid. And 7... exd4 8 lbxd4
But Black improves with 5... d6!.
Then 6 d4 .td7 transposes to the
Modem Steinitz line that runs 5 c3
.td7 6 d4 lbge7. Theory likes
White's position more today than it
did a generation ago but what may
be more important is Black avoided
the non-5 c3 options of the Modem
Steinitz.
Moreover after 4 ... lbge7 5 0-0
Black can safely transpose into the
fianchetto line (5 ... g6 6 c3 d6).
White may have to challenge
But 5... .te7 is playable if Black 4 ...lbge7 with 5 d4 exd4 6 lbxd4,
delays ... c5 by means of 8....td7 which so far has remained unclear
9lbc3 lbf6. Or, better yet, 9....tf6! after 6... g6 or 6... b5!? 7 .tb3 lba5.
... allows White to set up a bind
after 8... c5 9 lbc6! 'ifd7 10 lba5!
(10 ... lbf6?! 11 c4).
41
RuyLopez
EXCHANGE VARIATIONS
White can play Jlxc6 at move
four, five or six in the main Lopez
lines. Exchanging at move five
makes sense ifhe wants to avoid the
Open Defense and retain the 0-0-0
option.
Eduard Gufeld beat Mikhail Tal
in a game that went 1 e4 e5 2 ttJf3
ttJc6 3 Jlb5 a6 4 Jla4 ttJf6 5 Jlxc6
dxc6 6 d3 ttJd7?! 7 ttJbd2 Jle7
8 ttJc4 Jlf6? and then 9 'iVe2 c5
10 Jld2 0-0 11 g4 with a winning
attack (ll...bS 12 ttJe3 g6?! 13 h4
ttJb8 14 0-0-0 ttJc6 IS :dgl Jle6?
16 ~bl ttJb4 17 Jlxb4 cxb4 18 gS!
and ttJg4).
This can transpose to a Steinitz
Defense Doubly Deferred line
(7 :el d6) that Akiba Rubinstein
liked to playas Black and is similar
to the better forms of the old
Steinitz after 8 d4 exd4 9 ttJxd4
Jld7.
White might improve with
7 ttJxe5 ttJxe4 or 7 d4 ttJxe4 8 dxeS
0-0 (9 :e 1 dS 10 ttJd4 Jld7 11 f3
ttJgS as in I.Sokolov-S.Nikolic,
Pancevo 1987) but there's too little
experience to tell.
A major benefit of a delayed
Jlxc6 is that Black finds it harder to
defend his e-pawn. After 3 ... a6
4 Jla4 ttJf6 S Jlxc6 dxc6 he cannot
play... f6 as he would after 4 Jlxc6
dxc6. However, 6... Jld6! is a good
developing defense and Black
stands well after 7 ttJbd2 Jle6.
OPEN DEFENSE
Before getting to the main Open
line, 1 e4 e5 2 ttJf3 ttJc6 3 Jlb5 a6
4 Jla4 ttJf6 5 0-0 ttJxe4 6 d4 b5
7 Jlb3 d5, let's consider 7...Jle7!?
The reason the doubly delayed
exchange, 5 0-0 Jle7 6 Jlxc6 dxc6,
has a better reputation is that
7 ... Jld6 would cost a tempo.
Instead, Black obtains a solid game
with ... ttJd7. If he wants a
unbalanced middle game, with
greater winning chances, he should
consider the rare 6...bxc6!?:
42
RuyLopez
i..e3, lbbd2) and at Black's ( ... i.e7,
...lbc5, ... lba5) can be mixed in a
variety of orders, making transpositions inevitable.
This tries to lure White into
8 :el and then 8... d5 9 dxe5 .te6,
a dead even position that could
come about after 7 ... d5 8 dxe5 .te6
if White plays the inept 9 l:.el?!
.te7.
In the first edition of ECO,
Viktor Korchnoi endorsed the
conventional wisdom that 9 c3 .te7
10 .te3 and 9 c3 .te7 10 lbbd2
were so favorable that Black should
avoid them with 9 ... .tc5!. But
Andrei Sokolov showed how White
can trick his opponent into a
favorable line with 9 .te3!? and
then 9 ....te7 10 c3!. This is very
effective against booked-up 9 c3
.tc5 opponents.
The best try at punishing 7 ... .te7
is 8 dxe5. Experience with 8...lbc5
9 i.d5 .tb7 (or 9 ...lbe6) has been
far from convincing, e.g. 9 ... .tb7
10 lbc3 0-0 11 .tf4 lbe6 12 .tg3
f5!.
If Black likes these positions he
can get to them even if White
adopts a standard way of avoiding
the Open Defense, 5 d4. This often
favors a better-prepared White after
5... exd4 6 0-0 or 6 e5. But Black
has 5.•.b5!? instead:
For instance, 10 ... 0-0 11 lbbd2
'ild7 12 :el :ad8 13 .tc2 lbxd2
14 'ifxd2 .tg4 15 'ifd3 g6 16 .th6
:fe8 17 :adl lba5 18 h3 .txf3
19 'ilxf3 c5 20 'ilg3 lbc4 21 b3
lba3 22 e6! 'ifc6 23 .tf5! and wins,
Kamsky-Piket, Dos Hermanas
1995.
Then 6 dxe5! lbxe4! 7.tb3lbc5
8 .td5 i.e7 9 0-0 transposes to the
previous line.
The tabia of the Open Defense
arises after 1 e4 e5 2 lbf3 lbc6
3 .tb5 a6 4 .ta4 lbf6 5 0-0 lbxe4
6 d4 b5 7 .tb3 d5 8 dxe5 i.e6. The
moves at White's disposal (c2-c3,
There is no obvious antidote to
9 .te3 but attention has shifted to
9 lbbd2. It allows White to answer
9 ... .tc5 with 10 lbxe4 dxe4
43
RuyLopez
11 it.xe6 with at least a small White
edge. That often prompts 9 ... it.e7,
after which 10 c3! is another good
transposition.
8 ... l:tb8 9 axbS axbS, he can reach a
much improved Open Defense,
10 dxeS it.e6 11 c3.
But 8 a4? was knocked into a
footnote by the Lasker-Schlechter
match of 1910 when Black
equalized with 8... tiJxd4!'
The downside to 9 tiJbd2 is that
White has diminished control of d4,
compared with 9 it.e3 and 9 c3.
That's significant after 9.•.tiJcS!.
The appeal of the tweaked move
order, 8 dxeS it.e6 9 a4, is that
9 ... l:tb8 allows White to reach that
ideal version of 8 a4 following
10 axbS axbS 11 c3. For instance,
11...it.e7 12 tiJd4! tiJxeS (12 .. :iVd7
13 it.c2!) 13 f3 tiJcs 14 it.c2 it.d7
IS b4 with advantage, AlekhineRohacek, Munich 1941.
Then White would be happy to
see 10 c3 it.e7 11 it.c2 it.g4 because
that transposes to yet another good
9 c3 line (9 c3 tiJcs 10 it.c2 it.g4
11 tiJbd2 it.e7). For example,
12 .:tel 'ifd7 13 tiJf1 lId8 14 tiJe3
it.hS IS tiJfS.
But Black has been doing better,
after 10 c3 with 10•.• d4. That has
become a main line, with book
analysis stretching past move 20.
So what's wrong with 9 a4 ?
Vintage theory says 9•..tiJaS is the
answer. But 10 axbS axbS 11 tiJd4!
it.cs 12 c3 0-0 13 it.c2 is hardly
convincing. More testing is needed
(say of 9 •.• b4 10 as, threatening
it.a4) before we can properly
evaluate 9 a4!?
A trickster will prefer yet another
ninth move, 9 a4!?
MOELLERIARKHANGEL
DEFENSE
White is trying to slip into a
forgotten line that begins 1 e4 eS
2 tiJf3 tiJc6 3 it.bS a6 4 it.a4 tiJf6
S 0-0 tiJxe4 6 d4 bS 7 it.b3 dS and
now 8 a4, instead of 8 dxeS. If
White gets control of an open a-file,
44
There's no better example of how
transpositions can improve a group
of related openings than the
aggressive systems in which Black
mixes ... a6/... tiJf6 with ... i.cs and
possibly ... bS/ ... i.b7. Up until
fairly recently they seemed like
distant cousins unworthy of our
attention:
The Moeller Attack (1 e4 eS
RuyLopez
2 ~f3 ~c6 3 i..bS a6 4 i..a4 ~f6
S 0-0 i..cS) was considered too
flaky because of 6 c3 i..a7 7 d4 or
6 ~xeS ~xeS 7 d4.
The Arkhangel, S•.. bS 6 i..b3
i..b7, enjoyed 15 minutes of fame
in the 1960s. That expired when
White obtained an edge on both
wings with 7 .l:tel i..c5 8 c3 d6 9 d4
i..b6 10 i..g5 and a2-a4.
But then came the 'New
~
~
Arkhangel,' S...bS 6 ....
b3 ....
cS.
commonly come from 5 'iVe2 b5
6 i..b3 i..c5 7 0-0 and lead to
roughly equal chances.
If White abandons c2-c3/d2-d4
and plays 7 d3 Black can safely
revert to the old Arkhangel, 7... 0-0
8 a4 i..b7. That's an 'old' position
that occurs after 5 0-0 b5 6 i..b3
i..b7 when White passes up the
critical lines such as 7 .:tel. Again
Black gets a free ride in the
opening.
A key test of the New Arkhangel
is 7 c3 d6 8 d4 i..b6:
Black's cunning lies in holding
back his QB until he knows whether
it serves best on g4 or b7. For
example, on 7 .l:tel he avoids
7... i..b7 - which transposes to the
unfavorable old Arkhangel - and
plays 7... ~g4! in the Moeller spirit.
That's good for him, e.g. 8 d4?
~xd4 9 ~xd4 'ii'h4! or 8 .l:te2 ~d4
9 ~xd4 i..xd4, threatening
10 ... 'ii'h4.
The difference between new and
old is revealed by 9 .:tel. Then
9 ... i..b7? transposes to an inferior
'old' position. But after 9...i..g4! we
are 'new.' White can't support his
center on the current squares and he
has no edge after 10 d5 ~e7 11 h3
i..d7! and ... c6.
What about a2-a4, the usual
antidote to an early ... b5 ? That
shouldn't trouble Black in the last
diagram, e.g. 9 a4 i..g4! 10 axbS
axbS 11 .l:txa8 'iVxa8 12 h3 i..hS
If White concludes that he should
be playing 7 'i'e2 instead, he is
merely transposing into one of the
Worrall/Wormald
tabias
that
45
RuyLopez
has perfonned well, e.g. 13 "ifd3
exd4 14 ii.g5 ii.xf3 15 ii.xf6
ii.xg2!, Pavlovic-Carlsen, Reykjavik 2006.
face the more critical 8...l:tb8 and
8... ii.g4 lines.
Despite that, the .. .ii.c5 family is
respectable today - because so far
Black has found more good
transpositions than White.
Is there nothing but good news
for Black in the New Arkhangel?
Not quite. He has to know what to
do after 7 ttJxeS. The amount of
MARSHALL and
theory on 7 ... ttJxe5 8 d4 ii.xd4
ANTI-MARSHALL
9 'ilfxd4 d6 has exploded in recent
years and the debris hasn't settled
In the 1930s Frank Marshall's
yet.
gambit was considered refuted. In
Moreover, White can set his own the 1960s it was resuscitated by
transpo traps. With 7 a4 he forces Boris Spassky - as a drawing
Black to meet the threat ofaxb5. weapon. Today it is perhaps the
most dangerous Ruy Lopez
Then 7...ii.b7
defense. That adds huge significance to Black's seventh move:
1 e4 eS 2 ttJf3 ttJc6 3 .tbS a6
4 .ta4 ttJf6 S 0-0 .te7 6 l:tel bS
7 ii.b3.
8 d3 is another quiet line, an antiMarshall in which .. .ii.c5 has been
played in place of ... ii.e7. That
would seem to benefit Black but
White can confuse matters with
8 c3, hoping for 8...d6 9 d4 ii.b6
10 ii.gS!, a favorable 'old' position.
With 7 ... d6 Black commits
himself to a main-line variation
(8 c3 0-0 9 h3 etc.). But 7... 0-0
allows him to threaten to play the
Marshall Gambit, 8 c3 d5!?
What
happened:
We've
transposed into 7 c3 d6 8 a4!? ii.b7
9 d4 ii.b6 10 ii.g5. This helped
White because he didn't have to
It's just a threat because he can
46
RuyLopez
back out with 8... d6, tranposing into
the main line. Yet the threat is so
worrisome that today's strongest
GMs skirt the issue with modest, if
not cowardly anti-Marshall moves
like 8 d3 or 8 h3.
Why then should anyone play
7... d6 if it enables White to play
more aggressively without risk?
The only downside to 7... 0-0 is
supposed to be that 8 a4 is stronger.
It is - but not significantly, after
8 a4.i.b7.
That's why top GMs have been
hunting for an aggressive way to
sidestep the Marshall. One idea is
8 d4.
Alexander Grischuk among others.
And what about other antiMarshall lines? The simplest begins
with 8 d3. Then 8... d5 is considered
unsound since c3 is available for a
knight (9 exd5 tiJxd5 10 tiJxe5
tiJxe5 11 :xe5 c6 12 l:te 1 .i.d6
13 tiJc3!).
Black typically meets 8 d3 with
8...d6. Following 9 c3 White will
eventually play d3-d4. This ensures
he can execute the tiJf1 maneuver
that is so important to the Lopez.
He plans 10 tiJbd2, 11 tiJf1, 12 tiJg3
and eventually d3-d4.
He doesn't get to do that in some
8 c3 lines, such as the Zaitsev
Variation (8 ... d6 9 h3 .i.b7 10 d4
:e8 11 tiJbd2 .i.f8 12 tiJf1? exd4!
and Black wins a pawn).
The downside to 8 d3 d6 9 c3 is
White will be a tempo behind
traditional 8 c3 d6 9 h3 lines when
he eventually advances d3-d4. But
that tempo may be a minor concern
This works best when Black compared with dealing with the
becomes concerned about the looks Marshall.
of 8... tiJxd4 9 tiJxd4 exd4 10 e5
Note that one of the few active
tiJe8 and bails out with 8... d6.
responses Black has is 9 ... .i.e6.
Then 9 c3! transposes into 8 c3 Black seems to be doing well when
d6 9 d4 .i.g4. Books say that old you compare this with 8 c3 d6 9 h3
line is second-best to 8 c3 d6 9 h3. .i.e6 10 d4!, which theory says
But second-best doesn't mean it's favors White after 10 ... .i.xb3
bad and the old line is used by Gata 11 axb3 :e8 12 d5! and 13 c4.
Kamsky, Magnus Carlsen, and
But is Black equal after 8 d3 d6
47
RuyLopez
9 c3 .te6 and then 10 d4 ?
The test of Geller's order is
9... tbxd4, since 10 .txt7+ l1xt7
He is a tempo ahead of the
11 tbxe511f8! 12 'iixd4 c5 is a good
previous line. But all that means is
gambit and 10 tbxd4 exd4 11 e5
that h2-h3 is missing. Black best
tbe4 12 'iig4 d5 has performed well
may be 10... i.g4!, transposing to
for Black. Nevertheless, 9 d4 may
8 c3 d6 9 d4 .tg4, that old,
be worthwhile because allowing the
reasonable system for White.
Marshall is worse - and because the
Another anti-Marshall finesse is most natural response from Black is
8 h3. Then if Black continues fleeing with 9 ... d6 10 c3 into
8••• d6, as he often does after 8 d3, something more familiar to him.
White can shoot back 9 c3!. He has
transposed to 8 c3 d6 9 h3 while
Because of the forcing nature of
dodging the Marshall.
the main Marshall lines, there is
Books used to dispose of 8 h3
with the comment that 8 ....tb7
keeps the gambit alive. That is, 9 c3
d5! 10 exd5 tbxd5 11 tbxe5 tbxe5
12 l1xe5 tbf4! is a prohibitively
strong attack.
little room for transposition
following 8 c3 d5 9 exd5 tbxd5
10 tbxe5 tbxe5 11 l1xe5 c6. An
exception occurs in Fischer's idea
of 12 g3.
Yefim Geller pointed out the
virtue of 9 d4!? instead.
Then 9..• d6 10 c3! transposes to
the Zaitsev Variation. If White
prefers the Zaitsev to the Marshalland most Lopez players do - then
he's scored a minor victory.
48
RuyLopez
This is designed to bypass the
mega-theory of 12 d4 .td6
13 .:tel 'iVh4 14 g3 'iVh3.
Arthur Bisguier once explained
in Chess Review why he used 5•.. b5
6 .tb3 d6!?
The simplest way of handling
12 g3 is 12 ....td6 13 .:tel:
And now 13 ..:iVd7! and ... 'iWh3
transposes to the line White is
trying to avoid.
One of the few variables in the
Marshall is the timing of .txd5, if
that's White's wish. He can do it
with 12 d4 .td6 13 .:tel 'iWh4 14 g3
'iVh3 15 i.xd5 exd5 or wait for
First, he wanted to avoid his
15 .te3 .tg4 16 'iVd3 .:tae8 and opponent's
favorite
Delayed
then play 17 .txd5.
Exchange (5 ... .te7 6 .txc6)
Or he can wait further, 17 ttJd2
.:te6 18 'iWfl 'ifh5 19 .txd5. He can
Variation. Second, White 'couldn't
be sure I would transpose into the
main line and he felt compelled to
analyze the complexities of 7 ttJg5. '
Opinions vary about the soundness
of7 ... d5 8 exd5 ttJd4.
even reach the same positions with
12 i.xd5 exd5 13 d4 i.d6 14 .:tel
'iVh4 15 g3 'ifh3.
Which is best? Perhaps the last
order because - aside from the extra
14 .:te3 option - it bypasses another
Geller line, 12 d4 i.d6 13 .:te 1 'iVh4
14 g3 'iVh3 15 i.e3 i.g4 16 'iVd3
ttJxe3!? 17 .:txe3 c5.
White eventually selected 7 e3
and Black transposed after 7... i.e7
8 h3 ttJa5 9 i.e2 e5 10 d4. But by
then he had gained more than an
hour (!) on the clock.
ROAD MAPS TO THE
RUYMAIN
Black reaches the main Ruy tabia
after 1 e4 e5 2 ttJf3 ttJe6 3 i.b5 a6
4 i.a4 ttJf6 5 0-0 i.e7 6 .:tel b5
7 .tb3 0-0 8 e3 d6 9 h3. Other
orders give White more options but
that also means more to think about.
49
RuyLopez
That's an extreme case of 'giving
him something to think about. ' Like
other alternative routes to the Ruy
tabia, this one comes with a price.
White has saved a tempo compared
with the usual main lines.
He doesn't have to play l:.el,
since ~c2 protected the e-pawn. In
other routes he can avoid h2-h3
because ... ~g4 is not a concern. Yet
This position occurred several
as we'll see White may have to play times in the 1950s as a means to a
h2-h3 and l:tel after all and give Marshall-like gambit, 8... d5.
back the tempi. If he does, Black
can derive some short-term gains
However, Black also has 8.••d6,
through these alternative orders.
with the hope of reaching an
Another is 1 e4 eS 2 ltJf3 ltJe6 improved, or at least confusing,
3 ~bS a6 4 ~a4 ltJf6 S 0-0 bS version of the Ruy main line after
6 ~b3 and now 6... ~e7. Then 9 d4 ltJaS!? That leaves a pawn
hanging but no one who has faced
7ltJg5? O-O! is pointless.
this as White seems to trust 10 dxe5
This order has the usual ... b5 dxe5 llltJxe5 ltJxb3.
drawback of allowing 7 a4. Then
7... b4 has a poor reputation because
Instead 10 .te2 eS has been
of 8 'iVe2 0-0 9 a5! and 10 c3. played. When Paul Keres reached
Perhaps better is 7... ~b7 and this position he rejected 11 dxe5
8 axb5 axb5 9 l:txa8 and 10 ltJc3 dxe5 12 'i'xd8+ l:.xd8 13 ltJxe5
~xe4! and played llltJbd2.
but there is little experience to
prove it.
Another all-but-forgotten order
What's happened is that we've
(after 1 e4 e5 2 ltJf3 ltJc6 3 ~b5 a6 reached a main Lopez line but with
4 ~a4 ltJf6 5 0-0) is S•.• ~e7 6 l:.e1 the omission of h2-h3 and ... 0-0.
bS 7 .ib3 ~b7. This also comes That should favor White, not Black.
about via the Arkhangel if Black But it's hard to prove that if Black
puts his KB on e7 rather than c5. continues to delay castling with
White has nothing better than 8 e3: 1l ••• exd4 12 exd4 l:.e8.
50
Ruy Lopez
This is not just a trap, 13 dxe5
dxe5 14 liJxe5? 'iie7!. It deserves
the serious testing it hasn't gotten.
Instead he usually plays 14 i.b1
and then 14... 0-0 15 liJf1.
That transposes to a position that
normally comes from 5 0-0 ~e7
6 :el b5 7 ~b3 0-0 8 c3 d6 9 h3
liJa5 10 .tc2 c5 11 d4 'iic7
12 liJbd2 cxd4 13 cxd4.td7 14 liJf1
:ac8 when White passes up
15 liJe3! and 15 .l:!e2! and settles for
15 ~b1.
OLD MAIN LINE
The way the main Ruy tabia was
reached for much of the 20th
century was 5 0-0 ~e7 6 :e1 b5
7 ~b3 d6 S e3 and now S•••liJa5
9 .te2 e5. This was considered a
finesse because it avoided 8... 0-0
9 d4, at a time when that line was
regarded as potent. The 8... liJa5
order began to disappear in the
1950s but was revived by Bent
Larsen and Oleg Romanishin.
In this order White didn't get that
choice and the result is a roughly
equal position (15 ... l:.fe8 16 liJg3
liJc6).
Why was it revived? One reason
is that by delaying castling Black
can meet 10 d4 'iie7 11 h3 with
a quick ... l:tc8 that threatens the
c2-bishop earlier than usual.
Another plus of the Old Main
Line is that Black isn't castled after
8... liJa5 9 .tc2 c5 10 d4 'iic7 11 h3
liJe6. That means 12 d5? - which is
For example, 11 ...~d7 12 liJbd2
exd4 13 exd4 :eS. Since White
hasn't had time for liJf1, he can't
defend the bishop with :e2 or liJe3,
as he does in the orthodox main
line.
a good move when liJbd2 and ... 0-0
are added - turns out well for Black
after 12 ...liJbS 13 liJbd2.
Instead of 13 ... 0-0? 14 a4 ...
51
RuyLopez
use the 11.. ..id7 12 ltJbd2 cxd4
13 cxd4 :c8! 14.ib 1 0-0 trick.
MAINLINES
A modem main line runs 1 e4 e5
2 ltJf3 ltJe6 3 .ib5 a6 4 .ia4 tiJf6
5 0-0 .ie7 6 :e1 b5 7 .ib3 0-0 8 e3
d6 9 h3 and now 9...ltJa5 10 .ie2
e5 11 d4. Black typically defends
the attacked e-pawn with ...'fic7,
... ltJd7 or ...ltJc6, with or without a
trade of c-pawns.
... he can attack with 13 ... g5!
(14 ltJxg5 :g8), with advantage
according to Aleksandr Matanovic.
Often he adds ... :e8 so that after
... .if8 he smoothes out his
development and puts pressure on
e4, e.g. 11...'ifc7 12 ltJbd2 cxd4
13 cxd4 .ib7 14 ltJf1 :ac8 and
15 ...:fe8.
The drawback to the Old Main
Line is that White saves a tempo by
not playing h2-h3. But that may be
temporary. For instance, 11 ltJbd2
0-0 12 ltJf1?, rather than 12 h3!,
allows 12 ... cxd4 13 cxd4 .ig4! with
excellent play.
More economical is Yaacov
Murey's gambit idea, 11 ...:e8!?
It's based on 12 dxe5 dxe5
13 ltJxe5 .ib7 with compensation
for the sacrificed pawn.
Tal was plainly frustrated with
the Old Main Line when Larsen
adopted it against him in 1991. Tal
knew that the books claimed 10 d4
'ifc7 11 a4 was the refutation. But
as he studied the position he
concluded 11 ... e4! (instead of the
book 11...h4?) was not at all bad.
He also noticed that after
11 ltJbd2 Black can reply 11 ....id7
12 d5 0-0 and when White plays
ltJf1-e3, Black replies ... ltJg4!,
which looked good. So after
considerable thought Tal just played
11 h3 and transposed after all.
For example, 14 ltJd2 .id6 or
14 'iff3 .if8 15 .if4 :e6 16 ltJg4
ltJc4 17 ltJxf6+ :xf6 18 tiJd2 tiJe5,
Chandler-Hebden, Millfield 2000.
This was the final game between
these two great warriors and it
ended in a draw after Black got to
52
RuyLopez
But the principal idea of 11 .. J::te8
What made the Zaitsev playable
is to save a tempo by omitting was the realization that 9...:e8
... 'iic7. For example 12 liJbd2.tts 10 liJg5:
13 liJf1 gives Black good
counterplay after 13... cxd4 14 cxd4
exd4 15 liJxd4 i.b7.
Compare this with 11...'iic7
12 liJbd2 :e8 13 liJfl .tfS when
White gets the upper hand from
14 i.g5!.
By avoiding ... 'iic7 Black may
find a better use for the queen, e.g.
Now 10...:ts! is perfectly safe
11...l:te8 12 liJbd2 .tfS 13 b3 liJc6
for Black. White has nothing better
14 i.b2 .td7 15 a3 'iib6! with good
than 11 liJf3 !, repeating the
play. In contrast, 11 ... 'iic7 12 liJbd2
position. In fact, grandmaster draws
:e8 13 b3 i.fS 14 liJfl favors
have ended here or after a few more
White.
repetitions.
Igor Zaitsev's contribution to the
Ruy occurs when Black meets the
tabia's 9 h3 with 9...:e8 10 d4
i.b7. He prevents the liJd2-fl
maneuver that is so valuable in the
Ruy, as noted earlier.
Zaitsev's original move order is
inexact if Black wants to avoid a
draw as well as the complications of
9...:e8 10 a4 liJa5 11 .ta2. More
precise is 9....tb7! first and then
10 d4 :e8, transposing.
53
Chapter Three:
Sicilian Defense
Transposition tricks punish
routine moves more often in the
Sicilian than in any other opening.
That might be expected in sharp
Sicilian lines but it also applies to
the closed variations.
White has a strong sacrifice, 8 fS!
gxfS 9 tbh4 fxe4 10 dxe4 followed
by tbf5 or 'i'h5, e.g. 10 ... ~e6
11 tbf5 .ltxf5 12 exf5 f6 13 tbc3
0-0 14 tbd5, Korolyev-Varlamov,
correspondence 1981.
For example, White can try to
save a tempo compared with the
traditional closed order, 1 e4 cS
2 tbc3 tbc6 3 g3, with 2 g3. The
Russian postal champion Sergey
Korolyev showed the benefits of
this following 2...tbc6 3 .ltg2 g6
4 d3 .ltg7 S f4 d6 6 tbf3.
If Black chose 6... tbf6 instead,
Korolyev was content to transpose
to a normal closed with tbc3, since
many players think ...tbf6 is one of
the less effective closed defenses.
The drawback to 2 g3 is 2... dS!.
White doesn't have time for 3 d3
and 4 tbd2 because of 3... dxe4! and
Black has been playing auto-pilot he may not like 3 exd5 'iVxd5 4 tbf3
moves but if he continues 6..•eS .ltg4 5 .ltg2 'iVe6+!.
70-0 tbge7?!:
Alexander Morozevich refined
this order with 2 d3!? That enables
White to transpose, after 2... g6 3 g3
.ltg7 4 .ltg2 tbc6 5 f4 d6 6 tbf3 e5
7 0-0 tbge7 8 f5!, to Korolyev's
sacrifice. But it also allows him to
safely meet a quick ... d5 with tbd2!
and obtain a satisfactory King's
Indian Reversed.
Thanks to his saved (tbc3) tempo
54
Sicilian Defense
CHAMELEON SICILIAN
This elastic system is based on
White's ability to delay a decision
of whether to open the center with
d2-d4. It works because the
sophisticated Black moves of an
open line, such as ... a6 and ...'iic7,
will be out of place in a closed one,
as Vladimir Simagin first showed in
games like Simagin-Portisch,
Plovdiv 1959, 1 e4 c5 2 liJf3 e6
3 liJc3!? a6 4 g3 b5.
liJc6 3 liJge2 e6 4 g3.
Black indicates his willingness to
play a Taimanov or Scheveningen
variation and White is hinting at a
closed line. But if Black naively
responds 4... g6?, expecting 5 .tg2
.tg7
... he'll be surprised by 5 d4!
cxd4 6 liJxd4. This exposes a nasty
hole at d6 (6 ... .tg7? 7liJdb5!).
Similarly 4... liJge7 5 .tg2 g6?
6 d4!. Black never has this problem
in normal Closed Sicilian orders
such as 1 e4 c5 2 liJc3 liJc6 3 g3 g6
4 .tg2 .tg7 because he has too
much firepower focused on d4.
Black's setup would serve him
well after 5 d4 cxd4 but not in the
game's continuation, 5 .tg2 .tb7
6 d3! liJc6 7 0-0 d6 8liJg5. Despite
his opening expertise in the open
And suppose Black abandons
Sicilian, Lajos Portisch appeared
clueless after 8... h6 9 liJh3 liJf6 ... g6 and goes instead for a quick
10 f4 i.e7 11 f5! e5 12 liJf2 liJd4 ... d5, as in 1 e4 c5 2liJc3 e6 3 g3 d5
13 liJb I! liJd7 14 c3 liJc6 15 a4 4 exd5 exd5. That's a perfectly
i.g5 16 .txg5 'iixg5 17 axb5 axb5 valid anti-closed weapon.
18 lha8+ i.xa8 19 liJa3 liJa7
But 2... e6 is a more committal
20 d4!.
move than Black may suspect and
On the other hand, if Black after 3 liJf3 he cannot stop 4 d4. If
assumes he is headed into a closed Black is a Najdorf or Dragon
Sicilian, White can switch to an player, he'll find himself tricked out
open one. This is the case of 2 liJc3 of his repertoire.
55
Sicilian Defense
players are not familiar with. And
5... a6 is trickier than you might
think - 6 'ii'd3 tbc6 7 ..tf4 'ii'c7 8 00-0 tbe5?? 9 tbxe5 dxe5 10 ..txe5!
Resigns was Benjamin-Hrop,
The first version is the most Parsippany 2003.
deceptive. After 2 tbf3 a Taimanov
The Dragon player may feel at
Variation player might reply 2 ... e6 ease in this first Chameleon version
because he wants to avoid 2 ...tbc6 because he can meet 2 tbf3 tbc6
3 ..tb5, the Rossolimo Variation. 3 tbc3 with 3... d6 4 d4 cxd4
But 3 tbc3 poses a problem.
5 tbxd4 g6. But this is a Modern
Then 3... tbc6 allows one of the Dragon Variation with a slight
better editions of the Rossolimo, difference: ...tbc6 has been played
4 ..tb5!. And 3 ...tbf6 transposes in place of ... tbf6. That means
after 4 e5 tbd5 into the dubious 6 tbd5! is possible. Then 6.....td7
Nimzovich Variation (usually 7 ..tg5 is annoying and 6... e6 7 tbc3
reached via 2 tbf3 tbf6 3 e5 tbd5 exposes holes.
4 tbc3 e6).
There are three basic versions of
the Chameleon: White can bring his
knights to f3 and c3. Or he can play
2 tbe2 and 3 tbbc3. Or he can begin
with 2 tbc3 and then 3 tbge2.
The most flexible response to
2 tbf3 e6 3 tbc3 is 3... d6. Then 4 d4
cxd4 5 tbxd4 opens a path to
various mainstream Sicilians
(5 ...tbf6, 5... tbc6, 5... a6). However,
White can surprise him with
5 'i'xd4!?
If Black replies 6.....tg7 instead
he is not lost - as Alexey Suetin
claimed in the first edition of ECO.
But he is worse after 7 tbb5! :b8
8 c4.
The Dragon player might refme
his order and meet 2 tbf3 with
2..• d6 3 tbc3 g6. But he lands in
new and complicated territory after
4 d4 cxd4 5 'ii'xd4! (5 ...tbf6 6 e5
tbc6 7..tb5).
Then 5...tbc6 6 ..tb5 is a position
(usually reached via 2 tbf3 d6 3 d4
cxd4 4 'iixd4) that most 2 ... e6
56
Sicilian Defense
But life isn't so simple for White
But this is how Bobby Fischer
in this Chameleon order. A major outsmarted himself in his 1992
drawback to 2 tDf3 tDc6 3 tDc3 or rematch with Boris Spassky.
2... d6 3 tDc3 is 3... e5!. Then d2-d4 Spassky switched to a closed
is impossible and he is stuck in a Sicilian, 3 tDbc3 d6 4 g3! tDc6
closed Sicilian with his KN 5 i.g2 g6 6 0-0 i.g7 7 d3 0-0 8 h3
misplaced at D, where it blocks his and 9 f4 - an excellent weapon
f-pawn.
against Fischer, who had always
preferred ... f5 or ... e6/ ... tDge7
For example 2... tDc6 3 tDD e5
systems, not those with ... tDf6.
4 i.c4 d6 5 0-0 i.e7 6 a3?! tDf6
Perhaps Black's best policy after
7 :b 1 0-0 8 b4 a6 9 d3 b5! as in
J.Polgar-Kramnik, Paris 1994. 2 tDe2 is allowing an open Sicilian,
White can improve on 6 a3?! but so say with 2...tDc6 3 tDbc3 d6 4 d4.
He delays ... e5 until White has
far 3... e5! has performed well.
given up on tDd5/i.c4, such as 4 g3
Jose Capablanca and Paul Keres e5!.
tried to solve the ... e5 problem with
There are slight differences in the
the second Chameleon version,
2 tDe2 and 3 tDbc3. The difference third Chameleon order, 2 tDc3 and
is that 2... tDc6 3 tDbc3 e5 allows 3 tDge2. A Najdorf player can be
4 tDd5! followed by tDec3! and expected to reply 2..•d6 and 3... a6.
i.c4, which coordinates the White Then after 4 d4 cxd4 5 tDxd4 he
has a choice between transposing
pieces well.
into a normal Najdorf with 5... tDf6
The natural response to 2 tDe2 is and the extra option of 5... e6!?
2...tDf6. This attacks the e-pawn
That avoids the 6 i.g5 Najdorf
and invites the complications of
3 e5?! tDg4 or 3 ... tDd5 4 tDbc3 e6. and gets good versions of other
Najdorfs, such as 6 f4 b5 7 i.d3
i.b7 8 0-0 tDd7 9 f5? 'iib6!. It also
creates a reasonable Keres Attack
after 6 g4, e.g. 6... tDe7 7 a3 tDbc6
8 tDb3 b5 as in Vallejo PonsTopalov, Leon 2006.
The problem with the Najdorfwannabe order is Black has to
justify ... a6 after 4 g3!.
57
Sicilian Defense
Then 4 g3 .ltg7 is an innocuous
Closed Sicilian in which White's
KN is once again misplaced. Better
is 4 d4 cxd4 5 tDxd4 i.g7. If Black
has any Dragon instincts he'd be
happy here because he enjoys the
benefits of the accelerated order
without allowing White his
strongest weapon, the Maroczy
Bind.
Now 4 ... e5 is not nearly as
effective when White's f-pawn is
unblocked. He is virtually a move
ahead of normal closed Sicilians
after 5 .ltg2 tDc6 6 d3 g6 7 f4 .ltg7
8 0-0 tDge7 9 .lte3 0-0 10 'i'd2
tDd4 11 litf2.
Note that Black also gets an extra
option in both the 2 tDc3/3 tDge2
and 2 tDc3/3 tDf3 orders. After
2 ... tDc6 and 3 ... d6 4 d4 exd4
5 tDxd4 he can try 5... e5.
But White can also feel happy
with this move order, because he
discouraged Black from a slew of
other openings he might have
prepared, such as the Taimanov or
Sveshnikov.
This has the benefit of reaching a
Sveshnikov Variation (6 tDdb5 a6
7 tDa3 b5 8 tDd5) but with the
possibility of 8 ...tDce7!?, which has
tested well. However, White also
has an extra option, 6 tDf5!?, and
this entire 5 ... e5 line must be called
unclear.
GRAND PRIX ATTACK
The Grand Prix was once
considered a patzer's opening. But
no less a Sicilian authority than
Yevgeny Sveshnikov said that
allowing it by answering 2 f4 with
2 ... tDc6 was a serious mistake.
'Already Black is worse,' he wrote.
The biggest problem with all the
Chameleon orders is the accelerated
form of the Dragon.
White's basic plan in the attack is
tDc3 and f2-f4 followed by tDf3 and
posting his KB on c4 or b5. If Black
After 1 e4 c5 2 tDf3 tDc6 3 tDc3
g6!:
58
Sicilian Defense
could count on White not transposing, his best order would be 1 e4
c5 2 lLlc3lLlc6 3 f4 e6 4lLlnlLlge7:
What about a more flexible - and
very common - order such as 1 e4
c5 2 lLlc3 lLlc6 3 f4 d6 4 l'bn e6 ?
Then 5 i.b5 can be firmly answered
by 5... l'bge7 and 6... a6!. But on
5 g3!:
In that way he can meet 5 i.b5
with 5... a6! and 5 i.c4? with
5... d5!.
White gets a good Chameleon,
e.g. 5...lLlge7 6 i.g2 and if 6... g6
then 7 d4! exposes the d6 hole
again.
But the greatest strength of the
Grand Prix today is its transpotential. White can switch to an
open Sicilian, 5 d4 cxd4 6 lLlxd4.
The result is perfectly good version
of the Taimanov Variation. But
there's quite a bit of theory for
Black to know - and if he's not a
Taimanov player, he's been tricked
out of his repertoire.
Better is 4... a6. But 5 d4 cxd4
6 lLlxd4 lLlf6 would land Black in a
strange Najdorf. That is, 1 e4 c5
2 l'bf3 d6 3 d4 cxd4 4 lLlxd4 l'bf6
5 l'bc3 a6 and 6 f4 l'bc6?!, e.g.
7 i.e3 i.d7 8 'iYn g6 9 l'bd5! ':c8
10 l'bxf6+ exf6 11 c3 is excellent
for White (F 0 garasi -Ftacnik,
Austria 2006).
This kind of switch happens
because Black has to make pawn
and piece commitments. That
provides White with the information he needs to decide whether to
transpose into another Sicilian line.
In the last diagram for example,
White could play 5 g3, creating a
closed Sicilian in which Black has
given up on the recommended
... lLlf6 and ... e5 defenses.
Most likely the best answer to
1 e4 c5 2 l'bc3 l'bc6 3 f4 d6 4 l'bf3 is
neither 4... e6 nor 4... a6 but 4... g6.
Black should be happy with 5 i.b5
i.d7, a familiar Grand Prix
position. But he had better know a
lot of Dragon theory as well
because of 5 d4!?
59
Sicilian Defense
The Grand Prix is a particularly
good weapon against Najdorf,
Scheveningen and Modem Dragon
players because they'll meet 1 e4 c5
2 liJc3 with 2•.. d6 in the hopes of
transposing (3 liJf3 liJf6 4 d4) or
punishing White's order (3 liJf3
e5!).
The trickster can try to exploit
Kasparov's order with the
paradoxical 5 .tb5+ .td7 6 .tc4!?
His point is that Black's bishop gets
in the way of supporting ... d5.
There are only a few examples of
this so far (6 ... liJc6 7 0-0 e6 8 d3
liJge7 9 'i'el a6 10 f5!? gxf5
But there's a problem with 2... d6. 11 'i'h4 b5 12 .tb3 'iVb6,
Some of the best forms of the Grand Ramon Perez - Movsziszian,
Prix for White are those with ... d6. Balaguer 2006). But this may be a
For example, 3 f4 liJc6 4 liJf3 g6 finesse with a future.
5 .tb5! is a pin that rules out
And there is one more headache
... liJd4, a strong idea when the pawn
for Black after 3... g6. White can
is at d7.
reply 4 d4 cxd4 5 'i'xd4. This
Garry Kasparov's solution was
allows him a promising pawn sack
1 e4 c5 2 liJc3 d6 3 f4 g6 4 liJf3
after 5... liJf6 6 e5 liJc6 7 .tb5
.tg7:
(7 ... dxe5 8 'i'xd8+ ~xd8 9 fxe5!
liJxe5 10 .tf4 liJed7 11 0-0-0).
ALAPIN VARIATION
Simon Alapin - pronounced
ah-LAH-peen, in case you were
wondering - gets the credit for 2 c3.
But the German player H. W.Popert
played it, as well as the wily
This is far superior to 4 ... liJc6 5 2 liJf3 I 3 c3, more than 50 years
.tb5 .tg7?! 6 .txc6+!, which before Alapin.
favors White. Black can meet
The benefit of the second order is
5 .tb5+ with 5....td7!.
that White gets to see Black's
second move before he commits
himself to c2-c3. That's significant
because after 2 liJf3 liJc6 3 c3!?
Black has lost some of his best 2 c3
options.
He also invites a Dragon, 5 d4
cxd4, and can meet 5 .tc4
with 5 ...liJc6 6 0-0 e6 followed
by kicking the bishop with
... liJge7/ ... d5
60
Sicilian Defense
The difference is 2 ttJO e6 3 c3
ttJe7!? 4 d4 cxd4 5 cxd4 d5:
He can't play the best version
oLb6/... i.b7 - which arises in 2 c3
ttJf6 3 e5 ttJd5 4 d4 cxd4 5 cxd4 e6
6 ttJf3 b6 - because the QN is in the
way (3 ...ttJf6 4 e5 ttJd5 5 d4 cxd4
6 cxd4 b6 7 i.c4!).
And now 6 exd5 allows
6... ttJxd5!. What's happened is
Black slipped into a harmless SemiTarrasch Defense position which
Also 3... d5 4 exd5 'iixd5 leads to
more commonly occurs in orders
a 2 c3 d5 position in which Black's
like 1 ttJf3 ttJf6 2 c4 e6 3 d4 c5 4 e3
QN is developed early. White soon
cxd4 5 exd4 d5 6 cxd5?! ttJxd5.
threatens to push his center pawns,
White does better with 6 e5 even
e.g. 5 d4 i.g4 6 i.e2 e6?! 7 h3
i.h5 8 c4 'i'd6 9 g4 i.g6 10 d5! though 6...ttJbc6 and 7...ttJf5 is
quite a respectable version of the
with advantage.
Advance French for Black. It's
In practice, the delayed form of respectable - but perhaps not the
the Alapin appears infrequently kind of middle game that a
after 2 ttJO d6 because of 3 c3 ttJf6! Sicilianista would like to defend.
(4 d4? ttJxe4). It occurs most often
That's why you'll see 1 e4 c5
after 2 ttJf3 a6, when White tries to
2 ttJO e6 3 c3 d5 4 e5 being played,
prove that ... a6 is irrelevant.
even though 4 exd5 is objectively
It's also a good weapon against better. White is hoping a 1...c5
2... e6 but there is a counter-finesse player won't be comfortable in a
in 3 c3 ttJe7!? and ... d5. This French. And that's also why Black
favorite of Yevgeny Vasiukov's often avoids the objectively best
avoids ... d5/exd5 lines in which 4... ttJc6 5 d4 and prefers a gambit
Black has to retake with his queen with 4 ... d4!? 5 cxd4 cxd4 6 i.b5+
or e-pawn, that is 1 e4 c5 2 c3 d5 ttJc6 7 i.xc6+ bxc6 8 'ifa4 or
3 exd5 or 2... e6 3 d4 d5 4 exd5.
6 ... i.d7 7 ttJxd4!?
61
Sicilian Defense
MAIN ALAPIN LINES
unclear, especially after trendy
continuations such as 7 .te2 'fIe7
After 2 e3 tLlf6 play usually
continues 3 e5 tLld5 4 d4. But 8 'iVe2 g5!?
Sveshnikov, the leading authority
The main line of 2.•• d5 is 3 exd5
on 2 c3, prefers 4 tLlf3 and meets 'iVxd5 4 d4. Black should avoid
4 ... tLlc6 with 5 .i.c4 and then 5... e6 4... cxd4 because he's headed to one
6 0-0 d6 7 d4 cxd4 8 cxd4 to reach of the isolated pawn middlegames
a typical Alapin position.
we know to be inferior when they
His point is that this order averts arise in the Queen's Gambit
some troublesome lines such as Accepted and Semi-Tarrasch. For
example, 4 ... cxd4 5 cxd4 tLlc6
4 d4 cxd4 5 cxd4 d6 6 tLlf3 tLlc6
6
tLlf3 e6 7 tLlc3 'fId8 8 i.c4
7 .Jtc4 dxe5 and 7... tLlb6 8 .i.b3 d5.
transposes to a 1 d4 position with an
Another virtue of 4 tLlf3 is that it extra tempo for White.
gives Black a chance for 4..• e6
Better is 4... tLlf6 5 tLlf3. Both
5 ..te4 b6. He thinks he can
transpose into the 4 d4 cxd4 5 cxd4 5... e6 and 5... .i.g4 are common. The
e6 6 tLlf3 b6 line, which enjoys a sly alternative is 5•..tLle6!?
good reputation.
This rules out popular attacking
setups with .i.d3 because of 6 .id3
.i.g4! (7 .i.e3 cxd4 8 cxd4 i.xf3
9 gxf3 e6).
But White can cross him up with
6 .i.xd5! exd5 7 d3 and claim a
slight edge. The more accurate
route to the fianchetto line is 4 tLlf3
b6! 5 ..tc4 ..ib7.
Another benefit of 5... tLlc6 is that
6 .i.e2, which prepares c3-c4, can
be handled by 6.•. exd4! 7 exd4 e6
with rough equality. White cannot
take with his knight on d4 because
g2 is hanging.
The drawback to 4 tLlf3 tLle6
5 .Jte4 is Gedeon Barcza's idea,
5... tLlb6! 6 ..ib3 e4. That's not
necessarily bad for White. It's just
62
Sicilian Defense
In contrast, 6... e6 7 0-0 cxd4
8lbxd4! lbxd4 9 cxd4 followed by
lbc3 or i.f3 favors White.
8 i.g5! and 0-0-0.
However, 7 ... h6!? leaves him
without a plan. A cautionary tale,
Lopez-Dobrov, Neiva 2005, went
ROSSOLIMOIMOSCOW!'iIVxd4 8 i.e3 e5! 9 'ilVd3lbf6 10 0-0-0 i.e7
The two variations in which 11 h3? 'ilVd7 12 ~bl b5 13 lbd2?
White places his bishop on b5 at 'ilVb7 14 f4 0-0 15 fxe5 dxe5
move three are naturally related. If 16 lbf3? b4 17 lbe2 i.b5 White
Black is versatile enough to play resigns.
As a result White has been
open Sicilians with both 1 e4 c5
2lbn d6 and 2 ...lbc6 he can reduce anticipating 7 lbc3 h6 with the
his investment in midnight oil by immediate 7 i.g5!.
means of 2.•. d6 3 i.b5+ lbc6! and
2...lbc63 i.b5 d6!, transposing.
This almost always transposes
(7 ... lbf6 8 lbc3) and so far no one
has found a drawback to 7 i.g5!.
Another virtue is that 4 d4 cxd4
5 'ifxd4 i.d7 6 i.xc6 i.xc6 is a
main line that can also occur after
2... d6 3 d4 cxd4 4 'ifxd4 and then
4... lbc6 5 i.b5 i.d7 6 i.xc6 i.xc6.
The easiest way to deal with 1 e4
c5 2lbn d6 3 d4 cxd4 4 'ilVxd4 with
the prospect of c2-c4 is simply to
avoid it by means of 3 ...lbf6!
(4lbc3 cxd4!).
If Black spends his home time on
that position, he can be prepared for
all the major i.b5(+) and 'ilVxd4
variations.
Top GMs such as Vishy Anand
have tried to find an edge for White
in 4 dxc5 but failed after 4...lbxe4
5 cxd6 lbc6! 6 lbbd2 lbxd6.
Note that in the 6 i.xc6 i.xc6
position White's only bid for
advantage or even for the initiative
is supposed to be 7 lbc3 lbf6
There's a more direct, but little
known, way for White to seek the
63
Sicilian Defense
Maroczy Bind, 1 e4 cS 2 tiJf3 d6
3 d4 cxd4 4 tiJxd4 tiJf6 S .td3!?
Objectively this isn't dangerous.
Subjectively, it can be a major pain
for Black because the reason he
plays 4 ... tiJf6 is to prompt 5 tiJc3
and avoid c2-c4. Today we know
that the pawn structures after
5 iLd3 e6 6 c4 or 5... g6 6 c4 aren't
necessarily bad for Black. But they
can be very difficult to handle if
you're unfamiliar with them - as a
Modern Dragon, Scheveningen or
Najdorfplayer usually is.
For example, 8 ....tg7 9 .tb2 0-0
10 0-0 "ikc7 11 tiJc3 (FogarasiStarostits, Scanno 2005). So far a
simple antidote to 5 .td3!? hasn't
appeared.
MODERN DRAGON
VARIATION
The Modern Dragon was born as
a counter-finesse. When the Dragon
began scoring victories in the 19th
century it commonly came about
via 1 e4 c5 2 tiJf3 tiJc6 3 d4 cxd4
4 tiJxd4 tiJf6 5 tiJc3 d6 6 .te2 g6.
But White found an effective
deterrent in 6 .tg5 and then 6... g6?!
7.txf6!.
Black has two forcing replies to
. 5 .td3. One is 5... e5 when 6 tiJf3
Kurt Richter, who helped launch
and 7 c4 reaches a good version of
6 .tg5, also popularized the refined
the Kalashnikov Variation for
order of 1 e4 cS 2 tiJf3 d6 3 d4 cxd4
White.
4 tiJxd4 tiJf6 S tiJc3 g6!. Then
Perhaps better is S...tiJc6 and 6 .tg5 .tg7 is too late.
then 6 tiJxc6 bxc6 and 7... g6.
Unlike the 6 .td3 Najdorf we'll
consider later, Black hasn't spent a
tempo on ... a6. But experience
indicates White stands well after
7 c4 g6 8 b3:
The price Black pays IS
committing himself to a Dragon
before .te2. This gives White more
aggressive options, including the
Yugoslav Attack - .te3/fl-f3J'iid2/
0-0-0 - and some quiet ones like the
64
Sicilian Defense
fianchetto system, 6 g3. Even the
latter contains pitfalls after 6 g3
.tg7 7.tg2:
More common is 6 .te3. It sets a
small trap (6 ... lDg4?? 7 .tb5+) and
typically leads (6 ....tg7 7 0) to
another fork in the road. One path is
7... 0-0 8 'iVd2 lDc6 and another is
7... lDc6 8 'iVd2 0-0. Most players
see no difference.
But there is one because 7...lDc6
8 'iid2 allows Black to use
8....td7!? as a waiting move.
The natural 7...lDc6 walks into
8 lDxc6! bxc6 9 e5, a trap which has
snared many a master. Black should
either play 6... lDc6 (7 .tg2 lDxd4!)
before ... .tg7 - or he should delay
... lDc6 in favor of 6... .tg7 7 .tg2
0-0.
Then on 9 .tc4 he can transpose,
with 9... 0-0, into main lines. That's
useful during times, like these,
when Black is faring better against
.tc4 Yugoslavs than against non.tc4 ones. Thanks to this order he
can temporarily keep his king in the
center after 9 0-0-0 ~c8. Then
10 g4 lDe5 11 h4 h5! 12 g5 lDh7
In the Yugoslav Attack there is a makes it safer for Black to castle,
'something to think about' move, now that the g- or h-file can't be
6 D. Yefim Geller indicated in the forced open.
1975 ECO that '6 f3?!' was
Another situation arises when
punished by 6... 'i'b6 7 .te3 'iixb2.
White
plays .tc4 but delays 'iid2,
But that allows 8 lDdb5! with
as in 7..•lDc6 8 .tc4 in order to
threats to trap the queen with
accelerate the h2-h4-h5 attack.
9 ~b 1. This is strong enough to
This order tries to tempt Black
make 6... 'i'b6 the move deserving
into 8... 'i'b6. That looks good in
ofa '?!'.
This trap is important to
recognize because it arises from
other move orders, including 1 e4
c5 2 lDf3 lDc6 3 d4 cxd4 4 lDxd4
lDf6 5 lDc3 d6 6 g3 g6 when 7 .tg2
.tg7 8 lDxc6! transposes into what
White wants and 7... lDxd4! 8 'iixd4
.tg7 is what Black wants.
65
Sicilian Defense
view of 9 ii.b3? liJxe4 or 9... liJg4.
However 9 liJf5! 'ifxb2 10 liJxg7+
is such a promising sack that
masters abandoned 8... 'ifb6?! in the
1960s.
Black's best policy after 8 ii.c4 is
8... 0-0!, which usually transposes to
a main line after 9 'ifd2. If instead
9 ii.b3 Black can get his queenside
attack running before White inflicts
kings ide damage, 9...ii.d7 10 h4
liJxd4 11 ii.xd4 bS 12 hS as.
For instance 13 hxg6 hxg6
14 liJdS liJxdS 1S ii.xdS l::te8
16 'ifd2 ':xe2! 17 'ifxe2 ii.xd4
18 0-0-0 ii.g7 19 'itb1 'ib6 20 g4
e6 as in Perez Candel-Guseinov,
Kusadasi 2006.
Then 10 exd5 liJxd5 11 liJxc6
bxc6 12 ii.d4 is heavily analyzed
and remains controversial and
12 liJxd5 cxd5 13 'ifxd5 is risky.
White should avoid the alternative
10 liJxe6 bxe6 11 exdS.
This transposes to the pawn-grab
line after 11... liJxd5 12 liJxd5 cxd5
13 'ifxd5. But there's no reason to
grant a dangerous extra option,
Vasily Panov's l1 ...'ifaS 12 dxe6
ii.e6!. A 1954 Soviet game went
13 a3 l::tfd8 14 'ife2 l::txd1+
15 'ifxd1 l::td8 16 'i'e2 liJd5
17 liJxd5 'ifxd5 18 c7? ii.xb2+!
19 'itxb2 'i'a2+ and wins.
If Black doesn't play 9... d5, he
has two natural orders to reach a
tabia. One is 9...liJxd4 10 ii.xd4
ii.e6 11 'itb1! 'i'e7 (not 11...'i'a5
12 liJd5! 'i'xd2 13 liJxe7+).
Experience says White is better but
not enough to discourage Black
from seeking it.
DRAGON: 90-0-0
In the Yugoslav tabia (1 e4 eS
2 liJt3 d6 3 d4 exd4 4 liJxd4 liJf6
S liJe3 g6 6 ii.e3 ii.g7 7 t3 liJe6
8 'ifd2 0-0) White has three chief
alternatives. One of them, 9 0-0-0,
The other order is 9...ii.e6
allows only a little room for finesse followed by 1O ... liJxd4. Black may
be skittish about it because ...
after the forcing 9...dS.
66
Sicilian Defense
.. .instead of 10 'it'b 1 ttJxd4
11 i£.xd4 "ikc7, transposing, White
can try 10 ttJxe6 fxe6 11 g3! and
12 i£.h3, which puts e6 under fire.
But there's an upside to 9... i.e6.
Black steers clear of 9 ... ttJxd4
10 i.xd4 i..e6 11 ttJd5!?, which
many Dragonistas don't like to
face. So it's up to him to choose
which line he wants to avoid more,
the 11 ttJd5!? one or the 9... i.e6
10 ttJxe6 one.
Not at all. A natural continuation
is 12 ..."ika5 13 i.a4 l::f.fc8 14 i.b3.
Then we've landed in a position
that was reached in the 1950s via
9 i.c4 ttJxd4 10 i.xd4 i.e6 11 i.b3
'iia5 12 'it'bl l::f.fc8. It's been
analyzed well into a king-and-pawn
endgame (!) with the conclusion
that White is much better, if not
winning .
This underlines a danger to both
players in a Modem Dragon - the
possibility of a 9 0-0-0 line
suddenly turning into a risky 9 g4
position and vice versa. Or a
9 0-0-0 line transforming into a
9 i.c4 position.
That's the case with 9 0-0-0
ttJxd4 10 i.xd4 "iia5 and 11...i.e6.
This looks like just another routine
Dragon. But Black's order is faulty
because 11 i.c4! will transpose,
after 11 •••i.e6 12 i.b3, into that bad
1950s line.
Whatever the route to that tabia
White usually continues 12 g4 or
12 h4. A clever transposer may
prefer 12 i.b5!? The bishop is
headed for b3, its best square on the
board.
DRAGON: 9g4
The 9 g4 thrust was long been
considered a worthy alternative to
9 0-0-0. It's superior in some ways:
It stops 9 ... d5 with 10 g5!. It
discourages 9 ... ttJxd4 10 i.xd4
i.e6 because 11 ttJd5! is a better
version of 9 0-0-0 ttJxd4 10 i.xd4
i.e6 11 ttJd5.
But doesn't that just lose time?
But 9 g4 was dealt a serious blow
m the 1995 world championship
67
Sicilian Defense
match. What Team Kasparov came
up with was 9.•. il..e6! 10 0-0-0
tbxd4 11 il..xd4 'ilVa5!.
DRAGON: 9 il..c4
Some Dragon experts, like
Sergey Tiviakov, have circumvented this hyper-theoretical line by
meeting 6 il..e3 il..g7 7 f3 tbc6
S'ilVd2 with S..•il..e6!?:
We are headed toward a standard
position in the 9 0-0-0 tbxd4 line but with an extra tempo for Black.
He saved it by getting his queen to
a5 in one move, compared with
9 0-0-0 tbxd4 10 il..xd4 il..e6
11 ~bl! when l1..:iWc7 is played
because 11 ... Via5? stumbles into
12 tbd5!.
Black is willing to transpose into
non-il..c4 lines after 9 0-0-0 0-0 or
9 g4 0-0. The obvious drawback is
9 tbxe6 fxe6. But so far the pawn
weakness hasn't proven significant
in games that went 10 0-0-0 .a.c8
11 il..c4 'ilVd7 or 10 il..c4 ct;f7 11 a3
(else ... tba5) .a.c8 12 il..a2 tbe5 and
...tbc4.
An extra tempo is worth
platinum in the Dragon and this was
apparent in the first AnandKasparov game. After 12 <itbl.a.fcS
Black was already threatening
... l:.xc3/ ... tvxa2+. His attack would
be faster after 13 a3 ':'abS 14 h4 b5
15 h5?! b4!.
Black can try to improve on this
with S•..tbxd4 9 il..xd4 il..e6, when
White is denied both tbxe6 and
il..c4. But this fails to 10 il..b5+!.
Then 1O... il..d7? 11 il..c4 gets Black
into the position he was trying
to avoid - and a move down to
boot. There's been virtually no
experience with 1O... ct;f8 (11 0-0-0
'ilVa5) so we'll have to let the
tricksters test it for us.
There doesn't seem to be a
downside to 9 g4 il..e6! as there is in
the comparable 9 0-0-0 il..e6
position because 10 tbxe6 fxe6
11 g3 is, of course, illegal.
68
Sicilian Defense
In the main .tc4 line, 7...lDc6
8 'i'd2 0-0 9 i.c4, Black usually
plays ....td7 and posts a rook at c8
while White plays 0-0-0, .tb3 and
h2-h4. If Black is going to play a
... 'i'a5/ ...:fc8 system instead it's
more accurate to do it with 9...'i'a5
and 1O ... .td7 rather than the
reverse.
Now 9 f4 allows 9•• :Wb6!, which
threatens 1O ... ~xe4 or 1O ... ~g4/
11.. ..txd4. No better is 8 f4 'i'b6!.
This
underlines
White's
difficulty in reaching Classical
positions via the Levenfish (6 f4)
Variation. The Levenfish sets a trap,
6 ....tg7 7 e5!?, that is best avoided
by 6 ... ~c6. A devious White might
play 6 f4 to see if Black falls into it,
and then shift into the Classical line
after 6 ... ~c6 7 .te2 .tg7 8 .te3
0-0.
The reason is that this order halts
the quick attack of 10 h4? with
10 ••• 'i'b4! 11 i.b3 ~xd4 and
12 .txd4? ~xe4. White has to play
the 12 'i'xd4 ending and, while it's
not dreadful, it's certainly not what
he had in mind.
But he's outsmarted himself
because 9 O-O?! walks into 9•••'i'b6!
again.
DRAGON: CLASSICAL
When White plays .te2 he
usually continues fl-f4 and .te3.
An early tabia is 1 e4 c5 2 ~f3 d6
3 d4 cxd4 4 ~xd4 ~f6 5 ~c3 g6
6 .te2 .tg7.
There are few .te2 lines in which
White castles long, so 7 0-0 would
seem to best retain his options.
However, the g I-b6 diagonal is a
problem after 7... 0-0 8 .te3 ~c6:
69
The way to avoid that kind of
surprise as well as other liberating
ideas such as ... d5, ... ~xd4 and
... ~g4 is to retreat the knight
from d4 to b3. Siegbert Tarrasch
routinely gave ~b3 a question mark
in the Sicilian. But a case can be
made in the Modem Dragon for
6 .te2 .tg7 7 ~b3 rather than
7 0-0.
The reason is the natural 7... 0-0
allows White a sharp extra option,
Sicilian Defense
8 g4!?
. White may be gun-shy about 8 f4
because of 8.. :ii'b6. But this time
White has a strong 9 e5!, which
Sammy Reshevsky once allowed
(9 ... dxe5?? 10 tbxc6 wins).
Black's king posItion makes
8...tbc69 g5 tbd7 10 h4 or 10 ..te3
more dangerous than if he had
played 7 ...tbc6.
Some players don't like tbb3
because it allows Black to develop
... ..te6 without fear of tbxe6. But
the bishop faces more jeopardy
from pawns - 7 0-0 tbc6 8 tbb3
..te6? favors White after 9 f4! and
10 f5!, e.g. 9 .. J:tc8 10 f5 ..td7
11 g4.
After 8... 0-0 instead, White
reaches the Foltys Attack with
9 tbb3 and then 9... ..te6 10 g4. As
for Black, his way of dodging it
is 9... tba5. Then 10 0-0 ..te6!
transposes into a position normally
reached via 9... ..te6 10 0-0 tba5 that
is considered safe enough.
In truth, tbb3 serves too many
good purposes and the main
This order allows Black to meet
question is when to play it. It is
essential to the most double-edged 10 g4 with 10... b6!? There is no
Classical line, Jan Foltys' attack. bishop to kick with f4-f5, and
Perhaps the most accurate order for following 11 g5 tbd7 and ... e5 he
White is 6 ..te2 ..tg7 7 ..te3 tbc6 has counterplay, e.g. 12 'ild2 ..tb7
and then 8 f4 0-0 9 tbb3 ..te6 13 ..td4 e5 14 fxe5 tLJxe5 15 tLJxa5
10 g4!? (see next diagram)
bxa5 16 0-0-0 'ilc7 as in Oll-Van
der
Wiel, Dutch Team ChampionThis order avoids 8 tbb3 ..te6
9 f4 'iYc8!, which stops g2-g4. Then ship 1996.
White has nothing better than to
White, too, gets an extra option
transpose into the older lines with
10 0-0 0-0 that work well for Black from this order and it is 10 e5.
70
Sicilian Defense
Black can transpose into a
Modem Dragon with 6... g6 and
then 7 i.e3 or 7 ltJb3 is normal. But
there is also 6...ltJxd4!? 7 'it'xd4 g6.
Then 1O... dxe5?? loses a piece
(11 'it'xd8 and ltJxa5). But 10...ltJe8
needs testing to see if White is
overextended.
The most popular alternative to
Then on 8 i.e3 i.g7 White has
9... ltJa5 is 9... a5. Then on the semi- no edge because his queen is
automatic 10 a4 Black can safely vulnerable (9 'ifd2 0-0 10 0-0-0
play 10... i.e6 since 11 g4 d5! is i.e6). This would come about from
much stronger when Black's QN the Modem Dragon's 5... g6 6 i.e2
cannot be driven offb4 (12 f5 i.c8! ltJc6 7 i.e3 and 7 ... ltJxd4 8 'it'xd4?
13 exd5ltJb4 14 i.D gxf5).
If White is going to punish this
White usually plays 11 0-0 and
order the best bet is 8 i.g5 i.g7 9
transposes into a position normally
ltJd5. The critical position is likely
reached via 9... i.e6 10 0-0 but with
9... 0-0 10 i.xf6 gxf6. But Black's
the addition of the a-pawn moves.
bishops seem to offset the bad
There are plusses and minuses but
pawns, e.g. 11 0-0-0 f5 12 'ilfd3
the addition tends to help White in
fxe4 13 'it'xe4 :e8 14 'it'D :e5
lines such as 11 ...'it'c8 12 h3 ltJb4
15 i.c4 i.e6 as in Michiels-Van der
13 ltJd4 i.c4 14ltJdb5.
Weide, Amsterdam 2000.
Finally, there's a move order that
If that's true, 6 ... ltJxd4!? IS a
doesn't fit into any traditional
Dragon category. It arises from 1 e4 significant finesse - and a reward
c5 2ltJf3 d6 3 d4 cxd4 4ltJxd4ltJf6 by the Gods of Theory to Black for
5 ltJc3 ltJc6 when White responds reviving the old Dragon move order
after all these years.
with the vanilla move 6 i.e2.
71
Sicilian Defense
ACCELERATED DRAGONS
lines with ~hllf2-f4 in which .i.e3
The Accelerated Dragon, 1 e4 cS
2 liJf3 liJc6 3 d4 cxd4 4 liJxd4 g6,
becomes a Modem Dragon after
... d6. The benefits of the
Accelerated order lie in limiting
White options and enabling Black
to shoot for ... d5!.
is delayed.
Some books claim 6 liJb3 is
refuted by 6 ... .i.xc3+!? (7 bxc3liJf6
8 .i.d3 d5). But many GMs don't
believe it: Anatoly Karpov has
played 6 liJb3 anyway and Gata
Kamsky and Bent Larsen passed up
6 ... .i.xc3+ when they had the
chance as Black.
White's knight is immediately
attacked after S liJc3 .i.g7. If he
wants to fianchetto, he has to try
something like 6 liJde2liJf6 7 g3.
Nevertheless if White wants to
play .i.e2 but not .i.e3, a more
precise route is 5 .i.e2 or 5 liJb3
(after 1 e4 c5 2 liJf3 liJc6 3 d4 cxd4
4 liJxd6 g6). For example, 5 .i.e2
.i.g7 6 liJb3 and 6 ... d6 7 0-0 liJf6
8liJc3.
If you look up SliJc3 .i.g7 6 .i.e3
in books you will only find 6...liJf6.
The reason is this is how Black
sidesteps some Yugoslav lines. For
example, 7 f3?! 0-08 'iid2 and now
not 8 ... d6 but 8 ... d5! with excellent
chances. The same goes for 8 g4
'iib6!.
tl>
Now 7 ... 0-0 8 ... g2 d6 9 0-0
gets him to a respectable Modem
Dragon (1 e4 c5 2liJf3 d6 3 d4 cxd4
4 liJxd4 liJf6 5 liJc3 g6 6 g3 .i.g7
7 .i.g2 0-0 8 0-0 liJc6 9liJde2).
But Black improves with 7 ...bS!
(8 liJxb5 liJxe4). In contrast he
usually has to spend a tempo on
... l:.b8 or ... a6 to engineer ... b5 in
the Modem Dragon.
More often you'll see 6 liJb3.
White is seeking a Classical Dragon
(6 ... liJf6 7 .i.e2 0-0) in which he
retains the .i.g5 option as well as
But there is a rarely mentioned
downside to 6 ... liJf6. It grants
72
Sicilian Defense
White an extra option, 7 ltJxc6!? SCHEVENINGEN VARIATION
bxc6 8 e5, which leads to a space
The basic elements of the
edge following 8 ... ltJg8 or into
Scheveningen are Black pawns at
razor-sharp tactics (8 ...ltJdS) that
e6 and d6 and a ltJ on f6 facing
have been analyzed past move 20.
White knights at c3 and d4 and a
Do you really want to subject
pawn at e4. This can come about
yourself to memorizing lines such
from a variety of orders but many of
as 8... ltJdS 9 ltJxdS cxdS 10 'fixdS
them prematurely commit Black to
l:.b8 11 ii.xa7 ltxb2 12 ii.d4 ltxc2
moves like ... 'fic7.
13 ii.d3 e6! 14 ~a8 ltc6 IS ii.bS
lta6 ?
Why is it premature? After all,
Black is always playing ... ~c7 in
If the answer is no, the mental
the Sicilian, so 'it just transposes,'
hygiene move 6... d6 is preferable.
You may end up in the Yugoslav right?
this way but it may help you to
sleep at night.
No, not when White develops
ii.g2, for example. One of the
best features of the modem
Scheveningen order, 1 e4 c5 2 ltJf3
d6 3 d4 cxd4 4 ltJxd4 ltJf6 5 ltJc3
e6, is that after 6 g3:
Because of the ways that the
Accelerated move order degrades
White's options it would be a
perfect defense to 1 e4 - if it
weren't for the Maroczy Bind. To
avoid it Black has tried orders such
as 1 e4 c5 2ltJf3 g6.
Then if White tries to transpose
into the Bind with 3 c4 ii.g7 4 d4 or
3 d4 ii.g7 4 c4 (and then 4 ... cxd4
SltJxd4ltJc6) Black has a variety of
independent options, including
4 ... ~aS+ or 4 ... d6.
Those extra options have to be
weighed against the downside. To
play 1 e4 cS 2 ltJf3 g6 3 d4 ii.g7
Black has to feel comfortable in
other openings, including a Schmid
Benoni (4 dS) and a Panov CaroKann after (4 c3 cxd4 S cxd4 dS!
6 exdS!) as well as 4 dxcS!?
Black can dispense with ... 'iic7
and obtain equal play from 6...ltJc6
7 ii.g2 ii.d7!, e.g. 8 0-0 a6 9 a4 ii.e7
10 ii.e3 0-0 11 ~hl ltJeS 12 f4
ltJc4, Malakhov-Vogt, Feugen
2006. No better is 8 ltJdbS ~8
9 ii.f4 ltJeS and ... a6.
73
Sicilian Defense
The Scheveningen order is also
effective against White players who
try to escape to a more exciting
Sicilian with 6 ~c4 or 6 ~g5. They
expect 6 ... tiJc6 or 6 ... a6, reaching
familiar territory.
5 ... a6 6 ~c4 e6 7 ~e3?! ~e7. For
example, 8 'iWe2 b5 9 ~b3 b4 or
9 ~d3 .tb7 10 a3 tiJbd7 11 0-0
tiJc5 12 f3 0-0 and Black stood
splendidly
in
Ulko-Ulybin,
Moscow 2005.
But Black has a better move,
6 ... ~e7!. Then both sides risk being
trapped in an inferior Najdorf or
other Sicilian.
KERES AND ENGLISH
ATTACKS
Two major alternatives to quiet
Scheveningens involve g2-g4.
There is the Keres Attack, with an
immediate 6 g4, and the English
Attack, with ~e3, f2-f3 and 'iWd2
added.
For example, 6 ~c4 ~e7! 7 ~b3
and 7 ... a6? 8 f4 favors White, e.g.
8 ... b5 9 e5! with advantage or
8... 0-0 9 'iWf3. (In the Najdorf move
order this happens with 5 ... a6
6 ~c4 e6 7 .tb3 when Black plays
7 ... ~e7?! instead of7 ... b5!.)
For decades it was thought the
Keres Attack could only come
about from the traditional order.
There's no point to g2-g4-g5 if
Black hasn't put a knight on f6, the
thinking went. As a result a
1967 Shakhmatny Bulletin article
recommended 1 e4 c5 2 tiJf3 d6
3 d4 cxd4 4 tiJxd4 tiJc6 5 tiJc3 e6.
Black can get better counterplay
from ... tiJa6-c5 rather than ... a6, e.g.
7... 0-0 8 ~e3 tiJa6! 9 f4 tiJc5.
If White is alert he may try
7 ~e3 instead of 7 ~b3. He
wouldn't mind reaching a book
Sozin Variation (7 ... tiJc6). But
7... a6! is much better.
Then 6 ~e2 tiJf6 gets Black into
This is a superior Najdorf the Scheveningen while avoiding
position that could come about via the Keres.
74
Sicilian Defense
But White doesn't need to with a2-a3 and has already played
attack a knight. In the 1985 world g4-g5. That's enough to ensure an
championship match Karpov advantage, 8... tiJb6 9 :gl tiJ8d7
improved with the simple, but at the 11 f4 and 12 f5.
time, stunning 6 g4!. Then 6 ... ttJf6 7
Another example, is 6.....te7 7 g5
g5 is an excellent Keres. Avoiding
tiJfd7
8 h4 tiJc6 9 ..te3 when
... tiJf6 is better but not necessarily
10 ..tc4! gives White a favorable
equal (6 ... a6 7 ..te3 tiJge7 8 tiJb3).
Velimirovic Attack.
The sad truth is there is no
From the Sozin order (1 e4 c5
simple anti-Keres route to the
Scheveningen. Kasparov even 2 ttJf3 ttJc6 3 d4 cxd4 4 tiJxd4 ttJf6
resorted to the Najdorf (5 ... a6 6 5 tiJc3 d6 6 ..tc4 e6 7 ..te3) he
..te2 e6) to do it, although that could obtain something like that
allows White a wealth of sixth only if Black was very cooperative
- 7... a6 8 'i'e2 ..te7 9 g4? 'i'c7?
move alternatives.
(rather than 9... tiJxd4 10 ..txd4 e5!
In the normal Keres order, 1 e4 and ... ..txg4) 10 g5 tiJd7 11 h4.
c5 2 tiJf3 d6 3 d4 cxd4 4 ttJxd4 tiJf6
Another dangerous attacking
5 tiJc3 e6 6 g4, a consensus has
emerged that 6... h6 is the only move setup is begun by 1 e4 c5 2 tiJf3 d6
that offers Black a shot at equality. 3 d4 cxd4 4 ttJxd4 tiJf6 5 ttJc3 e6
Other moves allow White to reach 6 ..te3 or 6 f4. In both cases White
favorable versions of the English can carry out a Mikhail Tal plan of
Attack, the Velimirovic Attack or ..te3, £2-f4, 'i'D and 0-0-0.
somesuch. For example, 6... a6 7 g5
The pawn move seemed superior
tiJfd7 8 ..te3 b5 9 a3:
because 6... ..te7 7 ..tb5+ and 8 e5
poses problems for Black. The
bishop move, on the other hand,
allows White an extra option, after
6... tiJc6, of transposing to a Sozin
(7 ..tc4), without having to face Pal
Benko's 6 ... 'i'b6 in the normal
Sozin order.
But today 6 ..te3 is preferred for
another reason. It is the main route
into the English Attack.
This is an English Attack in
which White has replaced £2-D
After 6... a6:
75
Sicilian Defense
10 g4! dxe4 11 'iVf2 'iVc7 12 g5 with
advantage.
CLASSICAL VARIATION
If White adopts the f2-f4 plan
Black obtains quick counterplay
(7 f4 b5! 8 'iVf3 .tb7 9 .td3 ttJbd7
10 a3 ':c8).
But 7 g4 is a good response.
Then the natural 7... h6 transposes
into a kind of Keres in which White
has added the useful move .te3
while Black has added the slow
... a6. As a result White has good
chances whether he continues a la
Keres (8 h4), in English Attack
style (8 f3), or in a TaVTopalov
hybrid with 8 f4.
There is a drawback to 7 g4
inasmuch as 7••. e5 8 ttJf5 g6 forces
White to sack material (9 ttJg3?
.txg4). But 9 g5! gxf5 10 exf5, the
Bela Perenyi line, has been
prohibitively strong.
The wonderfully flexible system
that comes about via 1 e4 c5 2 ttJf3
d6 3 d4 cxd4 4 ttJxd4 ttJf6 5 ttJc3
ttJc6 should be named after Louis
and Wilfried Paulsen, who explored
it in the 1880s. But there already is
a Paulsen Variation so 'Classical'
may have to do.
Black has a choice of two orders,
2 ... d6 and 5 ... ttJc6 versus 2... ttJc6
and 5 ... d6. Today his decision
usually hinges on how he feels
about 2 ... ttJc6 3 .tb5 as opposed to
2 ... d6 3 .tb5+, as well as how he
would deal with 3 c3 or 3 ttJc3 .
The Paulsens used the Classical
to reach the Dragon, Scheveningen
and other lines that came to
prominence long after they were
gone, including Isaac Boleslavsky's
6 .te2 e5. Today it's also used as a
waiting policy. After 6 .te2, for
example, Black can play a
Scheveningen or a Dragon more
safely (or 6 ... ttJxd4!? 7 'iVxd4 g6 as
noted earlier).
But there is no absolutely correct
In the majority of games White
English move order after 6 .te3 a6,
since 7 f3 and 8 'iVd2 or 7 'iVd2 and plays 6 .tg5 and 6 .tc4 and so
8 f3 also have merit. In the 1990s Black spends his home preparation
White showed that his better mainly on the Rauzer and Sozin
development trumped ... d5 even Variations. This makes him more
when g2-g4 is delayed, as in 7 'iVd2 vulnerable to transpositional traps,
.te7 8 f3 ttJc6 9 0-0-0 d5 and now beginning with 6 f4.
76
Sicilian Defense
This move can be a problem for a
Black who uses the Classical to
dodge the Keres Attack and reach
the Scheveningen (6 i.e2 e6).
l:.e8 14 i.b3 a6 15 llJd4 llJa5
16 llJf5 llJxb3 17 axb3 llJxe4
18 llJh6+ gxh6 19 l:txe4 i.xe4
20 llJxe4 i.f8? 21 i.g5 with mixed
chances.
The safest response to 6 f4 may
be 6 ... g6, transposing into the
Levenfish Dragon. But while a
Dragonista will have memorized
those complex 7 llJxc6 bxc6 8 e5
and 7 i.b5 i.d7 8 i.xc6 lines, it's
unlikely a Classical player has.
That's another point in favor of
6 f4!?
If he replies 6... e6 White has an
ambitious plan in 7 i.e3 and 'if01
0-0-0. This transposes into 1 e4 c5
2 llJf3 d6 3 d4 cxd4 4 llJxd4 llJf6
5 llJc3 e6 6 f4 llJc6 7 i.e3 when
Black has lost the chance for the
sharper 6 ... a6 and ... b5/ ... i.b7.
Other tricky Classical options are
6 i.e3 and 6 0, which are both
aggressive and waiting moves.
They can transpose into a Dragon
(6 ... g6) or an English Attack (6 ... e6)
- but if Black really wanted to play
a Dragon he would have chosen
5 ... g6 and if he wanted an English
he would have played 5... e6.
A Black who prefers to meet
6 i.e2 with the Boles1avsky
Variation may also have a problem
with 6 f4. Books recommended
6 ... e5 because 7 llJxc6 bxc6 8 fxe5
llJg4! gives him excellent darksquare play whether White accepts
the pawn (9 exd6 i.xd6) or permits
9 ... llJxe5.
He loses some options this way.
Theory prefers ... a6 to ... llJc6 in the
English but after 6 i.e3 e6 Black is
already committed to ... llJc6. Also,
6 ... e6 7 i.c4! will be annoying to a
Black who usually meets 6 i.c4
with Benko's 6 ... 'iib6.
There are two clouds on 6 i.e3 's
horizon. One is 6 ... e5, which
transposes into a line of the
Boleslavsky that is considered even
(7 llJb3 i.e7 8 i.e2) or into a
Sveshnikov Variation that's rated
good for Black (7 llJdb5 a6 8 llJa3
b5).
However, 6 ... e5 is tested by
7 llJO! followed by i.c4. An
example of a Boleslavsky player
finding himself in unfamiliar
surroundings is Salmensuu-Atalik,
Groningen 1999: 7 ... i.e7 8 i.c4
0-0 9 0-0 exf4 10 i.xf4 i.g4
11 'iid2 i.h5 12 l:.ae1 i.g6 13 ~h1
77
Sicilian Defense
The other cloud is 6 ... lbg4!
which theory says is at least equal.
That's why 6 f3! is more accurate.
It normally occurs via 6 ... e6
7 'it'd2 .te7 8 0-0-0 0-0 9 lbb3
'it'b6. But White can keep it from
happening by playing 9 f4 or 9 f3.
Both moves enjoy good reputations.
White can still reach Yugoslav
Dragon and English Attack
positions (6 ... g6 7 .te3 or 6 ... e6
7 .te3) but he avoids 6 .te3 lbg4.
Experience with 6 ... e5 7 lbb3 .te7
8 .te3 .te6 9 lbd5 and 7 ... .te6
8 lbd5 is mixed.
That's why Black may prefer
6...'it'b6 and then 7 lbb3 e6 8 'it'd2
.te7 9 0-0-0 0-0, transposing. A
third route to the tabia is 6... e6
7 'it'd2 'it'b6 8 lbb3 .te7 9 0-0-0
0-0.
In the Classical order, 6 .tg5
discourages a switch to the Dragon
(6 ... g6 7 .txf6) or the Boleslavsky
Which order is best? The latter
two threaten ... 'it'xd4 as well as
... 'it'xb2. But they share the demerit
of allowing White to double pawns
with .txf6.
(6 ... e5 7 .txf6 'it'xf6? 8 lbd5).
Black usually replies 6 ... e6 and
experience has shown that White's
only way to seek an edge is 7 'it'd2
and 8 0-0-0.
In addition, White can answer
6.•. e6 7 'it'd2 'it'b6 with 8 0-0-0 (and
transpose after 8 ....te7 9 0, for
example). The only way to punish
this is 8 ... lbxd4 9 'it'xd4 'it'xd4
RAUZER VARIATION
Black makes major decisions 10 Ihd4. Unlike other early
about the timing or need for Sicilian endgames, this one favors
... lbxd4, ... h6, ... a6 and ... 'it'b6. If White, according to Peter Svidler.
White meets ... 'it'b6 with lbb3 he
might reach this position, one of the
best Rauzer tabias for Black.
He preferred 6•..'it'b6. But that
may have its own problem,
7.te3!?
78
Sicilian Defense
It's considered double-edged
enough to be attractive to both
White and Black and can come
about from 6 i.g5 e6 7 .d2 i.e7
80-0-00-09 f4liJxd4 10 .xd4.
If Black would rather defend this
position than the one that arises
after 9 liJb3, he can get where he
wants to go with 8•••liJxd4 9 .xd4
0-010 f4.
Declining the pawn is bad
(7 ...•a5? 8 liJb3 .c7 9 liJd5! or
8...•d8? 9 i.e2 e6 10 0-0, a
Scheveningen a tempo or two
down).
The problem with the second
order is White doesn't have to
cooperate by playing 10 f4. He also
The same goes for 7 ... liJg4 has 10 e5, which Jan Timman said
8 liJd5! liJxe3 9 fxe3 in view of is the only way to punish Black's
9...•a5+ 10 b4! liJxb4 11 liJb3!
sequence, e.g. 10... dxe5 11 .xe5
liJd3+ 12 ~e2, winning a piece. In
i.d7 12 i.e2 :c8 13 i.e3 a6 14 g4
contrast, 6... e6 7 .d2 'iWb6 8 i.e3
was unclear in Akopian-Jobava,
liJg4! is fine for Black.
Beer Sheva 2005.
So the answer to which order is
White also has a promising
best may depend on whether Black
can play 6... 'iWb6 7 i.e3 .xb2 alternative in 10 f3. That works
8 liJdb5. If not, 6... e6 7 .d2 i.e7 better here than with knights on the
and 8 0-0-0 0-0 9 liJb3 'iWb6 is board (6 i.g5 e6 7 'iid2 i.e7
better.
8 0-0-0 0-0 and then 9 f3 a6 10 g4
d5!).
Another tabia is:
That's the tradeoff: Black has to
decide which middlegame to shun the one with 8... 0-0 9 liJb3 or the
8... liJxd4 9 .xd4 0-0 10 e5 and
10 f3 ones.
A Black trickster may like the
rare 7 'iid2 i.e7 8 0-0-0 0-0 9 f4
i.d7!?
79
Sicilian Defense
His idea is to answer the natural
10 Jic4 with 1O ... liJxd4 11 'iixd4
Jic6 and reach a promising position
(12 .l:thel 'iia5).
That's a line that commonly
comes about via 9 ... liJxd4
10 'iixd4 'iia5 11 Jic4 Jid7 when
White passes up 12 e5! for the less
ambitious 12 ::thel.
This order is significant because
on 9 f4 Black transposes not into
the bad line (9 ... 0-0?) but into a
very good one, 9... Jid7!. This
comes about via 7 'iid2 a6 8 0-0-0
Jid7 9 f4 Jie7 when White has
bypassed 9 D.
If he insists on 9 f3 Black
responds 9...liJxd4 to 'iixd4 b5!
and posts his bishop more usefully
The apparent refutation of
on b7 than on d7, e.g. 11 h4 Jib7
9 ... Jid7 is to liJdb5, threatening
12 ~bl 'iic7 13 'iid2 .l:tc8 14 Jid3
liJxd6. But to ... d5!? offers good
h6 15 Jie3 b4 16liJa4 d5!, Sax-Acs,
play for a pawn, e.g. 11 exd5 liJxd5
Hungarian Championship 2003.
12 liJxd5 exd5 13 'iixd5 Jig4!
14 'iixd8 Jixd8 15 ::td2 Jib6. Or
And since 9 liJb3 allows 9... b5!
11 e5 liJe8 12 h4 a6 13 liJd4 f6,
under better than usual circumSutovsky-Zviagintsev, Essen 2000.
stances, the real test of Black's
Another deceptive order is move order is 9 Jixf6. Then
7 'iid2 Jie7 8 0-0-0 a6 (or 7... a6 9 ... Jixf6? drops a pawn to
8 0-0-0 Jie7). This bears a 10 liJxc6. But experience with
resemblance to 7 'iid2 Jie7 80-0-0 9... gxf6 10 f4 Jid7 and 10 Jie2
0-0 9 f4 a6? which Paul Keres 'iib6 has been inconclusive so this
refuted with 10 e5! dxe5 11 liJxc6 order may be a real contribution to
the Rauzer.
bxc6 12 fxe5.
80
Sicilian Defense
SOZIN VARIATION
Pal Benko's great contribution to
6 Jtc4 theory was realizing that
6...'iib6 7 lbb3 e6 transposes to a
fairly good Scheveningen for
Black, even though he'll probably
retreat his queen to c7 at the cost of
a tempo. One of the secrets of the
Sozin is that getting into the right
line is often worth a tempo.
Joseph Blackburne used 6 Jtc4
against the Classical Variation so
that 6 ... g6 could be punished by
7 lbxc6 bxc6 8 e5!. Later Os sip
Bernstein demonstrated the finesse
of6... Jtd7.
This is evident when White stays
within Sozin borders by answering
6.. :ii'b6 with 7 lbdb5!? and then
7... a6 8 Jte3 'i'a5 9 lbd4 e6 (not
9 ... lbxe4? 10 'iif3):
Books recommend 7 0-0 because
7 ... g6 8lbxc6 slightly favors White.
But 7... e6!? is an 'old' Sozin with
0-0 that isn't considered as
dangerous as 0-0-0 lines, e.g. 8 Jtb3
Jte7 9 Jte3 0-0 10 'i'e2 lbxd4
11 .ixd4 .ic6. And there's been
virtually no GM experience with
8lbdb5 'iib8 9 Jtf4 or 9 Jtg5.
White is a tempo down in an 'old'
Sozin. (It would occur after
6 Jtc4 e6 7 Jte3 a6 if Black had the
extra tempo .. :~i'a5.)
The subtle reply is 7 Jtb3, which
waits for Black to choose between
7... e6 and 7 ... g6. But after 7 ... g6
8 .ie3 Black has 8 ... lbg4! and
against 8 1'3 there is 8 ... lbxd4!
9 'i'xd4 Jtg7 with excellent play,
e.g. 10 Jte3 0-0 11 'i'd2 b5 12 Jth6
.ixh6 13 'i'xh6 b4 14 lbd5 e6!
15 lbe3 a5 16 h4 a4 17 Jtc4 d5!,
Shivaji-Motylev,
Minneapolis
2005.
But even a move down, White
has good chances following 10 0-0
Jte7 11 f4, theory says. In fact,
GMs often gives the tempo back,
say with 11...'i'c7 12 .tb3 0-0 to
reach a book line, 13 'i'f3 lbxd4
14 Jtxd4 b5, with equal chances.
81
Sicilian Defense
In the orthodox 6 .tc4 e6 White
chooses between the old 0-0 and
f2-f4-fS plan and Dragolyub
Velimirovic's attack with .te3,
'i'e2, 0-0-0 and g2-g4-gS. Black,
meanwhile, has to pick either quick
development ( ... .te7, ... 0-0) or
more
ambitious
but
slowdeveloping counterplay (... a6, ... bS,
.. .'iVc7).
.td7 10 .tb3 ""8 11 g4! :c8 12 gS
ttJe8 13 h4 ttJaS 14 g6! as in
Velimirovic-Milic, Belgrade 1965,
one of the earliest successes of
White's strategy.
Instead, Black does better against
the Velimirovic when he begins
queenside operations, e.g. 7 i.e3 a6
8 'ife2 "iic7 9 0-0-0 ttJaS or 9... i.e7
10 l:thgl ttJaS 11 i.d3 bS.
Matching the right defense with
each attacking plan and vice versa
is crucial. Queenside play didn't
fare well against the old Sozin when
6 .tc4 e6 7 0-0 was young. After
7... a6 8 .te3 'i'c7 9 .tb3 ttJa5 10 f4
b5:
What this means is each side
should be flexible enough to exploit
the plan his opponent chooses.
White's best waiting move is
7.tb3.
Then on 7....te7 8 .te3 0-0
Black often got crushed by 11 fS
ttJxb3 12 cxb3! (l2 ....te7 13
'iVd7 14 'i'f3 and IS eS or IS fxe6
fxe6 16 'iVh3).
... he can head toward the
Velimirovic with 9 "iie2!, having
the benefit of knowing where
Black's king lives.
:c1
An antidote was found in quick
development, 7....te7 8 .te3 0-0
9 f4? d5!.
The best waiting reply is 7... a6.
If White makes a commitment with
8 f4, then 8... .te7 9 .te3 0-0 heads
into an old Sozin tabia that offers
chances for both sides (10 0-0
ttJxd4! 11 .txd4 bS!).
But that policy doesn't fare well
against the Velimirovic, e.g. 7 .te3
.te7 8 'iVe2 0-0 and now 9 0-0-0
82
Sicilian Defense
The waiting game continues if
7 i.b3 a6 is followed by 8 i.e3.
Strong players, such as Vishy
Anand, have played 8.. :iic7 to keep
the queenside counterplay option
alive.
However switching to the old
Sozin with 9 f4! exploits Black's
tardiness and he is worse after
9...b5 10 ttJxc6 'iixc6 11 f5. Or
The novelty pays off when White
9... i.e7 10 0-0 ttJxd4 11 i.xd4 b5 plays 6 i.c4 and discovers after
12 e5!, which won quickly in 6 ... a6 7 i.b3 e6 that he's been
Golubev-Lerner, Odessa/Istanbul pulled into a popular Najdorf line
2006 after 12 ... dxe5 13 fxe5 ttJd7 that may be good for Black.
14 :xf7! ~xf7 15 'iif3+.
In the normal Najdorf order,
5 ... a6 6 i.c4 e6, White can
Both sides are finally out of
circumvent that position (7 i.b3
passes after 8 i.e3 and 8...i.e7!.
ttJbd7) by means of 7 O-O!, as
Then 9 0-0 0-0 is a good Sozin for played by that devious transposer
Black and 9 'iie2 0-0 10 0-0-0 is a Deep Fritz against Vladimir
Velimirovic tabia that is still being Kramnik in 2006.
evaluated.
But there are no other 5... ttJbd7
tricks
and plenty of drawbacks: On
RARE FIFTH MOVES
6 i.g5 Black's best is probably
... a6, reaching a Najdorf (5 ... a6
6
We know a lot about 1 e4 c5
2 ttJO d6 3 d4 cxd4 4 ttJxd4 ttJf6 6 i.g5 ttJbd7) that hasn't had
5 ttJc3 when Black replies 5... g6, success since the 1960s.
5... e6, 5... a6 and 5... ttJc6 but very
Also, White's d4-knight cannot
little about two alternatives. The be exchanged off by ... ttJc6xd4 so it
first is 5...ttJbd7, which has been dominates the center. He gets a
tried by Tigran Petrosian, Mark superior English Attack after
Taimanov, and Bent Larsen. In the 6 i.e3, e.g. 6... a6 7 g4 h6 8 f3 e6
Havana 1966 tournament book 9 'iid2 'iic7 10 0-0-0 and 11 h4, or
Petrosian wrote, 'The only 8...b6 9 'iid2 i.b7 100-0-0 e6 11 h4
advantage of this move... consists d5 12 i.h3! with a strong initiative,
Balogh-Bilek, Budapest 2002.
of its relative novelty.'
83
Sicilian Defense
The more successful alternative
is 5...i.d7, a specialty of Viktor
Kupreichik and Aloyzas K veinys.
the unclear 7 ltJdb5 i.c6 8 i.xf6.
Bottom line: 5... i.d7 has a few
advantages over the Classical order
but Black has to risk the worst-case
scenario, a somewhat questionable
English Attack or Rauzer.
NAJDORF VARIATION
The essential Najdorf move,
... a6, is useful in most Sicilian
variations and that gives Black
freedom to transpose into them.
Whether that makes sense depends
on which of the nearly a dozen
reasonable alternatives White picks
at move six. The move orders of
some of them, such as 6 i.e2, 6
i.g5 and 6 f4, are too complex to
consider here. Among the others:
This works as a waiting move in
cases such as 6 i.e2 g6, when Black
has dodged the Yugoslav Attack, or
6 ... e6, when he got a Scheveningen
without encountering the Keres or
English attacks.
The principle drawback to
5 ... i.d7 is the bishop occupies the
KN's best retreat square and that
improves the impact of g2-g4-g5 in
an English Attack (6 i.e3 or 6 f3).
(a) 6 i.d3
But in practice 6 i.g5 is White's
favorite. Then 6 ... ltJc6 transposes
into a 6 ... .td7 Rauzer. The only
apparent benefit to Black for doing
it that way, rather than 5 ... ltJc6
6 i.g5 i.d7, is to avoid the Sozin
and other Classical alternatives like
6 f4.
This virtually unbooked move
has great transpo value. After 6... e6
7 f4 or 7 0-0 White gets quite
reasonable Scheveningen positions
in which i.d3 replaces i.e2, e.g.
7 f4 ltJc6 8 ltJf3 i.e7 9 0-0 and
"iVel-g3. If Black never plays
For the sake of originality, GMs
prefer meeting 6 i.g5 with 6 ... e6.
That offers White a choice of a very
rare Najdorf line (7 f4 a6), another
Rauzer (7 "iVd2 ltJc6 8 f4 h6) and
84
Sicilian Defense
Scheveningen positions, because he
meets 6 ~e2 with 6 ... e5, then
6 ~d3 could be a problem.
2004. That can't be the best line for
Black after 6 ~d3. But what is?
(b) 6 a4
And if Black answers 6 ~d3 with
6... e5 White has a good retreat
square at e2, e.g. 7 ttJde2 ~e7
8 0-0 0-0 9 f4 ttJbd7 10 ttJg3
(10 ... exf4 11 ttJf5! ttJe5 12 ttJxe7+
'ilixe7 13 ~xf4 h6 14 ~g3! and
.th4, Gallagher-Karjakin, Panomo
2002).
This is White's best waiting
move in the Najdorf. Against 6... e5,
he can continue 7 ttJf3 so that
7....te7 8 ~c4 0-0 9 ~g5 or 7... h6
7 ~c4 will put him a tempo ahead
of 6 ~e3 e5 7 ttJf3 and 6 ~e2 e5
7 ttJf3 lines.
The drawback to 6 ~d3 would
seem to be 6•.. ttJc6, since the
attacked knight
cannot be
maintained on d4 (7 ~e3 ttJg4!).
However after 7 ttJxc6 bxc6 8 0-0:
Instead, Black may transpose to a
Scheveningen, Boleslavsky or
Dragon after 6 ... ttJc6. Evidence
indicates the a-pawn advances help
White in the Dragon. For example,
6•..ttJc6 7 .te2 g6 8 0-0 ~g7 9 ~e3
0-0:
White can continue ttJa4/c2-c4!
as he does in the Sozin line that runs
1 e4 c5 2 ttJf3 ttJc6 3 d4 cxd4
4 ttJxd4 ttJf6 5 ttJc3 d6 6 .tc4 'ilib6
7 ttJxc6 bxc6 8 0-0 and ttJa4 with a
bright outlook.
The hole at b6 gives White
chances for a queenside bind here
after 10 'ilid2 ttJg4 11 ~xg4 ~xg4
12 ttJd5! lic8 13 a5.
Also good is 10 f4 since
1O ... 'ilib6, which instantly equalizes
when the a-pawns are on their
original squares, can be met by the
strong 11 a5! ttJxe5 12 e5. If Black
has to settle for 10 ... ~d7 then
In the diagram he can answer
8... e5 with 9 b3 .te7 10 .tb2 0-0
11 ttJa4. Then 11...~e6 12 'ilie2
'ilic7 13 c4 ttJd7 14 c5! favored him
in Dgebaudze-Wemmers, Belgium
85
Sicilian Defense
11 ttJb3 and ttJd5 guarantees a
positional edge.
So Black turned to 7...b5. Then
White found 8 0-0 was good since
8 ... iLb7 9 :tel! ttJbd7 10 .tg5
ttJc5? 11 iLd5! is a sound sack, as
Fischer showed.
Another idea after 6...ttJc6 is
7 iLe2 e5. Books recommend
8 ttJxc6 and show how it only
equalizes. Better is 8 ttJb3! iLe7
90-00-0 and now 10 iLg5!:
To preempt that Black began to
play 7...ttJbd7 and ... ttJc5, much in
the spirit of the Scheveningen (1 e4
c5 2 ttJf3 d6 3 d4 cxd4 4 ttJxd4 ttJf6
5 ttJc3 e6 6 iLc4 i.e7 7 .tb3 0-0
8 iLe3 ttJa6!?). After 8 0-0 ttJc5
Black is faster than in the Fischer
line.
The counter-finesse is 7 0-0,
which had been discarded in the
1950s. Then 7 ... b5 8 .tb3
transposes to the 7 i.b3 b5 8 0-0
position that White wants and it
avoids the 7 i.b3 ttJbd7 one that
Black wants.
This favors White after 10...iLe6
11 iLxf6 and ttJd5. This is different
from the normal Boleslavsky
(5 ... ttJc6 6 i.e2 e5 7 ttJb3 i.e7)
when Black equalizes immediately
after 80-00-09 i.g5 ttJxe4!.
If Black plays 7...ttJbd7 anyway
then 8 i.g5!:
The addition of a2-a4 and ... a6
means that 10 i.g5 ttJxe4? dooms
the knight, 11 iLxe7 ttJxc3 12 iLxd8
ttJxdl 13 i.c7 ttJxb2 14 :tfbl.
(c) 6 iLc4
There's general agreement that
Black's best is 6... e6 but the
consensus about White's reply is
breaking down. The old favorite,
This transposes to a 6 i.g5
7 i.b3, was considered best
because it preserves the 0-0-0 Najdorf position that has been
option, e.g. 7 ...iLe7?! 8 f4 0-0 9 scoring well for White since the
'it'f3 and then 9... ttJbd7? 10 g4 or 1960s. Strong players such as
Leonid Yudasin have tried
9.. :ilc7 10 f5 ttJc6 11 i.e3.
86
Sicilian Defense
8... 'ii'c7?!. But 9 .txe6! fxe6
10 liJxe6 has been crushing Black
11 liJd5, as in Topalov-van Wely,
Wijk aan Zee 2007.
since a Keres brilliancy (l0 .. .'iic4
11 liJd5 rJif7 12 .i.xf6 rJixe6
~
13 .&.c3! and wins). To play this line
Black may have to risk 8... h6 9 .th4
g5 10 .i.g3 liJe5.
The crafty response to 6 .te3 is
6...liJc6.
He can avoid all this by
remembering why 7 0-0 was
abandoned half a century ago 7....i.e7! 8 .i.b3 0-0. Then 9 f4liJc6
reaches a fairly balanced main line
of the old Sozin and 9 .te3 b5 is an
equally double-edged line of the
Najdorf.
This avoids 6... e6 7 g4 and forces
an English Attacker to prepare
g2-g4. On 7 'iVd2 Black has an
active reply in 7... liJxd4 8 .txd4 e5
and .•..te6. And if White offers a
Scheveningen, with 7 .te2, Black
can force his way to a Classical
Dragon, 7... g6, in which White has
been deprived of the Yugoslav
Attack and .tg5 lines. The price he
pays is being committed to ... a6!?
(d) 6.te3
White is ready for a quiet
Scheveningen or an English Attack.
Some GMs, like Michael Adams,
would rather play against 6 .te3 e5
than 6 .te2 e5 so they begin with
6 .i.e3. If 6... e6, then 7 .i.e2 allows
them to transpose to 6 .te2 e6
7.i.e3.
Some English Attackers prefer
6 f3 and 6... e6 7 .i.e3. In this way
they avoid 6 .i.e3 liJg4. But they
also surrender options such as
6 .i.e3 e6 7 g4 and 6 .i.e3 e5 7 liJb3
.i.e7 8 'ii'd2!?
TAIMANOVIKAN VARIATION
The standard starting position of
the Taimanov Variation can be
reached in two different ways, and
Mark Taimanov used both of them,
1 e4 c5 2 liJf3 liJc6 3 d4 cxd4
4 liJxd4 e6 and 2... e6 3 d4 cxd4
4 liJxd4 liJc6.
In the last line White delays £2-f3
so that he rules out ... d5 and keeps
the £2-f4 option, as in 8 'ii'd2 .i.e6
9 0-0-0 liJbd7 10 f4. He can also
trick Black into a bad line after
8... 0-0 9 0-0-0 b5? 10 f3! .te6
The first order discourages the
King's Indian Reversed (3 d3),
which is promising against 2 ... e6.
87
Sicilian Defense
The second order eliminates the
Rossolimo, in view of 2 ... e6
3 .tb5? a6!. Black also may choose
an order based on how he intends to
meet 3 lLlc3 or 3 c3. For example,
1 e4 c5 2 lLlf3 e6 3 lLlc3 is a good
Chameleon for White but 2 ... lLlc6
3 lLlc3 e5! isn't.
Both orders, as well as a third
used by Salo Flohr, 1 e4 cS 2 lLlo
lLlc6 3 d4 cxd4 4 lLlxd4 'iVc7!?,
have the drawback of allowing a
Maroczy Bind. For instance, 4... e6
S lLlbS threatens lLld6+ and usually
buys time for 5... d6 6 c4!.
The timing of the three key
Taimanov moves, ... a6, ... lLlc6 and
... 'iVc7, has turned out to be more
complicated than it seemed a
decade or so ago.
Only recently was 1 e4 cS 2lLlo
lLlc6 3 d4 cxd4 4 lLlxd4 e6 S lLlc3
a6 tested by the English Attack, for
example. This has proven to be
dangerous after 6 .te3 lLlf6 7 g4 or
6... 'iVc7 7 'iVd2 lLlf6 8 0-0-0.
This provides an argument in
favor of S•.• 'iVc7.
Black can bailout of the bind
with S.•.lLlf6!?
Now on 6 ..te3 Black can
respond more vigorously with
6•••lLlf6 7 'iVd2 ..tb4! and then 8 f3
lLlxd4 9 .txd4 e5 or 9 'iVxd4 ..txc3+
10 bxc3 e5.
If White defends the attacked epawn with 6 lLllc3 then 6 ... d6
7 .tf4 e5 8 .tg5 is the Sveshnikov
Variation. So is 6 .tf4 e5 7 .tg5 d6
8lLllc3.
The drawback to this order may
be 7 lLldbS 'iVb8 8 f4 although
Black gets a tempo back after 8... d6
and ... a6.
Of course, Black has a simpler
route to the Sveshnikov if that's his
wish, and 5... lLlf6 makes most sense
when Black seeks a Taimanov and
is upset to see 5 lLlb5.
Fischer, who liked lLlb5/c2-c4
lines as White, reached the
Taimanov as Black through another
order, Ilya Kan's 1 e4 c5 2 lLlf3 e6
3 d4 cxd4 4 lLlxd4 a6. After SlLlc3
88
Sicilian Defense
he transposed (S ... ~c6) into the
S... a6 Taimanov.
and now S c4 ~f6 6 ~c3 .i.b4!) but
not versus the Kan. Another minus
Other Kan players delay ... ~c6 is that White can play S i.d3
further, by means of 5...fie7 in because his knight is not hanging on
order
to
disturb
White's d4 as it is in the Taimanov. That
development with ... b4 or ... .i.b4. enables him to create a delayed
For example, 6 i.e2 ~f6 7 i.e3? Maroczy Bind with c2-c4.
allows 7...i.b4!, e.g. 8 fid3 ~c6
Black
will
have
ample
threatening 9... ~eS.
opportunity in these orders to
Or 8 i.d2 i.xc3 9 i.xc3 ~xe4 transpose into a Scheveningen with
10 i.b4 'iVb6 11 i.a3 ~c6 with ... d6. But that makes the most sense
little compensation for the lost when White has played a move that
pawn in Bonte-Cabrilo, Timisoara is useful in the Taimanov but isn't
2006.
in a Scheveningen. An illustration is
Another plus of the Kan order
appears when White fianchettos,
5 ~c3 fie7 6 g3 and then 6....i.b4!:
a2-a3 to rule out ... i..b4 and to
anticipate ... b5-b4. A popular order
in the 1960s was 1 e4 e5 2 ~O ~e6
3 d4 exd4 4 ~xd4 e6 5 ~e3 a6
6 .te2 fie7 7 a3:
White cannot ignore ... i.xc3+.
But neither can he claim an
advantage after 7 ~e2 ~f6 or
Here 7... ~f6 8 0-0 .i.e7 9 .i.e3
7 i.d2 ~f6 (8 i.g2 ~c6 9 ~b3 d6! is a good decision. Compare it
i.e7! and ... d6 with an excellent with 1 e4 cS 2 ~f3 d6 3 d4 cxd4
Scheveningen).
4 ~xd4 ~f6 S ~c3 e6 6 .i.e2 a6
The drawbacks to the Kan order 70-0 fic7 8 f4 ~c6 9 .i.e3 i.e7 and
begin with S c4. This works badly now instead of normal lines such as
against the Taimanov (1 e4 cS 10 a4 or 10 fie1 White would
2 ~f3 ~c6 3 d4 cxd4 4 ~xd4 e6 transpose with 10 a3?!.
89
Sicilian Defense
But it's considered unwise for Mortensen). Black is ready to playa
Black to transpose into a normal Scheveningen after 9... lLlf6
Scheveningen after g2-g3 when and 10 ... 0-0.
he's already played ...'fic7. After
The only way to prevent that is
1 e4 c5 2 lLlf3 lLlc6 3 d4 cxd4
9 l:.el, with the idea of 9... lLlf6?
4 lLlxd4 e6 5 lLlc3 'fic7 6 g3 Black
10 lLlxc6!, reaching the 1960s
more or less has to play 6... a6 to
position. But the difference is
avoid 7 lLldbS 'iVb8 8 ..tf4.
9•••.i.f6!?
Then 6... a6 7 ..tg2 d6 8 0-0 lLlf6
Now 10 lLlb3, 10 lLlde2 and
9 liel! allows White to carry out his
10 .i.e3 aren't much after
strategic plan of lLlxc6 and e4-eS 10 ... lLlge7. The crucial lines are
which ... ~c7 is intended to prevent
10 lLlxc6 bxc6 11 .tf4 eS and
- by tactical means. That is, 9••. ..te7 10 lLlfS exfS 11 lLldS and unless
10 lLlxc6! bxc6 11 e5! dxe5 they prove more successful than
12 J::txe5 and 12 .. :ii'xe5 13 .txc6+. they have been, Mortensen's order
There have been no major is a significant strengthening of the
improvements in that line since the Taimanov.
1960s. For this reason many players
feel S... 'fic7 is a slight error that is
punished by 6 g3! and that S... a6 is
a superior waiting move (6 g3 d6!).
THE ...~6 VARIATIONS
There is a family of lines built
around ... 'iib6 and ... e6, often with
... ..tcS. They all use pressure on the
b6-f2 diagonal, at least at the start,
but often are used to reach the
Scheveningen. The family members
are:
But Black has another finesse
even if he's committed to ...'fic7
and ... a6.
(a) 1 e4 c5 2 lLlf3 lLlc6 3 d4 cxd4
4 lLlxd4 ~6,
(b) 1 e4 c5 2 lLlf3 lLlc6 3 d4 cxd4
4 lLlxd4 lLlf6 5 lLlc3 ~6,
(c) 1 e4 c5 2 lLlf3 e6 3 d4 cxd4
4 lLlxd4 ..tc5, and
(d) 1 e4 c5 2 lLlf3 e6 3 d4 cxd4
4 lLlxd4 ~6.
This can come about after
5•• :ii'c7 6 g3 a6 7 .i.g2 and then
7•.. d6 8 0-0 ..te7!? (Erling
The last two, popularized by
Michael Basman and Kveinys
90
Sicilian Defense
respectively, can transpose into one
another, e.g. 1 e4 cS 2 tLln e6 3 d4
cxd4 4 tLlxd4 iVb6 and now S tLlc3
i.cs 6 i.e3 tLlc6.
7 i.d3 i.e7 8 0-0 0-0 9 i.e3 "fic7
10 f4 d6 11 iif3 a6 12 a4 or 12 g4.
Theory says White is favored a bit
but Black's position is solid.
If that's what Black wants, which
is the best route to it? In order
(a) White has an extra option of
S tLlbS. But Black has a surprising
reply in S... a6 6 i.e3 'iVd8!.
Not 6.. :iVxb2? 7 tLldb5, which
was refuted in a Morphy-Paulsen (!)
game. Black has been getting quite
good positions after 6... tLlc6, e.g.
7 liJa4 'i'aS+ 8 c3 i.xd4! 9 i.xd4
tLlxd4 10 'i'xd4 liJf6 11 i.e2?! eS!
12 'i'c4 d6.
Black's most accurate route to
the diagram is Kveinys' . One
reason is that 4... i.c5 has not done
well when met by the surprising
5 .te3 'i'b6 6 c3!. For example,
6... 'i'xb2 7 'i'b3 'i'xal? 8 tLlc2 or
7 ... 'i'xb3 8 axb3. Or 6 ... tLlc6
7 tLld2! tLlxd4 8 tLlc4.
His point is that 7 tLld4 tLlf6
8 tLle3 e6 is a Taimanov Variation
with i.e3. That has a much better
reputation today thanks to 'iVd2 and
0-0-0 but it may not be something
White feels comfortable with,
especially ifhe normally plays i.g2
against the Taimanov.
This order may also confuse
White since 7 tLlSc3 and then 7... e6
8 tLld2 bS! is fine for Black. If
Black has any doubts about that, he
should prefer order (b) because
6 tLldbS a6 7 i.e3? is a mistake
(7 .. :iVa5 wins material) and 7 tLla3
e6 8 tLle4 "fie7 9 i.e3 bS has never
done well for White.
The three other orders often
reach Scheveningens after White
retreats his knight to b3. The benefit
to Black is avoiding the English and
Keres Attacks. For example, in the
Kveinys order, play may continue
5 tLlb3 tLlc6 6 tLlc3 liJf6 and then
91
Sicilian Defense
FOUR KNIGHTS VARIATION
For decades the variation that
begins 1 e4 c5 2 liJo liJc6 3 d4
cxd4 4 liJxd4 liJf6 5 liJc3 e6 was in
the Sicilian's 'dubious' basket
because the main line, 6 liJdb5 .tb4
7 a3, handed White the two bishops
and better prospects.
After 6 ... d6 Black says, 'I
outsmarted you. I reached a
Scheveningen in which your a2-a3
is a wasted tempo.'
This changed in the 1970s when
Black adopted the Four Knights as a
way of transposing, after 6 ... d6
7 .tf4 e5 8 .tg5 a6, to what was
dubbed the Sveshnikov Variation.
White replies, 'Not at all. My
goal was to avoid both the
Sveshnikov and 6 ... ..tb4. The added
This denies White options of the
normal Sveshnikov, 1 e4 c5 2 liJf3
liJc6 3 d4 cxd4 4 liJxd4 liJf6 5 liJc3
benefit is that I'll play 7 ..tc4! and
get a Sozin in which you've denied
yourself the Benko line.'
e5 6 liJdb5 d6 7 .tg5 a6, such as
6 liJf5, 6 liJdb5 d6 7 liJd5 and 7 a4.
Of course, there's a lot of theory
to know in the Four Knights as
well, if White decides not to join
him on the road to the Sveshnikov.
But Black can reduce that
significantly by shifting to a
Scheveningen with 6 ... d6.
Or Black can play 6 ... e5 and say,
'I'm the one who got what I
wanted, a Sveshnikov. In this
version you can't play 7 liJdb5 d6
8 .tg5 because your knight has no
retreat aft er 8... a6"..
To which White replies,
'Nonsense, I'm the one who came
out ahead because after 7 liJb3 it's a
Sveshnikov with an extra tempo for
t
me. That tempo sops
... A~b4'.. '
That is usually second-best. But
it avoids heavy analysis such as
6 g3 d5 7 .tg2 ..tb4 8 exd5 liJxd5 or
6 .te2 .tb4 7 0-0 .txc3 8 bxc3
liJxe4 9 .td3 d5 10 ..ta3. Instead,
Black is roughly equal after 6 g3
d6!? or 6 .te2 d6!?
There is no consensus on the best
reply to 6 a3 but 6 ... d6 can't be bad,
especially since 7 ..tc4?! really isn't
a good Sozin after 7 .. :ikc7! 8 ..ta2
a6 9 0-0 liJxd4 10 'iVxd4 liJg4.
There's a who-tricked-whom
quality to 6 a3.
92
Sicilian Defense
The Four Knights variation-into Zaitsev said he played this for the
Sveshnikov order was bound to first time in a five-minute game
spawn a counter-finesse. The ever- and later in a simul that went
creative Igor Zaitsev found one in 8...~xd5 9 exd5 exf410 dxc6 bxc6
11 'ii'f3.
6 ~db5 d6 7 i.f4 e5 8 ~d5!?:
Experience with 11...d5 12 0-0-0
remains mixed (12 ... .lid6? 13 'ii'c3!
cxb5 14 'iVc6+ .lid7 15 'ii'xd6
favored White in a ZaitsevSveshnikov game). But so do the
main lines of the Sveshnikov
Variation. The Sicilian is never easy
for Black - or for White.
93
Chapter Four:
Semi-Open Games
Black reveals something about
himself when he answers 1 e4 with
a move other than 1...e5 or 1...c5.
An Alekhine's specialist, for
example, almost certainly likes
positional imbalances and quick
counterplay. He may have a
problem with 1 e4 liJf6 2 liJc3!?
because the Vienna Game (2 ... e5) is
not likely to give him what he
wants.
can lead to a Steinitz French (4 f4
e6 5 d4). But the alternatives are
somewhat questionable, such as
3... liJe4 or 3... d4 4liJce2liJg4.
If White is going for a bigger
edge in the Alekhine's, the
traditional choice is the Four Pawns
Attack, 1 e4 liJf6 2 e5 liJd5 3 c4
liJb6 4 d4 d6 5 f4.
The standard response is 5... dxe5
6 fxe5liJc6. It sets some traps based
on a liJf3/ ... .tg4 pin. For example,
7liJf3 hopes for 7... .if5? 8 d5! liJb4
9liJd4 with advantage but 7... .tg4!
is excellent for Black.
Moreover,
few Alekhine's
players are prepared for a French
(2 ... e6 3 d4 d5) or a Pirc (2 ... d6
3 d4), and a Sicilian (2 ... c5?! 3 e5)
is dubious.
On 2... d5 3 e5:
The same goes for 7 .te3 .if5
8 liJc3 e6 9 liJf3 and 9... .ig4!,
despite the loss of time, e.g. 10.te2
.txf3 11 gxf3 'iih4+ 12 .tn 'l'f4.
White avoids this by simply
reversing the order, 9 .te2! and
10 liJf3.
A more deceptive Black will
delay both ... dxe5 and ... liJc6 as
Viktor Korchnoi did with 5....tf5
6liJc3 e6.
Black may be reluctant to play
the best move, 3... liJfd7! because it
94
Semi-Open Games
12 i.e3ltJc6 gets Black to where he
wants to go. When this order was
introduced in a 1960 GellerKorchnoi game White became
confused and was worse after
11 ~f4? ltJc6 12 exf6 ..txf6.
ALEKHINE'S: 4ltJf3
When White meets 1 e4 ltJf6
This pays dividends after 7 ~d3? 2 e5 ltJd5 with the more modest
i.xd3 8 'i'xd3 c5 (9 d5? dxe5 3 d4 d6 4 ltJf3 the timing of pawn
10 fxe5 'i'h4+ and ... 'i'xc4). Better exchanges on d6 and e5 become
is 7 ltJD, after which 7 ... dxe5 pivotal. After 4...i.g4 for example
8 fxe5 ltJc6 9 ~d3?! i.g4! is White looks for the right moment
for exd6, since ... cxd6 allows him to
another trap.
grab space with d4-d5!.
But after 9 i.e3! ~e7 White has
But on the immediate 5 c4 ltJb6
reasonable chances in either the
6
exd6
Black has 6... exd6!. That's
well-trod 10 i.e2 0-0 11 0-0 f6 or
the enterprising 10 d5!? (1O ... ltJb4 a satisfactory version of the
Alekhine's Exchange Variation,
11 ltJd4 i.g6 12 a3).
more commonly seen after 4 c4
Since 10 d5 demands exact ltJb6 5 exd6 exd6. More accurate is
preparation by Black, he may prefer 5 i.e2 e6 and then 6 c4ltJb6 7 exd6
to meet 7 ltJf3 dxe5 8 fxe5 with or 7 0-0 ~e7 8 exd6.
8•..i.e7!. This allows him to reach a
Another question concerns
Four Pawns tabia smoothly after
h2-h3. In most cases Black will
9 i.e2 0-0:
simply retreat to h5. But 5 ..te2 e6
6 h3 is mistimed in view of
6... ~xf3! 7 ~xf3 c6.
Now 10 d5 i.b4! collapses the
center and 10 0-0 f6 11 exf6 ~xf6
95
Semi-Open Games
Black wants to set up a target on
e5 (8 0-0 dxe5 9 dxe5 lbd7 10 'iVe2
'tic7). He benefits from White's
sloppiness because 8 c4 lbb6
9 exd6 allows the favorable
9••• lbxc4!.
If the bishop were still on g4
Black would stand well with ....tf6.
But here it's en prise and that gives
White time for 14... i.g6 15lbd4!.
His pieces dominated after
15 ... ..tf6 16 lbxc6 bxc6 17 'i'xc6,
as in A.Sokolov-Arzhenkov, Bor
2000 (17 ... i.xb2 18 ::tadl 'i'a5 19
'tixd6 'iVxa2 20 ..tf3 ::tac8 21 c5! a5
22 c6).
If White intends to put the
question to the bishop, a better
order is 6 O-O! i.e7 7 h3. Then
7.....txf3 8 i.xf3 c6 9 c4 lbb6 loses
material to 10 exd6 and 11 c5. The
reason Black needs 8 ... c6 is that
8... dxe5 fails to 9 c4! lb-moves
10 i.xb7.
A separate branch of the 4 lbf3
tree is 4... g6. Today 5 .tc4 and
5... lbb6 are considered best. But the
old 5... c6!? has a new use thanks to
Moreover 7...i.f5 permits White
to carry out the space-grabbing plan
transposition.
under ideal conditions. For example
8 c4 lbb6 9 lbc3 0-0 10 i.e3 lbc6
and now 11 exd6 cxd6 12 d5! exd5
13 cxd5! lbe5 14 lbxe5! dxe5
15 "iVb3 with advantage.
After 6 0-0 White seems to be
getting a good Exchange Variation
in view of 6.....tg7 7 exd6! exd6
8 ::te1+.
The main benefit of h2-h3 lies in
7... i.h5 8 c4 lbb6 9 lbc3 lbc6
10 exd6 and then 10... cxd6 11 i.e3
0-0 12 d5!. This is quite good
for White, particularly after the
natural 12 ...exd5 13 lbxd5 lbxd5
14 'tixd5:
However, 6... dxe5! 7 lbxe5 ..tg7
is the order of a slyboots. What has
happened is that Black transposed
into a line attributed to Edwins
Kengis. It usually comes about via
4 lbf3 dxe5 5 lbxe5 g6 6 ..tc4 c6
70-0 i.g7.
96
Semi-Open Games
11...0-0 12 b3 .tf6 13 .tb2 as
14 0-0 with a space edge, EmmsBaburin, Isle of Man 1997.
Black has been equalizing, e.g.
8 c3 lbd7 9 lbxd7 ..txd7 10 lbd2
0-0 11 lbf3 ..te6 12 'iie2 lbc7!
13 iH4 ..txc4 14 'iixc4 lbe6
IS ..teS 'iidS, Sarbok-Carlsen,
Gausdal 2003.
This possibility wouldn't arise in
a normal Exchange (4 c4 lbb6
S exd6 exd6 6 lbf3) if Black is
careful,
e.g. 6.....te7 7 lbc3 0-0
The point of the 4... g6 S ..tc4 c6
order is that Black avoids some of 8 ..te2 and 8 ...lbc6 9 .te3 .tg4!.
the options in Kengis's order, such
as (4lbf3 dxeS S lbxeS g6) 6 c4 and CENTER COUNTER DEFENSE
6 'iif3. If Black likes the Kengis
Once upon a time a player who
line, 4... g6 S ..tc4 c6 may be the answered 1 e4 with 1...d5 2 exd5
most precise way to reach it.
'iixd5 just wanted to get out of
Finally, there is 4 lbf3 lbc6. The book and was hoping White didn't
reputation of this line rests on the know that 3 lbc3 'iia5 4 d4 lbf6
latest analysis of a gambit, S c4 5 lbf3 .tg4 6 h3! favors him
lbb6 6 e6!? fxe6. If White isn't significantly (6 ... ..thS 7 g4! .tg6 8
confident about what to do then, he lbeS).
should bail out with 6lbc3.
Today Black usually plays 1 e4
dS with ... ..tfS to obtain a CaroKann pawn structure with more
active play. But he watches for a
chance to transpose into a good
version of ... ..tg4. For example
4 lbf3 lbf6 5 ..tc4 c6.
Then 6.....tg4 (or 6... dxeS 7 dS)
7 exd6 exd6 transposes to an
Exchange Variation in which Black
committed his queenside pieces too
early.
Now 6 0-0 ..tg4! has the benefits
of a superior Caro-Kann (7 d4 e6
8 .:tel lbbd7).
For example, 8 ..te2 ..te7 9 d5!
and 9... ..txf3 10 ..txf3 lbeS 11 ..te2
lbexc4? 12 ..txc4lbxc4 13 'iia4+ or
97
Semi-Open Games
But Black is the one most likely
to be outfoxed in the Center
Counter even when he thinks he is
the fox. After 4 tiJf3 he might be
tempted to seize the center with
4 ... e5. But then 5 d4! traps him in a
discredited old line that used to
arise after 4 d4 e5?! 5 tiJf3!.
The same idea succeeds,
however, after 9 0-0-0 tiJd5 in view
of 10 ~xd5 cxd5 11 'iib5+ tiJd7!
(12 'iVxb7 l:.b8 13 'iVc6 ~a3).
White also derives some extra
options from 9 tiJe5 !, such as
9 ..•tiJbd7 10 tiJxd7 tiJxd7 11 g4
~g6 12 h4, which has been
dangerous the few times it's
occurred.
The major finessing of the Center
Counter begins after 4 d4 tiJf6
5 tiJf3 (or 4 tiJf3 tiJf6 5 d4). The
waiting move 5 ... c6 anticipates
White most dangerous strategic
idea, d4-d5, and avoids queen traps
(6 tiJe5 tiJbd7 7 tiJc4 'iWc7!).
The most dangerous situation for
Black is when both ~c4 and 'iie2
are played early on so that the
strategic d4-d5! break is in the air.
For example, 1 e4 d5 2 exd5 'iVxd5
3 tiJc3 'iVa5 4 d4 tiJf6 5 ~c4 c6
6 ~d2 ~f5 and now 7 'iVe2! allows
White's bishops to take command
after 7 ... e6 8 d5! cxd5 9 tiJxd5 'iVd8
10 tiJxf6+.
The more common reply to 5... c6
is 6 ~c4. This reaches a tabia after
6... ~f5 7 iLd2 e6 8 'ife2 ~b4 when
White chooses between 9 tiJe5 and
castling. The two often transpose 9 tiJe5 tiJbd7 10 0-0-0 and 9 0-0-0
tiJbd7 10 tiJe5.
The best try to exploit White's
order is 5 ..• ~g4!?, rather than
5 ... c6.
But the more accurate knight
move cuts down Black's options.
The simplifying 9 tiJe5! tiJd5? is
dreadful (10 iLxd5! cxd5 11 'iWb5+!
'iVxb5 12 tiJxb5 iLxd2+ 16 <;i;xd2
tiJa6 17 tiJd6+ and wins).
Then 6 tiJf3? tiJc6! and 7•.. 0-0-0
gives Black the most aggressive
setup he can get in the Center
Counter, e.g. 7 ~b5 0-0-0 8 ~xc6
bxc6 9 0-0 e5 10 'ifd3 exd4 11 tiJe2
98
Semi-Open Games
.txf3 12 'iixf3 'iid5, DavidKosteniuk, Dresden 1999.
A more subtle order is 3 d4
tbxd5 4 .te2!?
And on 6 f3 .tf5 7 tbge2 e6
Black is secure, e.g. 8 .td2? 'iib6
9 g4 .tg6 10 h4 h6 11 .tb3 tbc6!
12 .te3 0-0-0 13 'iid2 .tb4
14 0-0-0 tbd5 15 .tf2 tba5 with
advantage in Movsesian-A.Kogan,
Nova Gorica 2000.
CENTER COUNTER: 2...tbf6
If you're going to seek a big edge
against 1 e4 d5 2 exd5 tbf6 you
probably have to master the
complex 3 .tb5+ lines. But if you
just want a reasonable middle game
you can save midnight oil with 3 c4.
Black's best is 3... c6 4 d4 cxd5,
transposing into the Panov
Variation of the Caro-Kann.
This rules out 4 ... .tg4, and
transposes into a main Whitefriendly line after 4 ... g6 5 tbf3.
Black might prefer 4....tf5 on the
grounds that he tricked White into
one of the quietest 3 .tb5+ lines
(3 ... .td7 4 .te2!? tbxd5 5 d4 .tf5 ).
But even if he's been tricked,
And if you're looking for White still enjoys good prospects
something in between, there is 3 d4 after 4...i.f5 5 tbf3 e6 6 0-0 i.e7
and 3 tbf3. After 3... tbxd5 Black 7 c4. e.g. 7...tbb4 9 tba3 tb8c6
will likely choose between two 10 .te3 0-0 11 ~d2 ~d7 12 .l:.fc1
basic plans of development - ... g6/ :fd8 13 tbb5! tba6 14 a3 ~e8
... .tg7 and ... .tg4/ ... e6. Against 15 :d1, Van der Weide-Buschke,
either plan White will operate with Senden 2006 .
.te2, 0-0 and c2-c4.
MODERN DEFENSE
The virtue of 3 tbf3 and then
3...tbxd5 4 d4 lies in avoiding the
sharpest versions of the Portuguese
Gambit (3 d4 .tg4!?). If Black
insists on meeting 3 tbf3 with
3....tg4, White seems to get the
upper hand with 4 .tb5+! in view
of 4... c6 5 dxc6 tbxc6 6 h3 .th5
7 .txc6+ bxc6 8 d3 or 4... tbbd7
5 h3 .th5 6 tbc3.
Black often uses 1.. .g6 and
2....tg7 as a form of reconnaissance. He waits for White to show
his cards before he decides whether
to transpose into a Pirc Defense
with ... tbf6.
This makes sense when, for
example, White has given up his
99
Semi-Open Games
most dangerous anti-Pirc weapons 2 d4 ttJf6 3 ttJc3 g64 g3) but can be
such as a quick .te3-h6xg7. It's interrupted in the Modem, 1 e4 g6
more a matter of taste after 1 e4 g6 2 d4 .tg7 3 ttJc3 d6 4 g3 ttJc6!.
2 d4 .tg7 3 ttJc3 d6 4 f4, when
Then 5 ttJge2? .tg4!, 5 d5 ttJd4
Black can choose between
and 5 .te3 e5, are not the kinds of
transposing into the Austrian Attack
position White wants when he plays
of the Pirc (4 ... ttJf6) or staying in
g2-g3.
the Modem with 4 ... c6, 4 ... ttJc6 or
4 ... a6.
But the Pirc option soon expires.
After 4 f4 c6 5 ttJf3 ttJf6?:
This is a bad Pirc because ... c6
doesn't fit in with 6 .td3 0-0 7 0-0
b5 and then 8 e5! dxe5 9 fxe5 ttJd5
10 ttJxd5 'iVxd5 11 a4.
For example, 5 d5 ttJd4 6 ttJbl!?,
to trap the knight, leads to doubleedged play like 6... c6 7 c3 ttJb5
8 .tg2 cxd5 9 exd5 ttJc7 10 ttJe2
ttJf6 11 0-00-0 12 c4 b5!, RaetskySakaev, St. Petersburg 1999.
A more precise White will prefer
Black was clearly worse after 4 ttJge2!. Then 4•••ttJf6 5 g3 gets
11...c5 12 'iVe2! .tb7 13 c4 bxc4 him to the desired Pirc position. If
14 .txc4 'ii'd8 15 e6, as in Ghinda- Black insists on 4 ... ttJc6, White
Bogdan, Rumanian Championship gains space with 5 d5! ttJe5 6 f4 and
1997, a case of Modem player .te3/ttJd4 gives him a much better
finding himself in an unfamiliar position than he normally gets in a
g2-g3 variation.
Pirc.
White has to make a key decision
at move four in the Modem.
Suppose he wants to devleop his
KB at g2. The fianchetto is carried
out smoothly in the Pirc (1 e4 d6
The .tc4 lines tell a similar story.
In the Pirc (1 e4 d6 2 d4 ttJf6 3 ttJc3
g6) White has had some success
with 4 .tc4 .tg7 5 'iVe2. But in the
Modem he faces 4 .tc4 ttJc6!?
100
Semi-Open Games
Then 5 d5 is inconsistent and on
other moves Black can shift into a
pleasant Pirc, such as 5 ttJf3 ttJf6
6 0-0 i.g4 or 5 i.e3 ttJf6 6 f3 0-0
7 ttJge2 e5.
The moral is that White needs
ttJc3 in the Pirc order to defend e4
but he should delay it in the Modern
if he wants the option of meeting
... ttJc6 and/or ... i.g4 with c2-c3!.
The favored order is 3 ttJf3 and then
3... d6 4 i.c4 ttJf6 5 'iie2.
Aside from Modem-into-Pirc,
Black can also try Modern-intoDragon by inserting ... c5. For
instance, I e4 g6 2 d4 i.g7 3 ttJc3
and 3 ... c5 seeks an Accelerated
Dragon (4 ttJf3 cxd4 5 ttJxd4 ttJc6)
without having to deal with the
Maroczy Bind.
In practice he succeeds about one
third of the time because I e4
players generally avoid the Schmid
Benoni, 4 d5, and they have doubts
about 4 dxc5!?
It also pays to play 3 ttJf3 d6
4 i.e2! before ttJc3 when White
seeks the Classical setup, since
4••.ttJf6 5 ttJc3 transposes to a
heavily trafficked Pirc.
This avoids some offbeat lines
such as 4 ttJc3 i.g4 or 4 ... a6. In
contrast, 4 i.e2 i.g4 5 h3 i.xf3
6 i.xf3 ttJc6 7 c3 e5 8 dxe5 favors
White's bishops. By delaying ttJc3
White also discourages ... a6/ ...b5
because there is no knight for ... M
to attack.
MODERN: 3... c6
When Black plays ... c6 before
... d6 he creates problems for a
White who meets 1 e4 g6 2 d4 JLg7
3 ttJc3 d6 with a move of the QB or
a KB fianchetto. For example
4 i.g5 is reasonable after 3... d6.
But after 3 ... c6 it's a dubious
gambit, 4 i.g5 'ifb6!.
And 3•.. c6 4 g3 invites 4••. d5!,
when White's bishop bites on
proverbial granite, e.g. 5 exd5 cxd5
6 JLg2 ttJf6 7 ttJge2 0-0 8 0-0 ttJc6
9 h3 b6 10 i.g5 e6 and ... iLa6, in
Ambrus-Eliseyev, St. Petersburg
2002.
But ... d5 and ... 'ifb6 aren't the
only ideas behind 3... c6. White has
to be wary of being trapped in an
unfamiliar Pirc if Black follows
with ... d6 and ... ttJf6, e.g. 4 ttJf3:
101
Semi-Open Games
Note that White can discourage
the Gurgenidze after 1 e4 g6 2 d4
.tg7 by means of 3 f4.
Black should transpose with
4 •.. d6! S .te2 ttJf6 into a main Pirc
line. This usually arises from 1 e4
d6 2 d4 ttJf6 3 ttJc3 g6 4 ttJf3 .tg7
5 .te2 0-0 6 0-0 c6 and is
considered fairly even. Thanks to
the 3 ... c6 order Black has dodged
the sharper Pirc lines, based on
f2-f4 and/or .tg5.
The sternest test of 3 ... c6 is 4 f4.
Then 4 ... d6 5 ttJf3 ttJf6? is a bad
Pirc, as noted above, and 5 ... .tg4 is
a somewhat dubious Modern. Black
often prefers 4 ... d5, seeking a
Gurgenidze Variation.
There are several routes to a
Gurgendize tabia, including 1 e4 g6
2 d4 .tg7 3 ttJc3 c6 4 f4 ~6!?
S ttJf3 dS and then 6 eSt, since
6 exd5 .tg4! would grant Black
plenty of compensation for a pawn.
Which route to the tabia is best?
The only thing that seems certain is
Black should avoid 1 e4 c6 2 d4 d5
3 ttJc3 g6 if he really wants a
Gurgenidze. White has too many
excellent alternatives to 4 e5 and
5 f4, such as 3 cxd5 and 3 e5.
And then 3 ••• c6 4 ttJf3 dS S eS.
Unlike the normal Gurgenidze,
when his knight is misplaced at c3,
here he can support his center with
c2-c3. And on 3 ... d6 4 ttJf3 White
has a promising Pirc. 'I felt like a
broken man after 3 f4,' Julian
Hodgson once said when it was
played against him.
Nevertheless Black has a good
response in 3 ... cS!, which is good
here because 4 c3? does a poor job
of supporting the center (4 ... cxd4
5 cxd4 'iib6!).
PIRC DEFENSE
The standard Pirc order is 1 e4
d6 2 d4 ttJf6. White's e-pawn is
attacked and he is more or less
limited to 3 .td3, 3 ttJd2, 3 f3 and
the main line, 3 ttJc3.
The first two moves are attempts
to reach Yefim Geller's quiet
system in the Modern Defense (1 e4
102
Semi-Open Games
g6 2 d4 i.g7 3 c3 d6 4 lbf3 and
i.d3/lbbd2). If he succeeds, it's a
small victory since Pirc players are
often uncomfortable in Modem
lines.
On 3 i.d3 Black can avoid the
Geller system with 3... eS. Then 4 c3
dS!? is the active way of disturbing
White's plan. The more solid way is
4...lbc6 Slbf3:
(12 dxe5 lbxe5 13 lbxe5 'i¥xe5
14 lbf3 ~5 15 e5 i.f5) in
Rublevsky- Azmai parashvili,
internet 2004.
With 3 f3 instead, White is
signalling he's willing to play a
Samisch King's Indian. An example
of what can happen to a strong GM
who likes the Pirc as Black but not
the KID was A.Rodriguez-Kuzmin,
Minsk 1982:
3... g6 4 c4! i.g7 5 lbc3 lbc6
6lbge2 0-0 7 i.e3 a6 8 lbc 1 e5 9 d5
lbd4 10 lbb3 c5?! 11 dxc6 bxc6
12 lbxd4 exd4 13 .i.xd4 l:i.b8
14 'i¥d2 'iia5? 15 l:i.cl .l::i.d8 16 i.e2
i.e6 17 0-0 'iib4 18 b3 and White
won.
Black can try to avoid that fate
with 3 ... eS. Then he's offering
White a tiny-edge endgame, 4 dxe5
dxe5 5 'iixd8+. But 4 lbe2!
followed by 5 c4 and 6 lbbc3 heads
to the Samisch.
Who is more uncomfortable
here? A Black who likes the sharp
tactics of a Pirc may not like
defending a Pseudo-Lopez after
If White plays the conventional
S...i.e7 6 h3 0-0 7 lbbd2. But 3 lbc3 after 1 e4 d6 2 d4lbf6 Black
White can find himself, after has an alternative to 3... g6 in 3... c6,
S...i.g4 6 dS lbe7 7 c4, in an Old a system worked out by the Omsk
Indian - a favorable one but one master Anatoly Ufimtsev.
that may not suit a 1 e4 player.
Black can also tum 3 i.d3 or
3 lbd2 into a Philidor by means of
... lbbd7, as in 3 lbd2 eS 4 c3 i.e7
S lbgf3 lbbd7. White's setup is so
passive that his edge usually
disappears by move 12.
For example, 6 i.e2 0-0 7 0-0 c6
8 l:i.el 'ikc7 9 i.f1 l:i.e8 10 'ikc2 i.f8
11 b4 d5! and Black was soon better
103
Semi-Open Games
Black waits: After 4 tiJt3 or 4 g3,
After 6 0-0 cxd4 7 tiJxd4 0-0,
he can bring about a tame version of he's transposed into a Classical
Dragon. On 6 d5 0-0 it's a Schmid
the Pirc with 4 ... g6.
Benoni but a Schmid in which
But if 4 ..tg5, then 4... tiJbd7 and White was denied the useful
S... eS/... ..te7! is a good Philidor ..tbS+!.
because White's QB is misplaced.
The tactical justification is
And on S f4 'iVaS 6 'iVd2 bS! 6 dxc5 'ifa5! with good play. This
Black obtains faster counterplay explains why S... cS is better than
than in a normal Pirc because he S... O-O 6 0-0 cS 7 dxcS dxcS, which
hasn't spent tempi on ... g6/.....tg7, is regarded as favoring White after
e.g. 7 ..td3 tiJb6 8 tiJf3 b4 9 tiJdl 8 'ifxd8 and 9 ..te3.
dS! 10 eS tiJe4 as in WellsThe most dangerous Pirc line is
Hodgson, Edinburgh 1989.
4 f4 and the usual continuation is
4.....tg7 5 tiJt3 0-0 6 ..td3. Black
The test of Ufimtsev's order is may try to avoid this with S... cS
4 f4!. Then neither the Pirc (4 ... g6?! 6 dxcS 'ifas and 7... 'ifxcs since
S tiJf3) nor Philidor (4 ... tiJbd7?! 7 exd6? allows a strong 7... tiJxe4.
S tiJf3) are active enough. Ufimtsev
But White can juggle the move
preferred 4...'iYa5!, which enjoyed
order, 4 f4 ..tg7 5 ..td3!?
some vogue in the 1980s but
remains slightly suspect.
The starting point for most Pirc
Defenses is 3 tiJc3 g6. Black looks
for a chance to attack the center
with ... cS. For example, 1 e4 d6
2 d4 tiJf6 3 tiJc3 tiJf6 4 tiJt3 ..tg7
5 ..te2 c5.
This avoids 5... c5 6 dxcS 'iYaS
because e4 is already protected
(7 cxd6!). Also bad is 6... dxcS 7 eS
(7 ... tiJdS? 8 ..tbS+).
If Black plays 5•.. 0-0 instead,
White transposes into the main line
with 6 tiJf3!.
104
Semi-Open Games
The downside may lie in the
extra options Black enjoys after
5 .td3, such as S...lbc6 6 dS lbb4
or 6lbf3 .tg4. But there's been too
little experience to tell.
CARO-KANN DEFENSE
The Caro-Kann contains transpo
traps even after the super-quiet 1 e4
c6 2 d3. One is set by 2... dS 3 lbd2
lbd7 4 lbgf3.
Another refinement, this time by
Black, occurs with 4 f4 .tg7 Slbf3
lba6!? This seems like an offer to
transpose, after 6 .td3 0-0, into a
variation that runs 4 f4 .tg7 5 lbf3
0-0 6 .td3 lba6 and is rated as
favoring White after 7 0-0 c5 8 d5
and f4-f5.
But Black has the extra option of
6 .td3 cst?~
We learned from a famous TalSmyslov game that 4... eS? S d4!
favors White. We can also see that
4...lbgf6 S eS! is a better-than-usual
French Defense (5 ... lbg4 6 d4 e6
7lbb3).
If Black prepares ... lbgf6 with
4••:ilic7, White plays S exdS! cxdS
6 d4 and has a favorable form of the
The idea is 7 dS .tg4 8 0-0 lbd7
or 7 .. :iic7 and ... c4 when his
counterplay runs faster than usual.
For example, 7 d5 'ikc7 8 'ike2
0-0 9 a3 e6! 10 dxe6 .txe6 was
Bareev-Christiansen, Biel 1991,
which went 11 .tc4 l:!.ae8 12 0-0
.txc4 13 'ikxc6 'ikc6! 14 e5 lbg4
with good play.
old Exchange Caro-Kann.
For example, 6 ... e6 7 .td3 lbe7
8 0-0 g6 9 :e 1 iL.g7 10 lbfl lbc6
11 c3 0-0 12 iL.g5 and TiviakovDreev, Gothenburg 2005 continued
12 ... e5 13lbe3! lbb6 14 dxe5lbxe5
15 iL.f4 lbxf3+ 16 'iVxf3 'ikc6
17 lbc2 and lbd4 with clear
superiority.
The flies in the 5... lba6 ointment
The safest path to equality after
are 6 e5 and 6 .tc4 0-0 7 e5. But 2 d3 is 2... dS 3 lbd2 eS 4 lbgf3
once again there's been virtually no .td6! (not 4 ...lbd7? 5 d4! which
transposes to Tal-Smyslov).
experience to judge them.
105
Semi-Open Games
Then 5 d4 resembles a Tarrasch
French after 5•.. exd4! 6 exd5 cxd5.
But Black is a tempo (....td6)
ahead of the Tarrasch version, 1 e4
e6 2 d4 d5 3 liJd2 c5 4 exd5 exd5
5 liJgf3 cxd4, and he should
equalize.
Then 4 ttJgf3 .tg4 eases Black's
game appreciably. White should
pass with 4 c3. Then 4... dxe4
5 liJxe4 .tf5 6 liJg3 .tg6 reaches a
Caro-Kann main line but with the
addition of .. :iib6 and c2-c3.
That makes it harder for White to
develop his QB without hanging the
b2-pawn. However, 7 h4 h6 8 liJh3
may make c2-c3 more useful than
.. :iib6, e.g. 8... liJd7 10 liJf4 .th7
11 .tc4 e6 12 'ilie2, Lane-Pedersen,
Trnava 1985.
You will often see GMs play 1 e4
c6 2 d4 d5 3 liJd2 and transpose
into main lines after 3 ... dxe4
4 liJxe4. This order sidesteps the
obscure 3 liJc3 b5, which can lead
to a version of the ... exf6 Caro, after
CARO: TWO KNIGHTS
4 a3 dxe4 5 liJxe4 liJf6 6 liJxf6+
exf6, that may be better for Black
The 1 e4 c6 2 liJc3 d5 3 liJf3
than usual because 7 .tc4 is not order transposes into main (3 d4)
possible.
lines, after 3... dxe4 4 ttJxe4 ttJd7
5 d4 or 4... liJf6 5 liJxf6+ and 6 d4.
But there's a more substantial However, there are two traps for the
reason for 3 liJd2. White can meet unwary Black.
3••. g6 with 4 c3, as Anatoly Karpov
One is notorious, 3••. dxe4
showed in a 1990 game with Deep 4 liJxe4 .tf5 5 liJg3 .tg6?, when
Thought that went 4....tg7 5 e5! f6 White uses his extra knight move
6 f4 with a plus for White.
with 6 h4 h6 7 ttJe5! .th7 8 'ilih5 g6
The drawback to 3 liJd2 is that
the d-pawn is unprotected after
3••• 'iVb6.
9.tc4.
The other trap, with 4...liJd7
5 .tc4 liJgf6 6 ttJeg5 e6 7 'iie2,
106
Semi-Open Games
wasn't known until 1954, largely
because Aron Nimzovich's ... tiJd7
wasn't played much before then.
If d2-d4 has been inserted in
place of tiJf3, Black obtains a solid
game from 7...tiJb6. But here that
would drop pawn to 8 tiJeS! and
Black has to accept the slight
inferiority of 7... tiJd5.
The main Two Knights line is
1 e4 c6 2 tiJc3 dS 3 tiJf3 .i.g4 and
then 4 h3 .i.xf3 S 'ilixf3. In the
1950s Black played 5... e6 until he
found that 6 d4 dxe4 7 tiJxe4! was a
sound gambit, e.g. 7... 'ilixd4 8.i.d3
tiJf6 9 c3 'ilid8 10 0-0 and .l:tdl.
So he adopted S... tiJf6 because
6 d4 dxe4 7 tiJxe4 'iVxd4! is
unsound.
Books used to dismiss this by
citing a 1951 game that went 7 .i.d3
.i.e7 8 e5 tiJfd7 9 'ilig3 with an
effortless attack. But Black
improves with 7..• dxe4! 8 tiJxe4
tiJxe4 9 'ilixe4 tiJd7.
For example, 10 c3 tiJf6 11 'iie2
'iVd5 equalizes after 12 0-0 .i.d6
13 ':'e 1 0-0 14 .i.c4 'iiia5 15 .i.d2
.l:tfe8 16 .l:tadl ':'ad8 17 .i.c1 'iih5!,
Krai-Kamsky, U.S. Championship
2004.
If White wants more after S... e6
6 d4 tiJf6 he may have to try the
French-like 7 eS or 7 .i.gS but with
his QB traded it should be a good
French for Black.
But Karpov has exclusively
played S... e6. Its benefits include
keeping f6 free, e.g. 6 d3 tiJd7
7 i..d2 .i.c5 8 'ilig3 'ilif6!,
Chemyshov-Lastin,
Voronezh
2006.
Meanwhile, 4....i.hS (after 1 e4
c6 2 tiJc3 dS 3 tiJf3 .i.g4 4 h3) has
been deeply analyzed, beginning
with S g4 .i.g6 6 exdS cxdS
7 .i.bS+. When an opponent plays
4 ... .i.hS he is usually boasting that
his memory is better than yours.
The main point of S... e6 is
revealed by 6 d4 tiJf6!:
The mental hygiene alternative is
S d4 e6 6 .i.d3.
107
Semi-Open Games
Then 5 tbf3 ~g4 6 tbc3 tbf6
transposes to a main Panov line.
But the drawback is 5 cxd5! 'iixd5
6 tbfJ.
White's point is that 6••. dxe4
7 tbxe4 ~xf3 8 'iixf3 'iixd4 9 0-0 is
a superior version of the 4 ... ~xf3
5 'iixf3 e6 6 d4 gambit.
Black should probably decline
with 6...tbf6. Then 7 e5 tbfd7 8 g4
~g6 9 i.xg6 and 10 tbe2 offers
mixed chances (Short-Khalifman,
Merida 2001).
CARO:PANOV
Then we've landed in an Alapin
Sicilian (1 e4 c5 2 c3 d5 3 exd5
'iixd5 4 d4 tbc6 5 tbf3 cxd4?!
6 cxd4) in which the early exchange
of c-pawns allows an annoying
tbc3, e.g. 6... e5 7 tbc3 ~b4 8 ~d2
~xc3 9 ~xc3 with an edge.
More common is 4.••tbf6 5 tbc3.
A key line begins 5••• tbc6 6 ~g5 e6.
Now 7 c5 ~e7 8 ~b5 0-0
9 ~xc6! bxc6 10 tbf3 threatens
'iia4/tbf3-e5 and gives White an
ideal version ofthe Panov bind. For
example, 1O ... tbd7 11 ~xe7 'iixe7
12 0-0 e5 13 tbxe5 tbxe5 14 dxe5
'iixe5 15 l:tel 'iif6 16 'i'd2 l:tb8
17 b4 (18 ... l:txb4? 19 tbxd5), RathBrinck-Claussen, Vejle 1974.
Vasily Panov's 1 e4 c6 2 d4 d5
3 cxd5 exd5 4 c4 becomes a
Since the Panov bind began
Queen's Gambit Accepted after threatening the health of the Caro... dxc4. But it's a slightly inferior Kann more than 70 years ago,
QGA and Black should delay Black has found ways to dodge it,
... dxc4 even at the risk of White such as 6... 'i'a5, 6... ~e6 and 6... e6
creating a favorable pawn wedge 7 c5 ~e7 8 ~b5 ~d7!. That's why
with c4-c5. And on 4.•.tbc6?!:
White may prefer 6 tbfJ to 6 ~g5.
108
Semi-Open Games
Then 6... e6? 7 cS! is good for
him, e.g. 7... iJ..e7 8 iJ..b5 0-0 9 0-0
and 9 ... tLle4 10 iJ..xc6 tLlxc3 11 bxc3
bxc6 12 'iVa4. But 6 ... iJ..g4 renders
easy equality following 7 iJ..e2? e6
8 c5 tLle4.
Instead White's best answer to
6...iJ..g4 is 7 cxdS tLlxdS 8 ~3.
This leads to a somewhat superior
endgame that became famous when
Bobby Fischer beat Max Euwe with
it (8 ... iJ..xf3 9 gxf3 e6 10 'iVxb7
tLlxd4 11 iJ..b5+ tLlxb5 12 'iVc6+!
<J;;e7 13 'iVxb5 tLlxc3 14 bxc3 'iVd5
or 14 ... 'iVd7).
This is an important line because
White can reach it in other orders
and because Black's inferiority is so
slight he often cooperates.
Another Black strategy is quick
castling after 4 c4 tLlf6 S tLlc3 and
S... e6. Now on 6 iJ..gS:
White can tranpose, 6 ... tLlc6
7 cS, to that 4 c4 tLlf6 5 tLlc3 tLlc6
position considered earlier.
Better is 6... iJ..e7! and then 7 tLln
0-0. On 8 iJ..d3 it's time for
8 ... dxc4! 9 iJ..xc4 a6 because this
time it's a good QGA, e.g. 10 0-0 b5
11 iJ..b3 iJ..b7 12 l:.el tLlc6 13 a3
l1c8 or 11 iJ..d3 iJ..b7 12 'iVe2 tLlc6
13l1adl tLlb4 14 iJ..bl tLlbd5.
In place of 6 iJ..g5 White can try
6 cS iJ..e7 7 iJ..bS+. But an alert
Black will see through 7 ... tLlc6? which heads toward the bind after
8 tLlf3 - and play 7 ... iJ..d7! 8 iJ..xd7+
'iVxd7 9 tLlf3 tLlc6.
Instead of 6 iJ..g5 or 6 c5, White
should consider an exchange of
pawns on d5. But the immediate
6 cxdS tLlxdS and 7 iJ..d3 iJ..b4! has
performed well for Black.
Instead 6 tLln makes sense since
6 ... tLlc6 is once again inexact
(7 c5!) and 6 ... iJ..b4 7 iJ..d3 dxc4
8 iJ..xc4 gives White a fairly good
QGA, actually a Nimzo-Indian.
But the main point is that against
6...iJ..e7:
White can play 7 cxdS! tLlxdS
without having to defend c3
compared with 6 cxd5 tLlxd5 7 iJ..d3
iJ..b4. He obtains a slight edge with
8 iJ..d3 or 8 iJ..c4.
109
Semi-Open Games
If Black retakes with the pawn
(7 cxd5 exd5) he has the worst of a
nearly
symmetrical
position
following 8 i.b5+! i.d7 (8 ... ~c6
9 ~e5) 9 'iib3 ~c6 10 0-0 or
9 •..i.xb5 10 'iWxb5+ 'iWd7 11 ~e5.
When White does not play 4 c4
after 1 e4 c6 2 d4 d5 3 exd5 cxd5
Black's difficulties chiefly lie in his
problem piece, the QB. White
usually plays 4 i.d3, which
stops ... i.f5 and makes ... i.g4
problematic. He stands well after
4 ... ~c6 5 c3.
CARO:PANOV
ACCELERATED
With 2 c4 White tries to bring
about a Panov in which Black can't
maintain a pawn on d5, nor can he
reach the relative safety of the
Fischer-Euwe endgame.
Instead, Black often accepts a
slightly worse isolated d-pawn
middlegame, such as after 2•.. d5
3 exd5 cxd5 4 cxd5! ~f6 5 ~c3
~xd5:
Black's eyes may light up when
he sees 4 c3:
Now 6 i.c4 ~b6 7 i.b3 ~c6 8
~f3 and 9 d4, or 6 ~f3 ~c6 7 i.b5
e6 8 0-0 i.e 7 9 d4 are promising for
White. Thanks to his move order he
avoided 6 d4 ~c6 7 ~f3 i.g4!, the
This seems careless in view of endgame variation.
4•..i.f5. But the trap is sprung by
White also gets the extra options
5 'iWb3!, attacking b7 and d5. of annoying checks, 5 i.b5+ and
Following 5 ... b6? 6 i.b5+ or 5 ~a4+, which theory regards as
5 ... ~d7 6 ~f3 ~c6 7 i.b5! and difficult for Black.
~e5 Black is clearly worse.
Black can sidestep these
If Black is alert he'll realize that problems by meeting 2 c4 d5
after 4...~c6 White has nothing 3 exd5 with 3...~f6!'
better than 5 i.d3, transposing into
the main line.
Then 4 dxc6 ~xc6 is a sound
gambit, better known in the Center
110
Semi-Open Games
Counter (1 e4 d5 2 exd5 liJf6 3 c4
c6 4 dxc6 liJxc6 and then 5 d3 e5
6 liJc3 i.f5 7 liJf3 i.b4 8 i.e2 e4!).
If White stops 4 ... cxd5 with
4 'iVa4, Black can offer a different
pawn, 4•.. e6! S dxe6 i.cS! 6 exti+
'ii;xti or 6 liJf3 liJg4 7 d4 i.xd4
8 liJxd4 'iVxd4 9 'iVc2 liJa6 as in
Huebner-Luther,
Sauerbrucken
2002.
matter because White was assumed
to be better in either.
'Practice shows that White's
chances are somewhat better and
now this defense is rare,' Alexey
Suetin said of the Rubinstein in his
1983 book on the French. Svetozar
Gligoric wrote that 3 liJc3 dxe4
4 liJxe4 liJd7 is best answered by
S liJf3 liJgf6:
More exact is 3 cxdS!.
Then '6 i.g5!' reaches a good
Burn, he said, since 6 ... c5? walks
into 7 dxc5 liJxc5? 8 'iVxd8+ and
9 liJxf6.
Then 3...liJf6 4 dxc6 liJxc6 is less
sound (5 liJc3 e5 6 liJf3 i.c5
7 i.b5).
Black can avoid all of this by
meeting 2 c4 with 2... eS. He is
betting that a 1 e4 player will be
temperamentally unsuited for an
Old Indian (3 liJf3 d6 4 d4 liJbd7).
FRENCH DEFENSE
The French (1 e4 e6) variations
that most often transpose into one
another are the Rubinstein (2 d4 dS
3 liJc3 dxe4 or 3 liJd2 dxe4) and
the Burn (3 liJc3 liJf6 4 i.gS dxe4).
That used to be a 'who cares?'
But today many GMs distrust
6 i.g5 because 6... h6 and 6... i.e7
7 liJxf6+ liJxf6 8 i.d3 c5 seem to
equalize. So White began to avoid
i.g5 in a Rubinstein - and Black
discovered that the Burn was his
way to trick White into a i.g5
Rubinstein.
That is 3 liJc3 liJf6 4 i.gS dxe4
S liJxe4 liJbd7 so that 6 liJf3
transposes to the last diagram. No
better is 6 liJxf6+ liJxf6 7 liJf3 h6!
(8 i.d2 c5! or 8 i.h4 i.e7 9 i.d3 c5
10 'iVe2 'iVa5+!).
111
Semi-Open Games
To avoid the Burn, White can
play entirely different lines, such as
3 lbd2 or 3 lbc3 lbf6 4 e5. But
Black found that 3.....te7 makes
... dxe4! stronger. For instance,
3lbd2 (or 3lbc3) ..te7 4 ..td3 dxe4
S lbxe4 lbf6:
Slbce2!? and then S... cS 6 c3lbc6
7 f4 'iYb6 8 lbo.
This is the same posItIon as
3 lbd2 lbf6 4 e5 lbfd7 5 c3 c5 6 f4
lbc6 7 lbdf3 ~6 once White plays
lbe2.
Black can retake with a bishop
on f6 and obtain an excellent
Rubinstein, e.g. 6 lbxf6+ ..txf6
7lbo lbc6! 8 c3 eS.
Or 6lbolbbd7 and 7 'iWe2lbxe4
8 ..txe4 cS! and then 9 0-0 cxd4
10 l:.dl lbc5 11 lbxd4 lbxe4
12 'iWxe4 0-0 (Becerra-Kaminski,
internet 2005).
Black managed to reach a
position that could come about after
3 lbc3 dxe4 4 lbxe4 lbd7 5 lbf3
lbgf6 6 ..td3 ..te7 but only if White
rejects 6lbxf6+!.
FRENCH: STEINITZ
In this way White dodges other
Tarrasch lines. When Vishy Anand
wanted to reach the diagram in the
2000 FIDE championship finals he
knew that if he chose 3 lbd2, his
opponent, Alexei Shirov, wouldn't
cooperate by playing 3... lbf6
because he preferred other moves.
So Anand chose 3 lbc3!, reached
the Tarrasch position he wanted and
won.
Quite a different situation arises
after 3 lbc3 lbf6 4 eSlbfd7 slbo.
White is ready to liquidate his
center to obtain tactical chances,
such as S••• cS 6 dxcS lbc6 7 ..tf4
..txcS 8 ..td3 and 8... 0-07 9..txh7+.
A more open French results, e.g.
The Steinitz Variation has
become a favorite way of 8 ... f6 9 exf6 lbxf6 10 'iWe2 0-0
transposing into a sharp line of the 11 0-0-0 a6 12lbe5lbb4 13 ..tg5 b5
Tarrasch, 3 lbc3 lbf6 4 eS lbfd7 14 a3 lbxd3+ 15 lIxd3 'iWc7 16 f4
112
Semi-Open Games
as in Berg-Akesson, Gothenburg
2006.
7 d4, or try for more with 6 dxc3!?,
e.g. 6... c5 7 h4 tbbc6 8 h5 h6 9 i.f4
'ikc7 10 'ii'd2 as in Rogers-Tondivar,
Hoogoveen 2006.
If White likes this he can reach it
while discouraging the Rubinstein
and perhaps the Winawer as well.
He begins with 1 e4 e6 2 tbf3!?
FRENCH: WINAWER
As noted in the Introduction
Black can sidestep many of the rare
Winawer (3 tbc3 i.b4) lines by
meeting 4 e5 with 4...tbe7 5 a3
i.xc3+ 6 bxc3 c5.
Francophiles regard that with
contempt and will want more than
2•.. d5 3 tbc3 dxe4. And they'll see
that Black is a bit worse after 3... d4
4 ttJe2 c5 5 c3.
So they'll play 3 ...tbf6 and then
4 e5 ttJfd7 5 d4 gets White to where
he wants to be. And on the natural
3... c5
The 'something to think about'
quality of 4 ... tbe7 often prompts
responses like 5 tbe2? c5 6 a3
i.xc3+ 7 tbxc3?, which favors
Black after 7... cxd4 8 'ii'xd4 tbbc6
9 'ikg4 0-0 10 'fib5 d4!, ByrneUhlmann, Reykjavik 1968.
And on 5 tbf3:
... then 4 exd5 exd5 5 d4! traps
Black in one of Frank Marshall's
lesser ideas (1 e4 e6 2 d4 d5 3 tbc3
c5 and 4 exd5 exd5 5 tbf3.)
What should Black do after
2 ttJf3 d5 3 tbc3 if he doesn't like
3... ttJf6 ? His best bet may be
3.....tb4.
It's unclear whether White
should transpose to a Winawer, with
4 e5 ttJe7 5 a3 i.xc3 6 bxc3 and
Black can transpose (5 ... c5) into
a normal Winawer. But also
attractive is 5... b6, since White has
missed his 'ikg4 opportunity,
compared with 5 a3 i.xc3+ 6 bxc3
b6 7 'ikg4!.
Chances should be equal after
5... b6 6 i.d2 i.xc3!? 7 i.xc3 i.a6
113
Semi-Open Games
8 ..txa6 ttJxa6 9 0-0 c5 (BokrosRustemov, Kircheim 2006).
The counter-finesse to 4 ... ttJe7 is
S ..td3. Then S... cS 6 a3 ..txc3+
7 bxc3 seems to be a good version
for Black of a main line Winawer in
view of White's prematurely
developed bishop (7 ... ttJbc6 8 ttJf3
c4 or just 7... c4 and ... ..td7-a4).
3 ttJc3 i.b4 4 e5 c5 5 a3 i.xc3+
6 bxc3 ttJe7 7 dxc5) if White had an
extra tempo.
But White improves with 6 dxcS!
ttJbc6 7 ttJO. He's transposed into
the 4 ... c5 5 dxc5 variation made
famous by Fine-Botvinnik, AVRO
But is dxc5 a good extra move?
1938. White's compensation is still
That
depends on positions like
rated highly after 7•.. d4 8 a3 i.aS
9 b4 ttJxb4 10 axb4 ..txb4 in view 8 'i'g4 ttJd7 9 ttJO 0-0 10 i.d3.
of 11 O-O!, rather than Fine's
The thematic 9••.'i'c7 10 'i'xg7!
11 ..tb5+.
:tg8 11 'i'xh7 ttJxeS favors White
This is significant because today after 12 'i'h5 ttJxf3+ 13 'i'xf3 i.d7
White rarely gets there via Fine's 14 i.f4, Shamkovich-Gipslis,
4 e5 c5 5 dxc5 because theory urges Soviet Championship 1961.
5.. .'iic7!? or 5... ttJc6 6 ttJf3 d4!.
In the main line Winawer, 4...cS
S
a3 i.xc3+ 6 bxc3 ttJe7, the
A major virtue of 4 ... cS over
4 ... ttJe7 is that it gives Black an choices are 7 ttJf3, 7 a4 and 7 h4.
extra option, the Armenian S a3 But Oleg Romanishin has made
good use of7 ..td3!?
..taS. Then instead of the main line,
6 b4, the trickster may prefer
This often transposes to the
6 dxcS!? This threatens 7 b4 and others but for the moment it retains
transposes into a superior version of the option of 'i'g4. For example,
Fine's line after 6... d4 7 b4 dxc3 7 ... 0-0 8 'i'g4 gets White to a
8 'iVg4!.
popular 7 'i'g4 0-0 8 ..td3 position
without having to deal with 7 'i'g4
Perhaps best is 6.....txc3+ 7 bxc3
'i'c7 and 7 ... cxd4.
ttJe7. The result is a strange position
that could come about from a
For example, 7 i.d3 ttJbc6
normal Winawer (1 e4 e6 2 d4 d5 8 'i'g4 c4 9 ..te2 ttJf5 10 a4 'i'a5
114
Semi-Open Games
We've transposed into a version
of Viktor Korchnoi's gambit
(3 4Jd2 4Jf6 4 e5 4Jfd7 5 i.d3 c5
6 c3 4Jc6 7 4Jgf3 'iVb6 8 0-0) in
which Black plays 8 ... i.e7 instead
of grabbing the pawn.
11 i.d2 i.d7 12 'iWh5 h6 13 i.g4 as
in Bagheri-Apicella, Chartres 2005.
The bishop move was once
thought dubious but now it's a valid
alternative, e.g. 9 dxeS 4JxeS
10 i.e2 4Jd7 11 l:te1 ~e7 12 ~e2
gS 13 ~a4 4JeS!, KabanovThe main drawback to 7 i.d3 is
7•.. b6. Why? Because often when
Black tries to trade his bad bishop
with ...b6/ ... i.a6, White's best way
to cross him up is i.b5+ as in 7 4JO
b6 8 i.b5+! i.d7 9 i.d3!. But after
spending a tempo on 7 i.d3 there is
little reason to play 7... b6 8 i.b5+
i.d7 9 i.d3.
FRENCH: TARRASCH
In the Tarrasch Variation (3 4Jd2)
Black can use 3•••i.e7 as a waiting
move. After 4 4Jgf3 4Jf6 S eS 4Jfd7
White has nothing better than 6 e3
eS 7 i.d3 4Je6 8 0-0 'iib6.
Kravstov, Tomsk 2003. In other
words, 3 ...i.e7 can lure White into
a gambit he never intended.
If White likes the gambit he can
try to reach it from other orders. For
example, 3... c5 seems to rule it out.
But 4 4Jgf3 and then 4 ... 4Jf6 5 e5
4Jfd7 6 c3 transposes. If Black
wants to sidestep the gambit he has
other good lines, including 4 ... 4Jc6
and 4 ... cxd4.
There is also 3... 4Jc6 and 4 4Jgf3
4Jf65 e5 4Jd7 6 i.d3. Then 6... 4Jb4
is natural and after 7 i.e2 c5 8 c3
4Jc6 9 ~d3 'iVb6 10 0-0 we've
reached the gambit again - although
this time White may do better with
9 0-0 and 9 ... cxd4 10 cxd4 'iVb6
11 4Jb3.
In the sharpest Tarrasch line,
3...4Jf6 4 eS 4Jfd7, White usually
selects 5 f4 or 5 i.d3. If he prefers
the latter, he may profit by delaying
i.d3 with S e3.
115
Semi-Open Games
3 ... a6 4 lLIgf3 c5 S exdS exdS.
Which is best for him?
The drawback to the first order is
that White can check at move five.
Both S c3 and S il.d3 reach a
main line after S ... cS. However
Black has an alternative strategy in
S...b6 and 6 ... il.a6, to trade his bad
bishop.
That makes S c3 more accurate
because if Black pursues the Btrading plan White will be a tempo
ahead - S c3 b6 6 'iig4 (or 6 lLIh3)
il.a6 7 il.xa6 - of S i..d3 b6 6 c3
il.a6 7 il.xa6.
After 3•.•cS the major move order
questions revolve around il.bS+ and
the timing of exdS. Black has been
doing well recently when he stops
the check, e.g. 4 exdS exdS S lLIgf3
a6.
He can also reach this position
via 4 lLIgf3 a6 S exdS exdS and
The drawback to 3 ... a6 is 4 eS.
Some French players don't relish
the prospect of shifting to an
Advance Variation (4 ... cS S c3) in
which they've played ... a6 in place
of the usual Advance moves such as
... 'iib6 and ... lLIc6.
In the third order, 3... cS 4 lLIgf3
a6, Black has to deal with S dxcS
il.xcs 6 il.d3, which gives White a
nice setup after 7 0-0 and 8 a3/9 h4.
So it's up to Black to rate the
worst-case scenarios in all three
orders and pick the best of them.
But how bad is that check? One
school of thought says a trade of
light-squared bishops is good for
White and that 3 ... cS 4 exdS exdS
S il.bS+ must be met by S... lLIc6.
But Lev Psakhis insisted S... il.d7
is fine, arguing that a bishop trade
helps Black. Max Euwe agreed and
said S...lLIc6 6 lLIgf3 was so good
for White that he should try to trick
Black into that position with
S lLIgf3 and ifS ... lLIc6, then 6 il.bS!.
If Euwe is right about the pin,
White may do better with another
order, 3 •.• cS 4 lLIgf3 so that he can
meet 4 •••lLIc6 with S exdS exdS
6 il.bS. The drawback is giving
Black options such as 4 ... lLIf6
S exdS lLIxdS.
116
Semi-Open Games
But if White reaches his goal, the
most common reply is 6... i.d6.
A trade of pawns is coming,
either on c5 or d4. Two lines that
often transpose are 7 dxe5 i.xe5
8 0-0 and 7 0-0 liJe7 8 dxe5.
Mikhail Botvinnik called the
second order inaccurate because
Black can improve on 7 ... liJe7 with
7 ... cxd4!. Then 8 liJb3
9 liJbxd4 0-0 is equal.
liJe7
What has happened is we've
transposed into 7 dxc5 i.xc5 8 0-0
liJe7 9 liJb3 i.d6 10 liJbd4 0-0.
But in that order White has more
options such as 10 .l:le 1, 10 i.g5
and 10 i.xc6+ bxc6 11 'ifd4. The
7 0-0 cxd4! order denies him the
choice.
FRENCH: ADVANCE
VARIATION
The Advance Variation tabia
used to be 3 e5 e5 4 e3 liJe6 5 liJf3
'ifb6 6 i.d3:
Following a Howard Staunton
game in 1841, 6 ...i.d7 was
considered best. But this was shown
to be inexact in Nimzovich's
celebrated victory over Georg
Sa1we at Carlsbad 1911, when
7 dxe5! i.xe5 8 0-0 and b2-b4
Black's
seriously
hampered
development.
Black improves with 6 ... exd4!
7 exd4 i.d7. Then White has no
convenient defense of d4 and the
8 0-0 liJxd4 gambit is regarded as
not quite sound.
These two basic positions, the
Nirnzovich-Salwe one and the other
with 6 ... cxd4, influence a lot of3 e5
theory. For example, you may see
4 .. .'iVb6, instead of 4 ... liJc6, and
then 5 liJf3 i.d7.
Black retains the option of
... i.b5, to get rid of his bad B.
White has two strange but possibly
good replies, 6 i.d3 i.b5 7 i.e2!?
(7 ... liJc6 8 i.e3) and the devilish
7 dxe5 i.xe5 8 b4!?, based on
trapping the bishop after 8 ... i.xf2+
9 'ite2.
117
Semi-Open Games
7 .. :iixb2? with 8 ..te3! 'ifxal
9 'ifc2.
Black might avoid both replies
by meeting 6 ..td3 with 6...tiJc6?
But then 7 dxc5! puts us back in
Nimzovich-Salwe.
Then 9... cxd4 10 cxd4? is bad
because of 10... ..tb4+ and the queen
can never be trapped.
He can try to repair matters with
6... cxd4 7 cxd4? tiJc6, reaching the
Black-friendly position.
But 10 tiJxd4! gives White a big
edge after 10 ... tiJxd4 11 ..txd4..ta3
12 ..tb5+ ~d8 13 0-0 or 1O ... il.a3
11 tiJb5!.
But thanks to his delay in ... tiJc6
White can do better with 7 tiJxd4!,
e.g. 7... tiJe7 8 tiJd2 tiJbc6 9 tiJxc6
and ife2/tiJf3 with a slight edge.
Therefore the proper order is
6... cxd4 7 cxd4 tiJh6, reaching the
last diagram. White should not play
8 ..txh6 because he is transposing
into the poor 10 cxd4 line. Instead,
he should go into one of the
mediocre alternatives such as
8 tiJc3 or 8 b3 .
Because of 6 ... cxd4! in the last
diagram, 6 ..te2 replaced 6 ..td3.
Black's most common reply is
6 ... cxd4 7 cxd4 tiJge7. He has no
major problems in lines that run
8 tiJc3 tiJf5 9 tiJa4 'ifa5+ 10 ..td2
..tb4 11 ..tc3 b5! or 8 b3 tiJf5 9..tb2
..tb4+.
White's best may be 8 tiJa3 tiJf5
9 tiJc2 ..tb4+ 10 ~fl. But Black can
avoid that if he reaches this:
David Bronstein called 5•..il.d7
a 'clever waiting move after 3 e5
c5 4 c3 tiJc6 5 tiJf3. White must
reveal his intentions with 6 a3,
6 tiJa3 or a bishop move.
Then 8 tiJa3 allows 8.....txa3!
9 bxa3 tiJf5. But how does Black
get here?
The wait pays off after 6 tiJa3?
cxd4 7 cxd4 ..txa3! 8 bxa3 'ifa5+
with advantage.
The problem with 6 ..te2 tiJh6 is
7 .iLxh6!. White can punish
Another benefit is that by
delaying ... 'iib6 Black ensures that
118
Semi-Open Games
dxc5 will not be forcing. That can
be a problem after 5.. .'iVb6 because
6...lbge7? allows 7 dxc5! 'iVxc5
8 .ltd3, which favors White.
for a
as 7
9 0-0
with
bind.
knight in 6... c4
liJbd2 liJa5 8
h6 10 :e1 liJc8!
prospects for a
lines such
..te2 liJe7
and ... liJb6
queenside
But in the diagram 6 ..te2 liJge7
Finally, if White feels that
7 dxc5 liJg6! is excellent for Black
and so is 7 liJa3 cxd4 8 cxd4 liJf5 5... i.d7 means there's no pressure
9 0-0 i.xa3 10 bxa3 0-0 (Sax- on d4, he may play the natural
Dreev, Tilburg 1992).
6..td3.
If White opts for 6 a3, which is
likely his best, Black benefits from
his move order because b6 is free
But then 6... cxd4 7 cxd4 'iYb6!
tricks him into the bad 5.. .'~'b6 6
i.d3 cxd4! position one more time.
119
Chapter Five:
Double QP Openings
But Black had better be ready for
Most 1 d4 d5 games become
Queen's
Gambits
after an 3 c4.
immediate or delayed c2-c4. The
reason White would delay is to rule
out certain 2 c4 defenses and to
weigh the option of a Queen's Pawn
Game once he sees Black's next
few moves.
With 1 d4 d5 2 lbf3 he denies
Black a few sharp lines such
as the Albin Counter Gambit,
offbeat Slavs and Queen's Gambit
Accepteds like 1 d4 d5 2 c4 dxc4
3 lbf3 a6.
Black can try to transpose into
the latter with 2... a6, hoping for
3 c4 dxc4. But White can make ... a6
look like a wasted move. e.g. 3 i.f4
lbf6 4 e3 e6 5 i.d3 c5 6 c3 i.d6
7 0-0 lbc6 8 dxc5 i.xc5 9 lbbd2
i.d6 10 i.g3 0-0 11 e4 with a plus,
Vladimirov-Iuldachev, Mumbai
2003.
This obvious move allowed a
minor Swiss master, Hans Fahrni,
to confound elite players a century
ago. Fahrni-Spielmann, Barmen
1905 went 3...lbc6? 4 cxd5 'i'xd5
5 lbc3 'i'd8 6 d5 lbb8 7 e4 with
obvious superiority. Even the
natural 3...lbf6?! and 3... cxd4?
favor White after 4 cxd5!.
Of course, there is nothing wrong
with
Black's game after 3... dxc4! or
But 2lbf3 doesn't commit White
to a QPG because he can shift gears 3 ... e6!. But to play those moves he
with 3 c4. Suppose 2... c5, a good would need to know a lot about the
move if White heads into something Queen's Gambit Accepted (3 ... dxc4
like 3 c3 lbf6 4 i.f4 lbc6 5 e3 4 e3 or 4 d5) or Tarrasch Defense
'ifb6!.
QGD (3 ... e6 4 cxd5 exd5).
120
Double QP Openings
If there are so many pitfalls for
Black, why isn't 2 ttJf3 played by
every d-pawn player? The answer is
White deprives himself of several
options and gives Black some
useful ones:
but landing in another opening, one
he's never played or simply doesn't
like.
How can that happen after such
quiet moves? Well, suppose Black
replies 3.•. cS. He had better be
prepared
for a Semi-Tarrasch (4 c4
White rules out the QGA lines of
1 d4 d5 2 c4 dxc4 3 e4 and 3 e3, as e6) or a Queen's Gambit Accepted
well as QGD variations with ttJge2. (4 ... dxc4).
He gives away his best chances of
And if he prefers 3••. g6, he needs
crushing the Tchigorin Defense to know a bit about the Gruenfeld in
QGD (2 c4 ttJc6 3 cxd5 or 3 ttJc3) view of 4 c4!? Yefim Geller, who
and the Baltic Defense QGD (2 c4 rarely played the Gruenfeld, once
discovered as Black that 4.•• it.g7
.tf5 3 cxd5 or 3 ttJc3).
S cxdS ttJxdS 6 it.e2 cS isn't an
At the same time he grants Black
easy position to handle.
extra options such as the Ragozin
and Semi-Tarrasch QGD variations,
which are not possible after 2 c4
unless White cooperates. That's a
lot of plusses and minuses to total
up.
Note that 1 ttJf3 dS 2 d4 reaches
the same position as 1 d4 d5 2 ttJf3.
This can be exasperating to an
aggressive Black who likes to
unbalance the position quickly with
1 ttJf3 d5 2 c4 d4 or 1 d4 ttJf6. That
was the case when Lajos Portisch,
an Indian specialist, faced a young
Anatoly Karpov, who began 1 ttJf3.
Portisch found himself in a
relatively unfamiliar Slav Defense
after 1...d5 2 d4! and lost.
COLLE SYSTEM
Black's greatest danger after 1 d4
dS 2 ttJf3 ttJf6 3 e3 isn't the Colle
Soon after 7 e4! Geller was in
muddy waters (7 ... ttJb6 8 d5 0-0
9 a4) and eventually lost.
The other major peril for Black
is a switch from the Colle,
characterized by c2-c3, to the
Colle's sibling, the Zukertort
system, characterized by b2-b3 and
J..b2. This is a problem because the
best defenses against a Colle often
involve ... ttJbd7 - which can be
dubious in the Zukertort.
121
Double QP Openings
Here's how: After 1 d4 d5 2 ttJf3
ttJf6 3 e3 e6 4 iLd3 c5 5 c3 ttJc6
6 ttJbd2 White usually answers
6 ...iLe7 or 6 ... iLd6 with 7 0-0 0-0
8 dxc5! iLxc5 9 e4, with a good
game. But if Black develops his QN
at d7:
This time 8 dxc5? is met by
8... ttJxc5!. White's bishop is
attacked and he can't carry out the
vital e3-e4 break. He's already a bit
worse.
But in the Zukertort, 1 d4 d5 2
ttJf3 ttJf6 3 e3 e6 4 iLd3 c5 5 b3,
Black stands well following
5...ttJc6! 6 iLb2 iLe7, e.g. 7 0-0 0-0
8 ttJbd2 'ilc7 9 ttJe5 ttJd7!.
The danger to both players is
tipping their hand. If Black had
played ... ttJbd7 instead of ... ttJc6, he
would be worse after 9 ttJe5!, e.g.
9... b6 10 f4 iLb7 11 c4 ttJe4 12 cxd5
exd5 13 ttJxe4 dxe4 14 iLc4 as in
I. Rabinovich-Makogonov, Soviet
Championship 1937.
What this means is Black's QN
should be watching White's QB and
vice versa. On 1 d4 d5 2 ttJf3 ttJf6
3 e3 e6 4 il.d3 it is inexact to play
4.•.ttJbd7:
White hasn't chosen between
Colle and Zukertort yet, so 5 0-0 c5
6 b3! gives him somewhat the better
chances. More precise is 4 ... c5.
Then 5 b3 and 5 c3 commits White,
and Black can respond 5... ttJc6!
and 5... ttJbd7! respectively.
The crafty response to 4.••c5 is
5 0-0, which was played by Jose
Capablanca and Akiba Rubinstein.
They waited for a Black
commitment, 5... ttJbd7 6 b3! or
5... ttJc6 6 c3!.
But they weren't tested by
5•••c4!, e.g. 6 iLe2 b5 7 b3 il.b7
8 a4 a6 9 axb5 axb5 10 ':xa8
iLxa8 11 bxc4 bxc4 12 c3 ttJbd7
with equality, Inkiov-Groszpeter,
Copenhagen 1988.
OTHER QUEEN PAWN
GAMES
The Trompowsky Attack (1 d4
ttJf6 2 iLg5) has a poor cousin in
122
Double QP Openings
1 d4 dS 2 i.gS. The lack of a i.xf6
threat makes Black's life easier, and
books say 2 ... h6 3 i.h4 c6 and
... 'iib6 is good.
edge of a symmetrical position after
4 ..tf4. The easiest way to equalize
against 3 c3 is likely 3.....tf5 4 "iVb3
'iic8 .
If Black doesn't like that he
should beware of c2-c4 traps. One
arises when he avoids doubled
pawns with 2...ttJd7 and 3... ttJgf6.
Then 3 c4!.
ALBIN COUNTER GAMBIT
The main line of the Albin begins
with 1 d4 dS 2 c4 eS 3 dxeS d4
4 ttJf3 tbc6. White characteristically fianchettoes his KB and
plays tbbd2, 0-0, 'i'a4 and a2-a3.
Black, meanwhile, develops his
QB, usually at e6 or g4 and then
decides between ...'iid7/ ... 0-0-0 and
the alternative ... ttJge7/ ... 0-0. Albin
aficionados have tried and failed to
establish which QB move is correct.
They should consider meeting S
ttJbd2 with S•.. ..tfS!?
Now 3••• dxc4 4 ttJc3 is an
inferior QGA and 3..• c6 is a dubious
Slav, 4 cxdS! cxdS S ttJc3, because
his QN should be at c6.
For example, 5... ttJgf6 6 e3 g6
7 ttJf3 i.g7 8 i.d3 0-0 9 0-0 b6
10 .:tel i.b7 11 'iib3 with
advantage in Skembris-Vasovsky,
Skopje 2002.
The Torre Attack is 1 d4 ttJf6
2 ttJf3 e6 3 i.gS. Normally Black
can avoid it simply with 1...d5 since
2 ttJf3 ttJf6 3 i.g5 allows him quick
equality with 3... ttJe4!.
But there is a sneaky waiting
move, 3 c3. Then on 3... e6 4 i.g5
an unwitting Black has been
dragged into the Torre. If 3... c6
instead White has the usual slight
Unlike 5... i.e6 and 5... i.g4, this
makes a threat, 6... ttJb4. That wins
time for queenside castling after
6 a3 'iid7, under better circumstances than usual.
The test of this is order is 7 b4
i.xb4!? (8 axb4 ttJxb4 9 "tWa4
ttJc2+ 10 ~dl ttJxal 11 'iixal c5).
Declining the sack is harmless,
123
Double QP Openings
8 'ilVb3 ..ta5 9 e3 l::td8, SpaisKatsaris, Athens 1999.
1990s, White did well when he
reached them via the Exchange
What happens after 7 g3 0-0-0
8 ..tg2 is virtually unexplored, e.g.
French or Petroff, so Black has a
real decision at move three.
8... d3 9 e3 ..th3 10 ..txh3 "ifxh3
(11 b4 ..txb4!). White avoids this
with S g3! and if S.••..te6 then
6lbbd2.
White makes it somewhat easier
for him when he chooses 3 lbc3
because after 3... lbf6 he has
nothing better than 4 e3, heading
towards main QGA lines. But in
those lines, lbc3 is better delayed,
as we'll see shortly.
QUEEN'S GAMBIT
ACCEPTED
After Joseph Blackburne popularized 3 lbf3, following 1 d4 dS 2
c4 dxc4, White's alternatives
almost vanished. But they're
reappearing, mainly as transpo
devices. Garry Kasparov played
3 e3, for example, when he wanted
to avoid 3 lbf3 c6!? 4 e4 b5.
Another story entirely is the
immediate 3 e4. It was once
considered dubious because of
3 ... e5, since White cannot maintain
a pawn on d4. But 4 lbO! revived
this line and poses a move order
problem.
Both 3 e3 and 3 lbc3 get Black
thinking about whether to transpose
into main lines with 3...lbf6 or to
try for quicker equality with 3... e5.
For example, 3 lbc3 e5 4 dxe5?
'iVxdl + is excellent. The key line is
3 lbc3 eS 4 e3 exd4 S exd4 lbf6
6..txc4.
A controversial position arises
after 4 ... ..tb4+ 5 lbbd2 exd4
6 ..txc4 and by transposition from
4 ... exd4 5 ..txc4 ..tb4+ 6 lbbd2.
Which is best for Black?
Theory used to call such
positions boringly even. But in the
Around 1980 the answer was
4 .....tb4+ because 5 lbbd2 is
punished by 5... c3! and S i..d2
i..xd2+ 6 'i'xd2 exd4 peters out to
124
Double QP Openings
equality after 7 'iVxd4 'iVxd4 or
7 l2Jxd4 'iVe7 8 f3 l2Jf6 9 iLxc4 0-0
and ...:d8.
But if Black waits to check,
4..• exd4 5 iLxe4 iLb4+, then White
had a choice between 6 iLd2 and
the more aggressive 6 4:Jbd2!, the
thinking went.
Black can sidestep that with
4..• e6 5 e4 iLb4. This an invitation
to a Vienna QGD (6 iLgS! cS!) and
one that should be accepted because
6 eS 4:JdS 7 iLd2 l2Jb6! makes it
hard for White to regain his pawn.
Within 10 years, theory reversed
itself and declared 4...iLb4+ to be
second-best because White was
getting good results from 5l2Je3!.
Another Black option after
4 l2Jc3 is 4•..l2Je6. Then S e4 Jt.g4 is
one of the better lines of the
Tchigorin QGD. It's as if White
played I d4 dS 2 c4 l2Jc6 3 4:Jc3
dxc4 4 l2Jf3 l2Jf6 S e4 iLg4 rather
than 4 dS!?
Today neither order is clearly
best. It's a matter of which position
Black most wants to avoid,
4 ... iLb4+ S l2Jc3! or 4 ... exd4
S iLxc4 iLb4+ 6l2Jbd2!.
But if Black played 2 ... dxc4 to
stay in the QGA, he should consider
the three alternatives to 3... l2Jf6.
They not only avoid the gambit but
speed his counterplay.
QGA: 3l2Jf3
The first is 3 ...e5. After 4 e3
Black can safely head towards a
QGA tabia with 4 ... l2Jf6 S iLxc4 e6.
At any tournament you're likely
to see Black rattle off his first half
dozen moves in a QGA as if they
didn't matter. But his third move is
significant. The popular choice,
3.•.l2Jf6, allows White to offer a
gambit, 4 l2Je3 and S e4, based on
attacking the knight with e4-eS.
He also get an extra option,
4.•• exd4 with the idea of keeping
the pawn, S exd4 iLe6!? The key
line is 5 iLxe4 "ike7 (not S... dxe3??
6 iLxf7+):
This was considered refuted by
6 'iVb3 because 6... iLe6? is met by
125
Double QP Openings
7 .t.xe6! 'ii'xc 1+ 8 \t>e2 with a
fierce, probably winning, attack.
However QGA and QGD lines
are replete with 'little' moves that
make a big difference, such as
6... e6!. Then 7 lLlxd4 a6 is equal
and 7 exd4 lLlc6 8 0-0 a6 reaches a
more normal QGA position but with
the addition of the queen moves.
That helps Black because he
threatens 9 ... lLla5! and gets a fine
position after 9 'tid! lLlf6, e.g.
10 lLlc3 .t.e7 11 d5 ttJa5! or 11 'tie2
0-0 12 l:tdllLlb4 as in WojtkiewiczYermolinsky, San Francisco 2002.
But that helps Black. He can
transpose into Rubinstein's a2-a4
line of the main QGA, 4...lLlf6 5 e3
e6 6 .t.xc4 c5, which is currently
considered too quiet for an edge.
Or he can try 5•...t.g4! 6 .t.xc4 e6
7lLlc3,
a superior version of the pin
That's two plusses for 3... c5!? avoiding the 3 .. .ti:Jf6 4 lLlc3/5 e4 variation. In the original pin order,
gambit and offering Black a good 3 lLlf3 lLlf6 4 e3 .t.g4, White gets
alternative to the QGA tabia. The the edge from 5 .t.xc4 e6 6 lLlc3
minus is 4 d5!? and then 4..• e6 particularly after 6... lLlc6 7 .t.b5!.
But that's not possible in this
5 lLlc3 lLlf6 6 e4, which is unclear.
order and 7... lLlc6 Black either
The second alternative is 3••. a6,
equalizes with ... e5 or does well
which often transposes into the
in complications such as 8 h3 .t.h5
third, 3... e6. It was considered an
9 0-0 .t.d6 10 .t.e2 0-0 11 g4 .t.g6
important finesse - to discourage
12 lLlh4 lLlb4, Hillarp Persson4 'iia4+ with 4 ... b5! - back in the
C. Hansen, Malmo 2003.
1930s when White was getting a
If there is a serious drawback to
plus from 3 ... lLlf6 4 'tWa4+.
3... a6 (and 3... e6) it is 4 e4!? when
Today the check seems harmless
4•..b5 5 a4 c6, as Kasparov has
and 3... a6 is usually just a prelude
played, is double-edged.
to 4 e3 lLlf6 5 .t.xc4 e6, another
tabia.
QGA: 3•••lLlf6
There is also a risky extra option,
4 e3 b5, as well as Capablanca's
The chief merit of3 ... lLlf6 is that
4 e3 e6 5 .txc4 b5!? and 6... i.b7. it stops 4 e4 and leads in the vast
That's often enough to frighten majority of games to a tabia after
White into 4 a4?!.
4 e3 e6 5 .t.xc4. The inevitable ... c5
126
Double QP Openings
allows Black to isolate White's dAnother version of the delayed
pawn (... cxd4/exd4) or expand on ... c5 is 1 d4 d5 2 c4 dxc4 3 ttJO
the queens ide with ... a6/ ... b5/... c4.
ttJf6 4 e3 a6 5 ~xc4 b5, a
But this also gives White a Capablanca idea.
chance for dxc5, possibly followed
by a trade of queens into a slightly
favorable endgame. If Black is the
higher-rated player he can avert an
endgame by delaying ... c5, e.g.
5.•. a6 6 0-0 ttJbd7.
Then 7 'ii'e2 c5 8 dxc5 ~xc5 is
harmless and Black can even get the
edge after 7 a3?! b5 8 il.a2 il.b7
On the routine 6 ~b3~b7 7 0-0
e6 8 'ii'e2 Black can reach a roughly
equal main line position with
8... ttJbd7 9 l:.dl c5. This order is a
bit of a bluff because it's based on
White heeding a QGA rule of
thumb that says when the bishop is
attacked by ... b5, it's almost always
best to retreat to b3.
9 'i'e2 c5 10 dxc5 ttJxc5!.
This is an exception: 6 ~d3! and
7 a4! favorably probes Black's
The problem with this order is
queenside, e.g. 6 ... ~b7 7 a4! b4
7 a4!, creating a better-than-usual 8 ttJbd2 ttJbd7 9 0-0 e6 10 'ii'e2 c5
version of the Rubinstein line. 11 e4 cxd4 12 e5 ttJd5 13 ttJb3 ttJc5
Unlike the usual Rubinstein Black 14 ~g5! 'ii'd7 15 ttJfxd4, Topalovcannot develop his QN on c6, its Narcisco Dublan, Barcelona 2000.
best square once the hole at b4
appears. He doesn't quite equalize
White's strategy works because
after 7... c5 8 ttJc3 cxd4 9 exd4 Ji.e7, his knight was at bI. If he violates
e.g. 10 'i'e2 0-0 11 il.g5 ttJb6 another rule of thumb and brings it
12 il.b3 ttJbd5 13 ':'fel and ttJe5 as out early, we get a position like
in Lechtynsky-Kantorik, Plzen 4... e6 5 Ji.xc4 c5 6 ttJc3 and 6... a6!
2003.
7 0-0 b5:
127
Double QP Openings
Experience with 8 i.d3 lLlbd7
shows that Black has an easy time,
e.g. 9 'iVe2 i.b7 10 l:dl 'iib6
11 i.c2 l:d8 12 a4 b4 13lLlbl 'iic7
14 i.d3 cxd4 15 exd4 a5, BenitahFressinet, Paris 2004.
Now on 6 0-0 a6 7 dxc5 Black's
best is 7.. :ifxdl and a draw is the
typical result of 8 l:xdl i.xc5.
But 6 'iie2 makes dxc5 more of a
threat, since it wouldn't allow a
trade of queens. Then 6.•. cxd4
What happened: White was 7 exd4 lLlc6 avoids that and tries to
tricked into a toothless line of punish White for endangering the
Meran Variation of the Semi-Slav. d4-pawn.
Compare it with 1 d4 d5 2 c4 c6
But another QGA rule of thumb
3 lLlf3 lLlf6 4 lLlc3 e6 5 e3 lLlbd7
6 i.d3 dxc4 7 i.xc4 b5 8 i.d3 a6 says Black shouldn't exchange on
when he plays 9 O-O?! c5! 10 'iVe2 d4, as Wilhelm Steinitz habitually
did, until he can safely threaten
i.b7, not 9 e4!.
the d-pawn or obtain other
White improves in the diagram compensation. In this case,
with 8 i.b3! when 8... i.b7 9 'iie2 accepting the gambit is risky, 80-0!
lLlbd7 10 l:dl 'iib8 or 10 ...'iVc7 are lLlxd4 9lLlxd4 'iixd4 10 l:dl 'iib6
main QGA lines that have been 11 i.b5+ i.d7 12 lLlc3, Pelletiertested for decades without finding a
Arencibia, Obeda 1998.
significant edge.
But declining it, 8... i.e7 9 lLlc3
Attention has turned instead to an 0-0 10 i.e3, transposes into the
early 'iie2. Books used to say that somewhat discredited Steinitz
the order of 0-0 and 'iie2 didn't
Variation. As a result 6... cxd4 looks
matter because - you guessed it suspect and there is no consensus
they 'just transpose.' But it does
on what Black should do instead
matter because of the possibility
after 6 'iie2.
of dxc5 and e3-e4-e5, Semyon
There are also good and
Furman's plan. After 1 d4 d5 2 c4
dxc4 3 lLlf3 lLlf6 4 e3 e6 5 i.xc4 c5: bad transpositions when White
128
Double QP Openings
anticipates ... b5 with a bishop
move, 6 0-0 a6 7 .i.b3.
Now 7...b5 8 a4! corrupts the
Black queenside, e.g. 8... b4 9 tLlbd2
and 8....i.b7 9 axb5 axb5 10 lIxa8
.i.xa8 11 'iie2, 11 tLla3 or 11 tLlc3.
The reason they often transpose
is that 8 tLle3 exd4 does not win a
pawn. After 9lIdl! Black has been
trapped in another Steinitz (9 ....i.e7
10 exd4 0-0 11 d5! exd5 12 tLlxd5
tLlxd5 13 .i.xd5 'iic7 14 'iie4) .
Max Euwe awarded 8 tLlc3 an
Moreover on 7... .i.e7 White
exclamation point for a different
plays 8 'iie2 and prepares to get a
reason. White can preserve his
good version of Furman's line with
bishop, after 8 tLle3 b5 9 .i.b3 .i.b7
9 dxe5!.
10 lIdl e4?!, with 11 .i.e2 tLlb4
If Black stops that with 8... cxd4 12 .i.bl!. This favors him when he
he transposes into another dubious gets to push his e-pawn, e.g.
Steinitz, 9 lIdl tLlc6 10 exd4 0-0 12 ... .i.e7 13 e4 tLld3 14 .i.xd3 cxd3
11 tLlc3. The same goes for 7... tLlc6 15 'i'xd3 b4 16 e5! as in
8 'iie2 cxd4 9 lId!.
Moskalenko-Sulava, Aosta 1990.
The correct order is 7•.• exd4!
8 exd4 tLle6. White doesn't get time
for 'i'e2/lIdl because this time
9 'iie2 tLlxd4! is unsound. He has to
settle for a tiny edge such as 9 tLlc3
.i.e7 10 lIel 0-0 11 a4 .i.d7, as in a
Topalov-Kramnik match game in
2006.
The final QGA issue concerns
6 0-0 a6 7 'i'e2 tLle6. Then 8 tLlc3
and 8 lIdl often transpose, e.g.
8 l:tdl b5 9 .i.b3 tLle6 10 tLle3.
In contrast, 8 ':dl b5 9 .i.b3 c4
10 .i.c2 tLlb4! equalizes for Black,
e.g. 11 tLlc3 tLlxc2 12 'iixc2 .i.b7
13 e4? (13 d5!?) b4 14 e5 bxc3
15 exf6 gxf6 as in a celebrated
Szabo-Euwe game won by Black.
QUEEN'S GAMBIT
DECLINED
Players have been finding
finesses in the QGD since Pierre St.
129
Double QP Openings
Amant showed that after 1 c4 Black
A third way for Black to prompt
could reach it with 1...e6! and ... d5. liJf3 is a bluff, 1 d4 d5 2 c4 c6.
Even the recently popular order Actually it's two bluffs: Black is
of 1 d4 liJf6 2 c4 e6 and then 3 liJf3 counting on White to heed the
d5! is old. It became famous in the majority view that 3 liJf3 is best
games of Boris Kostic and Heinrich against the Slav Defense. Then he
can continue 3... liJf6 4 liJc3 e6,
Wolf in the 1920s.
hinting that he is armed to the
That was the 'Nimzo-threat'
order, based on White's fear of theoretical teeth with the latest
3 liJc3 .i.b4. As long as the Nimzo- Meran (5 e3 liJbd7 6 .i.d3 dxc4)
Indian is performing well - as it is analysis. White often flees from
now - this will be useful to Black. that, with 5 .i.g5, into main QGD
He deprives White of liJge2, which lines - which a bluffing Black
is often superior to liJf3, and of a wanted all along.
tempo he might put to better use in
He is taking two small risks. He
orders like 1 d4 d5 2 c4 e6 3 liJc3 might end up in the rather
liJf6.
dull/drawish
Exchange
Slav
Black has another way of (3 cxd5). And he can't reach the
inducing liJf3: 1 d4 d5 2 c4 e6 best version of the Orthodox
3 liJc3 and now 3... i.e7.
Defense (5 .i.g5 .i.e7 6 e3 0-0)
because an unforced ... c6 doesn't
work well with ... i.e7. A KramnikDeep Fritz match game in 2002
began 1 d4 liJf6 2 c4 e6 3 liJf3 d5
4liJc3 c6 5 i.g5 .i.e7?!.
Since White cannot play 4 i.g5
he usually picks the most natural
developing move, 4 liJf3, and life
goes on after 4... liJf6, as if 3... liJf6
4liJf3 i.e7 had been played. On the
minus side, this order denies Black
White was so surprised that he
non-....i.e7 variations, such as the
Cambridge Springs, Vienna and allowed equality after 6 e3 0-0
7 .i.d3 liJbd7 8 0-0 dxc4 9 i.xc4
Botvinnik.
130
Double QP Openings
~d5 10 ~xe7 tixe7 11 l:tc1?!
~xc3 12 l1xc3 e5. He forgot what
Black has closed some doors in
the diagram but kept others
has been known since the 1930s, open. The Tartakower-Makogonovthat 11 ~e4! favors White, e.g. Bondarevsky Variation and the
11...~5f6 12 '~g3 e5 13 ~f5 or Tarrasch Defense work badly with
12 ... c5 13 tie2 ~b6 14 ~b3 cxd4 ... ~bd7. But Black can still playa
15 ~xd4.
Semi-Slav (5 e3 c6) or, after 5 ~g5,
an Orthodox Defense (5 ... ~e7), a
Witnesses described Krarnnik as Cambridge Springs (5 ... c6 6 e3
'visibly deflated' when he realized a
tia5) or a Manhattan Variation
computer, hardly a master of
(5 ... ~b4).
psychology, had tricked him into a
This tabia also helps Black in a
line whose theory he'd forgotten.
Lasker Defense, 5 i.g5 h6 6 ~h4
~e7 7 e3 ~e4, as Vlf Andersson
ALTERNATIVE TABIAS
showed. In the more familiar
The QGD tabias we all know Lasker order, 1 d4 d5 2 c4 e6 3 ~c3
(and love, hate, etc.) typically ~f6 4 ~g5 i.e7 5 e3 0-0 6 ~f3 h6
appear after 1 d4 d5 2 c4 e6 3 ~c3 7 ~h4 ~e4, White can misplace the
~f6 4 ~g5 and then 4 ... ~bd7 or enemy bishop, 7 i.xf6!? i.xf6, and
4 ... ~e7. But there are others:
hold a small pull.
But in the Andersson order
~xf6? is bad because Black can
retake with the knight. A second
virtue of this order appears after
5 ~g5 h6 6 i.h4 i.e7 7 e3 ~e4
8 ~xe7 tixe7 9 cxd5 ~xc3
10 bxc3 exd5 11 tib3:
This one can come from the
Nirnzo-threat, 1 d4 ~f6 2 c4 e6
3 ~f3 and then 3... d5 4 ~c3 ~bd7.
It also arises from 1 d4 d5 2 c4 e6
3 ~f3 ~d7!?, a means of
discouraging the Catalan (4 g3 dxc4
and ... ~b6), and then 4 ~c3 ~gf6.
131
Double QP Openings
This is a standard anti-Lasker
strategy. After Black defends the
attacked d-pawn, with 11.. .c6 or
11..:iVd6, White tries to liquidate it
favorably with 12 c4!.
But here Black can defend it with
the superior 11 •..ttJf6!. His pieces
are better coordinated than normal
after 12 c4 c6 13 i.d3 i.e6 (14 0-0
l:tc8 15 l:.abll:tc7 16 cxd5 i.xd5 as
in P.Nikolic-Andersson, Leningrad
1987).
This tabia sets a trap: 4... cS?! falls
into a Tarrasch line, S cxdS exdS
6 ttJc3, that's been known to
The chief drawback to the 1 d4
d5 2 c4 e6 3 ttJc3 ttJf6 4 ttJf3 ttJbd7 be bad since Mr. MacDonnell
tabia is that White hasn't committed met Monsieur LaBourdonnais. A
his QB yet. He can create threats century-plus of experience with
with i.f4 and ttJb5. In most 1 d4 d5 6... i.e6 7 e4! and 6...ttJc6 7 i.xf6
lines the best response to i.f4 is gxf6 8 e3 confirms that.
... c5!. Butthat's not nearly as strong
when a knight is at d7.
Another virtue is White can
favorably
evade the Cambridge
That means Fritz Samisch's
S cxdS exdS 6 i.f4 should be good. Springs and Manhattan Variations
The natural 6 ... i.e7? walks into by answering a check with ttJbd2.
7 ttJb5! i.b4+ 8 ttJd2 i.a5 9 'iiVa4. For example, 4... ttJbd7 5 e3 i.b4+
After 6... c6 White can build a good 6 ttJbd2! and then 6... 0-0 7 i.d3 h6
foundation for the middlegame with 8 i.h4 is a good Orthodox Defense,
7 'iiVc2 i.e7 8 h3 0-0 9 e3 l:te8 e.g. 8... dxc4 9 i..xc4 b6 100-0 i.b7
10 i.d3, a better-than-usual 11 'iiVe2 i.e7 12l:tfdl ttJe4 13 ttJxe4
Exchange Variation.
i.xh4? 14 d5! with a big edge,
Gabriel-Stangl, Altensteig 1993.
Moving on to another tabia:
Mihai Suba liked the fluid center
Black gets in a Semi-Tarrasch, 1 d4
d5 2 c4 e6 3 ttJf3 ttJf6 4 ttJc3 c5
5 cxd5 ttJxd5!. But in 1977 Tigran
Petrosian 'caught me out' with
4 i.gS!?, he recalled. Suba had to
handle this unfamiliar position and
lost.
But 4... i.b4+ is more exact.
Then 5 ttJbd2 dxc4! is fine for
Black and so is 5 i.d2 i.e7!. In
1934, Hans Mueller recommended
4... i.b4+ 5 ttJc3 dxc4 6 e4 c5 - and
the Vienna Variation was born.
132
Double QP Openings
Black can play mind games with
the immediate 4... dxc4!?
This pays off after s... lbbd7 6 e3
'iliaS? 7 a3! and 8 b4 with
advantage. Or after 5•.. i.e7 6 e3
'Maybe I'm going to play the
Vienna, by meeting S lbc3 with lbbd7, an Orthodox Defense in
S... i.b4,' he says. 'But maybe I which he can meet 7 i.d3 0-0 8 0-0
intend S... c6, which means we'll get dxc4? with 9lbxc4! as Capablanca
to see who memorized more of the used to do.
Botvinnik Variation. And if I really
For example, 9... lbdS 10 i.xe7
wanted to play the Vienna, why
didn't I give the check at move lbxe7 11 l::tc1 lbg6 12 i.xg6! hxg6
13 e4 and Black cannot achieve the
four?'
needed ... eS break (CapablancaWhite often refuses to be drawn Tylor, Hastings 1930).
into this guessing game and replies
But whenever White plays lbbd2
S e3. Then S... cS is an offbeat QGA
and S... bS leads to a very double- there is a chance either ... cS or
edged and relatively unexplored ... i.d6/ ... eS will equalize quickly.
line that runs 6 a4 c6 7lbc3 i.b4!. Here S... cS has less impact because
it's a loss of tempo. But 5...lbbd7
Black's other option in the 1 d4 6 e3 h6 7 i.h4 i.d6! is fine, e.g.
d5 2 c4 e6 3 lbf3 lbf6 4 i.g5 tabia 8 i.d3 0-0 9 0-0 eS! 10 i.g3 iie7
is 4••• c6, again hinting at a 11 cxdS cxdS 12 dxeSlbxeS, RivasBotvinnik (S lbc3 dxc4) but also a Yusupov, Minsk 1982.
Cambridge Springs (s ... lbbd7 6 e3
'iliaS).
White may be tempted by
5lbbd2.
The previous QGD tabias allow
White to transpose into the tried and
true Exchange Variation with cxdS.
This one doesn't:
133
Double QP Openings
3 liJf3 liJf6 4 liJc3 liJbd7 S cxdS,
said S ..igS was better. 'White can
always exchange at a later stage
when the 'economic climate' is
more favorable for him,' he
explained.
It can come about via 1 d4 liJf6
2 c4 e6 3 liJf3 d5 4 liJc3 dxc4.
Black invites a QGA (after S e3), a
Vienna Variation QGD (after S e4
or 5 ..igS) or even a Nimzo-Indian
(after S ..igS ..ib4 6 e4 cS 6 a3!?).
The deceptive choice is
S 'it'a4+!?, trying to reach a superior
QGA. In the normal QGA, after
1 d4 dS 2 c4 dxc4 3 liJf3 liJf6 Black
can safely meet 4 'it'a4+ with
4 ... liJc6 or 4 ... c6. But 4 ... liJbd7
S liJc3 e6 is regarded as inferior.
This means 5 'it'a4+ liJbd7?!
traps Black in the bad 4 'it'a4+ line,
e.g. 6 e4 a6 7 ..ixc4 c6? 8 'it'dl!
..ie7 9 0-0 0-0 10 a4 b6 11 ..id3
..ib7 12 eS! and White soon had a
winning attack in Kasparov-Short,
London 1993. Better is 5.•. c6
6 'it'xc4 b5.
QGD: EXCHANGE
The first issue in the Exchange
Variation is when to exchange.
Today it's commonly done at move
four. But Alexander Alekhine,
commenting on 1 d4 dS 2 c4 e6
An example of that is I d4 dS
2 c4 e6 3 liJc3 liJf6 4 ..igS liJbd7
S e3 c6. Black is advertising his
interest in the Cambridge Springs.
If White can't remember the theory,
6 cxdS! makes sense. (A move later
is too late, 6 liJf3 'tiaS 7 cxdS
liJxdS!.)
But exchanging on move three is
too early. After 1 d4 d5 2 c4 e6
3 cxd5? exd5 4liJc3:
Black solves his main QGD
problem, how to develop with QB,
with 4... c6! and ..•..if5!, e.g. S 'tic2
..id6 6liJf3 liJe7 7 ..igS ..ifS.
It's also inaccurate to exchange
at move four if White is already
committed to liJf3. One of
Petrosian's most impressive wins as
Black began with 1 d4 liJf6 2 c4 e6
3 liJf3 dS 4 cxdS?! exdS S liJc3 c6
134
Double QP Openings
5 e3 0-0 6 cxd5? tDxd5! 7 .ixe7
'iixe7 8 tDf3 tbxc3 9 bxc3 b6!
(10 .ie2 .ib7 11 0-0 c5 12 tDe5
Controlling the bl-h7 diagonal is
tbc6 13 tDxc6? .txc6 14.tf3 ~ac8
vital in the Exchange. If Black gets
and Black won in Alatortsevcontrol, he usually equalizes. But if
Capablanca, Moscow 1935).
he's played an early ... tDbd7, then
White also has an easy time after
cxd5! usually gives White enough
tDf3
0-0 6 'ifc2 c6 7 e3 tbbd7
5
time to stop him, e.g. 1 d4 d5 2 c4
8
cxd5
tDxd5! 9 .ixe7 'iixe7
e6 3 tbc3 tbf6 4 .ig5 tbbd7 5 cxd5
exd5 6 e3 and .id3 or 'iic2 will (10 .te2 ~e8 11 0-0 ctJxc3 12 bxc3
make ... .if5 impossible. Salo Flohr, e5).
6 .ig5 .ie7 and ... .if5! (7 'iic2 g6
8 e3 .if5).
a QGD connoisseur, played cxd5
And bear in mind that a delayed
immediately after ... tbbd7 even if cxd5 may be conflict with White's
other early moves. Alekhine felt the
that was at move eight.
'economic climate' was favorable
The reason many GMs make the
after 1 d4 d5 2 c4 e6 3 tbc3 tDf6
capture early (1 d4 d5 2 c4 e6
4 .ig5 .ie7 5 e3 tbbd7 6 l:.c 1 0-0
3 tbc3 tbf6 4 cxd5) is that once
for 7 cxd5 because 7 ... tbxd5?
.tg5 and ... .ie7 are played, Black 8 tbxd5 drops a pawn.
can favorably swap two pairs of
But after 7 ... exd5 8 .ltd3 c6
pieces after ... tbxd5. The result
often resembles an Orthodox 9 'iVc2 ~e8 10 tDf3 he discovered
Defense in which Black simplifies his QR would have been better
placed at b I to support the minority
with ... dxc4 and ...tbd5.
attack ofb2-b4-b5.
Take the case of 1 d4 d5 2 c4 e6
Another issue is when to bring
3 tbc3 tbf6 4 .ig5 .ie7:
out White's KN. Consider 1 d4 d5
2 c4 e6 3 tbc3 tDf6 4 .ig5 c6 5 e3
tbbd7 6 cxd5 exd5 7 .td3 .ie7.
Black equalizes after 5 cxd5?
tbxd5! 6 .ixe7 'ifxe7. Similar is
135
Double QP Openings
This was Smyslov-Keres, World
Championship Tournament 1948.
White played 8 tbf3. Black replied
8... 0-0. Today both moves are
considered mistakes.
Black should play 8...tbe4!
because 9 tbxe4? dxe4 wins a piece
and because 9 ..txe4 ..txg5! or
9 ..txe7 'iixe7 10 'iVc2 f5 are fine
for him.
If White wants to play tbf3 he
should prepare it with 8 'iie2! so
that 8 ... tbe4? loses a pawn to
8 ..txe7 and 9 tbxe4.
By delaying tbf3 White retains
the tbge2 option and can seize
control of bl-h7. There is some
further subtlety if Black pursues
... ..tf5 in another way, 1 d4 d5 2 e4
e6 3 tbe3 tbf6 4 exd5 exd5 5 ..tg5
..te7 6 e3 e6 7 'ile2 tbbd7 and now
8 tbf3 tbf8 9 ..td3 tbe6 10 ..th4 g6.
He can meet the regrouping plan
with fl-f3 and e3-e4 or g2-g4, e.g.
9 •••tbe6 10 ..th4 g6 11 0-0-0 tbg7
12 f3!, e.g. 12 ... ..te6 13 g4! or
12 ... tbf5 13 ..tfl 'iVa5 14 ~bl ..te6
15 h3 0-0-0 16 e4! with a fierce
attack
in
Bronstein-Medina,
Gothenberg 1955.
But
Black
has
another
simplifying option after 8 ..td3 8•••tbh5 9 ..txe7 'fIxe7. This works
because 10 ..txh7 is not a check but
rather an unsound sack after 10 ... g6.
Also 10 tbf3 and 10 g4 allow
10 ... tbf4. White should play
10 tbge2 and keep a small edge.
His idea is ...tbg7 and ... ..tf5.
QGD EXCHANGE: ...tbbd7,
Black is at least equal after 11 0-0
.•• e6 and ... h6
0-0 12 l:tabl a5 13 a3 tbg7! 14 b4
We've seen how an early ... tbbd7
axb4 15 axb4 ..tf5! 16 tbe5 ..txd3
impairs
Black's chances for .....tf5!.
17 tbxd3 tbf5 18 ..tg5 tbd7 19 ..tf4
tbb6 and ... tbc4, as in a Najdorf But there are benefits to quickly
connecting the knights. Black can
game.
meet ..txf6 with ... tbxf6, which is
important
when 'fIc2/..td3 is lined
Black exploited a premature
8 tbf3 in that case. The careful up against the h7-pawn.
transposer will prefer 8 ..td3! and
then 8..•tbfS 9 tbge2.
It's also significant after 1 d4 d5
2 e4 e6 3 tbe3 tbf6 4 exd5 exd5
136
Double QP Openings
5 lbf3?! c6 6 i.g5 i.e7 7 e3 i.f5!
and then S i.d3 i.xd3 9 'iVxd3
lbbd7!:
White was sloppy twice, at move
five and again when he missed
9 i.xf6! i.xf6 10 'iVxd3, which
misplaces enemy pieces. It would
cost Black time to get his bishop to
the superior b8-h2 diagonal and to
find a good square for the knight.
c5!, with good play in MinoginaBelavenetz, Moscow 1990.
White's
counter-finesse
S lbge2 liIeS 9 0-0:
IS
The absence of ... c6 means
9... lbrs can be met by 10 b4!.
Usually White must invest a tempo,
such as l:tb 1, to prepare this
thematic push. Here it's tactically
But in the diagram Black's pieces
are well-coordinated thanks to
... lbbd7! and he should have no
difficulty equalizing, e.g. 10 0-0
0-0 11 l:tabl a5 12 a3 l:teS 13 b4
axb4 14 axb4 b5!.
based on 10... i.xb4 11 i.xf6 'iVxf6
12 lbxd5 or 11...gxf6 12 lbxd5!
'iVxd5 13 'iVa4.
For example, 1 d4 d5 2 c4 e6
3 lbc3 lbf6 4 cxd5 exd5 5 i.g5
i.e7 6 e3 0-0 7 i.d3 lbbd7 S'iVc2
l:teS and now 9 0-0-0 lbf8 10 Wb 1
i.e6 11 lbf3 a6 12 lbe5 l:tc8 13 h3
looking for an alternative to i.h4.
For example, 1 d4 d5 2 c4 e6 3lbc3
lbf6 4 cxd5 exd5 5 i.g5 i.e7 6 e3
lbbd7 7 i.d3 c6 S 'YWc2 and now
S... h6?
The final area of move order
finesse in the Exchange involves
... h6. Black often inserts this
Another move that Black often because of the danger of
defers is ... c6. He can't play the 'iVc2/i.d3/i.xh7+. But he has to be
freeing ... lbe4 without it, because careful that he gets the response the d-pawn would hang. But that is
usually i.h4 - that he wants.
offset by the possibility of ... c5 in
one step.
Consequently White should be
137
Double QP Openings
Chances would be roughly even
after 9 i.h4 lbf8 10 lbge2 lbe6 and
...lbg5!. Also 9... 0-0 10 lbge2 l:te8
11 0-0 lbf8 and ...lbe4.
However 9 i.f4! is much more
promising after 9... 0-0 10 0-0-0
or 10 lbge2 :e8 11 0-0-0 and
h2-h3/g2-g4.
White should be looking to
transpose from an Orthodox
Defense in which ... h6 has been
played to an Exchange. For
instance, 1 d4 d5 2 c4 e6 3 lbc3
lbf6 4 i.g5 lbbd7 5 lbf3 h6. Now
on 6 i.h4 i.e7 7 cxd5 Black has
7... lbxd5! with equality.
But there is an immediate
6 cxd5!?
If Black ducks the challenge,
6... exd5 7 i.h4 i.e7 8 e3, he is
slightly worse, e.g. 8 ... c6 9 'ilVc2 0-0
10 i.d3 l:te8 11 i.g3 lbf8 12 h3
i.e6 13 0-0 lb6d7 14lba4 a5 15 a3,
LSokolov-Seirawan, Dutch Team
Championship 2002.
Accepting the sacrifice, 6... hxg5
7 dxe6 fxe6 8 lbxg5, grants White
ample compensation (8 ... i.d6 9 e4!
as in Torre-P.Nikolic, Leningrad
1987).
Another rendering is 1 d4 d5
2 c4 e6 3 lbc3 i.e7 4 lbf3 lbf6
5 i.g5 0-0 6 e3 lbbd7 7 'i'c2 and
7... h6. That provoked the routine
8 i.h4 from such QGD authorities
as Viktor Korchnoi, Boris Spassky
and Rubinstein.
But their opponents missed a
good 8 ... c5!, e.g. 9 :dl 'i'a5
10 cxd5 lbxd5 11 i.xe7 lbxe7
12 i.e2 lbd5! 13 0-0 lbxc3 14 bxc3
b6 and ... i.a6, Carlsen-Hansen,
Skanderborg 2005.
However in 1961 an amateur
demonstrated 8 cxd5! is strong. His
opponent was shocked into
8 ... lbxd5?, losing a pawn, because
he feared 8... hxg5 9 dxe6 offered
too much compensation for White.
Kasparov revived the idea against
Portisch who played 8... exd5 and
was a little worse (9 i.f4 c5
10 i.e2).
QGD:ORTHODOX
In the Orthodox Defense, 1 d4 d5
2 c4 e6 3 lbc3 lbf6 4 i.g5 i.e7 5 e3
0-0 6 lbf3 lbbd7, White's finesses
begin with his QR. Its primary
function, either at c1 or dl, is to
discourage ... c5, Black's best
freeing device.
138
Double QP Openings
13 .tbS cxd4 14 'ii'xd4liJf6 IS liJeS
liJe8 16 .td7 (Bogolyubov-Mueller,
Zurich 1934).
Black equalizes after 7 a3?! c5!
and 7 'iib3 c5! (8 cxdS cxd4
9 liJxd4 liJcS! 10 'iic2 liJxdS,
Poluljahov-Tregubov, Krasnodar
2000).
White would love to force his
way to these positions. But if Black
is careful with ... c6 he can only try
to trick him. One way is an early
'iic2 in place of e2-e3. For example
1 d4 d5 2 c4 e6 3 liJc3 liJf6 4 .tg5
.te7 5liJf3 0-0 6 'ii'c2:
That explains 7 ltel. It usually
prompts 7... c6, which safeguards
the c- and d-pawns and prepares
other freeing ideas such as ... liJe4.
But if ... c6 has already been
played, because Black uses an order
such as 1 d4 d5 2 c4 c6 3 liJf3 e6
4 liJc3 liJf6 5 .tg5 liJbd7 6 e3
i.e7?!, there is no pressing need for
ltel.
Rubinstein showed that White
can use his heavy pieces more
effectively with 7 'iic2! 0-08 ltd1!.
Then 8... cS transposes into a book
position in which Black is a move
behind and under strong pressure
after 9 cxdS exdS 10 dxcS.
And thanks to the Q+R tandem,
the familiar liquidation, 8... dxc4
9 .txc4liJdS 10 .txe7 'iixe7 11 0-0
liJxc3 12 "ilxc3, doesn't allow ... eS
and leaves Black with serious
problems after 12 ... cS 13 "ila3 or
Now 6... c6?! 7 e3 liJbd7 8 ltd1!
achieves White's goal. This order
also opens a 0-0-0 option, e.g. 6 ... h6
7 .txf6! .txf6 8 0-0-0 and 9 e4 with
good chances.
And after 6...liJbd7 he can switch
to an Exchange Variation since
7 cxd5 liJxd5? loses a pawn
(8 liJxdS exdS 9 .txe7 "ilxe7
10 'iixc7 or 8 ... .txgS 9 liJxc7).
But the problem with an early
"ilc2 is that White's rook isn't in
place yet so 6... c5! is good.
The situation is similar when
Black uses the 1 d4 d5 2 c4 e6
3 liJc3 .te7 order to delay .tgS and
White voluntarily delays it further,
139
Double QP Openings
4 tbf3 tbf6 5 'iVc2. Then 5.•. 0-0
6 iLg5! h6 allows him to reach the
promising 7 iLxf6 iLxf6 8 0-0-0.
Another natural move, 5•••tbbd7,
allows an excellent Exchange
Variation, 6 cxd5! exd5 7 iLf4! c5
8 e3 0-0 9 iLd3, since Black's QN is
misplaced.
But once again Black can strike
with ... c5 before White has deterred
it with l:.dl, e.g. 5... c5!? 6 cxd5
cxd4! 7 tbxd4 exd5 8 iLg5 tbc6
9 l:tdl h6 10 iLh4. 'iVa5 as in
Rustemov-Ubilava, Olite 2006.
... dxe4 attacks it. White can avoid
that by delaying either tbf3 or tbc3.
For example, 1 d4 d5 2 c4 e6 3 tbf3
tbf6 4 iLg5 iLe7 5 e3:
Now 5...tbe4 has been under a
cloud since W.Cohn-Leonhardt,
Ostende 1907 went 6 iLxe7 'iVxe7
7 tbbd2! 0-0 8 iLd3 and 8... f5
9 0-0 tbd7 10 'iVb3 c6 11 lIac 1 gave
White an excellent Stonewall
Dutch.
However, 5... 0-0 puts the ball
back in his court. He can continue
to withhold tbc3 with 6 'iic2. Then
he gets that ideal Orthodox Defense
QGD: LASKER
after 6... tbbd7 7 tbc3 c6 8 l:.dl!.
A typical starting point of But once again 7••. c5! is good and
Emanuel Lasker's defense is 1 d4 should equalize.
d5 2 c4 e6 3 tbc3 tbf6 4 iLg5 iLe7
5 e3 h6 6 iLh4 0-0 7 tbf3 tbe4. QGD: CAMBRIDGE SPRINGS
Unlike similar positions, the retreat
White has various means of
to g3 isn't promising here (8 iLg3?! avoiding the Cambridge Springs in
iLb4 9 l:.cl c5).
traditional orders. But most of them
The trade-seeking ...tbe4 works
best when White has one knight at
c3, so ... tbxc3 is possible, and
another at on f3 so that tbxe41
allow Black to fall back into a good
Orthodox or Lasker Defense. If
White plays 1 d4 d5 2 c4 e6 3 tbc3
tbf6 4 iLg5 tbbd7 5 e3 c6 6 a3:
140
Double QP Openings
Bogolyubov game won by Black
and 13 'iVc2 exd4 14 exd4 tLl7b6
15 tLle5 it.e6 ended in a quick draw
in Piket-Ivanchuk, Melody Amber
1993.
He can get a reasonable version
of Rubinstein's Orthodox line,
6... it.e7 7 tLln 0-0 8 'iVc2 and ~d1.
But instead, Black can transpose
to a good Lasker with 7...tLle4 or
7... h6 8 it.h4 tLle4. Then a2-a3
plays little role.
The strength of the main line
Cambridge Springs, 6 tLln 'iVa5,
rests today on whether White gets
an edge from 7 cxd5 tLlxd5 8 'iVd2
it.b4 9 ~c1. Black has to decide
when to play ... e5, ... h6 and ... 0-0 or
he will be distinctly worse. The
crucial tabia is:
This has performed well for
Black, e.g. 13 e4 tLlf4 14 it.c4 tLlg6
15 a3 tLlxh4! was an Alekhine-
The hard part for Black is
reaching the tabia. The 1983
Kasparov-Smyslov
candidates
match tested various orders and
showed that 9•••e5?! is bad in view
of 10 a3! it.xc3 11 bxc3 'iixa3
12 e4 with a dangerous attack.
That kind of sacrifice is a major
Black concern. Alekhine said
Black's best order to reach the tabia
was 9..•0-0 10 it.d3 e5 and then
11 0-0 h6 12 it.h4 ~e8. But this was
based on his claim that 11 a3!?
it.xc3 12 bxc3 'ii'xa3 is unsound.
If a Cambridge Springs player
has doubts about that, he should
consider the slightly different
9••• 0-0 10 it.d3 h6.
Then 11 it.h4 .l:.e8 12 0-0 e5
arrives at the tabia and 12 a3 it.f8 is
also safe. Black avoids 11 ... e5
12 a3!?, which may be a good
version of the gambit.
141
Double QP Openings
But this order has a potential
problem in 11 i.f4!? ttJxf4 12 exf4,
which stops ... e5. For example
12 ... ttJf6 13 0-0 c5 14 dxc5 i.xc5
15 a3 i.e7 16 ttJe5 l:td8 17 'iie2
was promising for White in Hillarp
Persson-Vera, Yerevan 1996.
Some authorities say 9 ... h6 is
best because 10 i.f4 is less
promising here and 10 i.h4 0-0
11 i.d3 e5 reaches the tabia. But
White might improve with
11 i.c4!? as in an Alekhine game
that went 11...ttJxc3 12 bxc3 i.a3
13 l:tbl e5 14 i.g3.
QGD: BOTVINNIK
He means about 5 i.g5 dxc4 or
5 e3 ttJbd7 6 i.d3 dxc4. If White
doesn't like his answers to
those questions, he'll also be
disappointed when he looks for
alternatives and sees that 5 cxd5
exd5! is an inoffensive Exchange
Variation (6 i.g5 i.e7 and ... i.f5!).
And even ifhe knew Black's intent,
White would be reluctant to play
cxd5 at move 3 or 4 because he'd
end up in another super-quiet line,
you
the Exchange Slav.
Therefore Black should pick the
order that avoids what he wants to
avoid. If he doesn't mind 9... 0-0
10 i.d3 h6 11 i.f4!?, that's his best
bet. But ifhe's willing to accept one
of the gambits, other orders are
superior.
If
look
up
Mikhail
Botvinnik's variation in books you
may find it under 1 d4 d5 2 c4 e6 3
ttJc3 ttJf6 4 ttJf3 c6 5 i.g5 dxc4.
But the little secret of Botvinnik
specialists is very few of them use
that order. It grants White too many
promising alternatives, including
4 i.g5 and 4 cxd5.
It is other orders, such as 1 d4 d5
2 c4 c6 3 ttJf3 ttJf6 4 ttJc3 e6, that
make the Botvinnik the weapon it
IS.
'Do you really know as much
about the Botvinnik as I do?,' Black
is asking. 'Or about the Meran?'
So what a Botvinnik player is
really saying in this order is, 'You
don't have much choice. It's the
Botvinnik for you or quick equality
forme.'
When White goes into 5 i.g5
dxc4 he justifies the lost pawn with
6 e4 b5 7 e5 h6 8 i.h4 g5 9 ttJxg5
hxg5 10 i.xg5, with mind-boggling
complications. One of the less
analyzed positions occurs when he
retreats instead with 9 i.g3.
142
Double QP Openings
1 d4 d5 2 c4 e6 3 4:Jf3 4:Jf6 4 4:Jc3
dxc4 5 iLg5 iLb4 6 e4 c5, but the
conservative 7 iLxc4 is the main
line today. A resulting tabia is:
There's a lot happening and it's
happening on both wings, e.g.
9... 4:JdS 10 4:Jd2 iLe7 11 iLe2 iLb7
12 4:Jde4 4:Jd7 13 h4 cS 14 hxgS
4:Jxc3 IS 4:Jxc3 cxd4 16 'i'xd4
.txgS 17 0-0 as in Hillarp PerssonHector, Gothenburg 2006, won by
White.
Antique theory claimed White is
much better (11...'i'xc4+ 12 'iitgl
gxf6? 13 l:.c1 or 12 ... 4:Jd7 13 l:.c1).
But 11 •.• gxf6! turned out to be fine
If this seems attractive to you it's for Black and today his concern is
worth adding 9 iLg3 to your the best route to the diagram.
repertoire because you can get to
One way is 7 iLxc4 cxd4 8 4:Jxd4
the diagram via other orders. For
'i'a5
after which 9 iLxf6 iLxc3+
instance after 1 d4 dS 2 c4 e6 3 4:Jc3
bxc3
'i'xc3+ 11 'iitn gets there.
10
c6 4 4:Jf3 4:Jf6 S iLgS a Black who
doesn't like the Botvinnik or The other is 8...iLxc3+ 9 bxc3 'i'a5
Cambridge Springs will try S... h6. and then 10 iLxf6 'i'xc3+.
Books used to say 'if 6 iLh4? then
The advantage of the first order
6... dxc4! wins a pawn.'
is that Black preserves his KB
for
dark square defense. That's
However, 6 iLh4 dxc4 7 e4 is
important in the second order when
promising even after 7... gS 8 iLg3
White tries 10 iLbS+ instead of
bS. Then White has various means
10 iLxf6.
of seeking compensation and they
Then 10 ... 4:Jbd7 11 iLxf6 'i'xc3+
include 9 eS, reaching the last
12 'iitfl gxf6 13 h4! and l:.h3 offers
diagram.
good chances.
QGD:VIENNA
But in the first order the check is
There have been many attempts harmless, 8... 'i'aS 9 iLbS+ 4:Jbd7
to refute the Vienna Variation, 10 iLxf6 gxf6.
143
Double QP Openings
The problem with 8 ... 'iVa5,
according to authorities going back
to the 1930s, is '9 ..td2! and then
9... e5 10 l2Jc2.'
But 9 ..td2 is so passive that
9... 0-0 makes White's pieces look
misplaced (10 'iVe2 .:f.d8 11 l2Jf3
l2Jc6 12 a3 ..txc3 13 ..txc3 'iVh5,
Summerscale-Wells,
Millfield
2000). Instead, White may have to
risk 10 l2Jc2 ..txc3 11 ..txc3 'iVg5!
12 'iVe2 'iVxg2 13 0-0-0.
The correct response is S...l2Jc6!.
Then 6 ..tgS loses its punch to
6.....te7!.
Black can also fall into the same
trap via the Semi-Tarrasch, 1 d4 d5
2 c4 e6 3 l2Jc3 l2Jf6 4 l2Jf3 c5, if he
plays 5 cxd5 exd5? 6 ..tg5! instead
of 5 ... l2Jxd5!.
SLAV DEFENSE
The early moves of the Slav
conceal
quite a bit of subtlety
TARRASCHISEMI-TARRASCH
because of the strong feelings
Today's Tarrasch tabia usually players have about the main lines comes about via 1 d4 dS 2 c4 e6 the Open Slav and the Meran and
3 l2Jc3 cS 4 cxdS exdS S l2Jf3 l2Jc6 Tchigorin variations of the Semi6 g3 l2Jf6 7 ..tg2 ..te7 8 0-0 0-0 and Slav. As a result, a Slav is often a
Black's d-pawn becomes a focus of silent battle between one player
the middlegame after a pawn trade. trying to avoid, say, the Open Slav
That's why Black may defend it and the other trying to avoid the
quickly, with S•••l2Jf6?!.
Meran, with success going to the
better transposer.
His punishment is 6 ..tgS!. He
has transposed into 1 d4 d5 2 c4 e6
3 l2Jf3 l2Jf6 4 ..tg5 when Black
plays the sloppy 4 ... c5?! and allows
5 cxd5 exd5 6 l2Jc3.
Take the odd-looking 1 d4 l2Jf6
2 c4 c6 and 3... dS. It doesn't get
much attention although it was used
by Paul Keres, Yefim Bogolyubov
and Alekhine. White often responds
3 l2Jc3 automatically and then
realizes after 3 ... d5 that he's ruled
out anti-Meran lines in which he
can play l2Jbd2 or delay a QN move,
as in 1 d4 d5 2 c4 c6 3 l2Jf3 l2Jf6
4 e3 e6 5 ..td3 l2Jbd7 6 0-0.
Or suppose 1 d4 dS 2 c4 c6 and
now 3 e3:
144
Double QP Openings
Black can transpose into a good
version of the Schlechter Variation
of the Gruenfeld (1 d4 d5 2 c4 c6
3 tZJf3 tZJf6 4 e3 g6). The
Schlechter's stodgy reputation is
based on lines with the more
enterprising .if4 or .ig5.
SLAV: EXCHANGE
Why would White play such a
quiet move? The reason is that an
early e2-e3 is a good way of
avoiding the high-maintenance
Open Variation (1 d4 d5 2 c4 c6
3 tZJc3 tZJf6 4 tZJf3 dxc4). To get an
advantage White has to play 5 a4
.if5 6 e3 or 6 tZJe5, or risk the
5 e4!? gambit. Either way he has to
memorize a huge amount of book.
White can initiate the Exchange
Variation at move three (as most
players do today), at move four (as
Botvinnik preferred) or later. The
rationale for delay is that Black may
prematurely commit his pieces
before he knows that cxd5 is
coming, e.g. 1 d4 d5 2 c4 c6 3 tZJf3
tZJf6 4 ttJc3 .if5? 5 cxd5 cxd5
6 ~3! favors White.
But if Black plays ... dxc4 after
e2-e3 White can retake on c4
without spending a tempo on a2-a4.
He transposes into a favorable
QGA, as if Black played 1 d4 d5
2 c4 dxc4 3 e3 c6?! 4.ixc4.
In case of 1 d4 tZJf6 2 c4 c6 3 e3,
Black has a good alternative,
3... g6!. Then he transposes into one
of his best versions of the King's
Indian Defense, e.g. 4 tZJc3 .ig7
5 tZJf3 0-0 6 .id3 d6! 70-0 tZJbd7 8
'ilc2 e5 9 l:tdl 'ile7 10 .in e4 11
tZJd2 l:te8.
There are two other drawbacks to
e2-e3 in any Slav order. One is that
... .if5! is safer to play when
White's QB is blocked in. Also,
The same goes for 4 ... ttJbd7
when 5 cxd5! cxd5 6 .if4 leaves his
QN misplaced.
White has the better pieces after
6... e6 7 e3 .ie7 8 h3 0-0 9 .id3,
e.g. 9... tZJb6 10 0-0 .id7 11 ttJe5
l:tc8 12 'ilb3! .ie8 13 l:tfc1 tZJfd7
14 a4!, Reshevsky-Bernstein, U.S.
Championship 1951.
145
Double QP Openings
In the normal Exchange, 3 cxd5
cxd5 4 lbc3 there is a small trap for
the unwary Black who thinks this
order allows 4•.. lbc6. He'll be
surprised by 5 e4!. White enjoys a
clear advantage after 5... dxe4? 6 d5
lbe5 7 'iVa4+ iLd7 8 'iVxe4.
Black should play 5... lbf6,
hoping White will allow the
endgame variation of the Panov
Caro-Kann (6 exd5 lbxd5 7 lbf3
iLg4 8 'iVb3 iLxf3 9 gxf3 etc.).
But 6 e5 lbe4 7 iLd3 is much
more promising, e.g. 7 ... lbxc3
8 bxc3 e6 9 lbf3 h6 10 0-0 lba5
11 .l::tbl iLd7 12 lbd2 and 'i'g4,
Gaprindashvili-Murshed, Polanica
Zdroj 1986.
SLAV: 3lbc3
This was once preferred to 3lbf3
(after 1 d4 d5 2 c4 c6) because it
does a better job of discouraging
... iLf5, as in 3 lbc3 lbf6 4 e3 iLf5?!
5 cxd5 cxd5 6 'iVb3.
Today when White chooses
3 lbc3 it's usually for a different
reason - to reach the Semi-Slav,
after 3...lbf6 4 e3 e6 5lbf3, without
allowing the Open Variation, 4 lbc3
dxc4.
Books say the downside to
3 lbc3 is 3.•. dxc4 because ... b5-b4
will attack the knight.
But in practice Black often plays
3 lbc3 dxc4 with no intention of
... b5. He knows all about the 3lbc3
lbf6 4 e3 trick to avoid the Open
Slav. 'But I'm the one doing the
tricking, ' he thinks to himself,
'because I'll get there after 3... dxc4
4 lbf3 lbf6!.'
White can avoid the Open
with 4 e4!? But then 4... b5 has
more justification (5 a4 b4 6 lba2
lbf6!).
Instead he often prefers 4 e3 .
There follows 4 ... b5 5 a4 b4 6lba2
e6 7 iLxc4lbf6 and since he has no
e4-pawn under attack he can
continue 8lbf3.
What happened: This often
comes about after 1 d4 d5 2 c4 c6
3 lbf3 lbf6 4lbc3 dxc4 when White
does not try for advantage with 5 a4
or 5 e4!? but plays the innocuous
5 e3.
146
Double QP Openings
He wants to pull White into
6 ttJc3 .tb4!, a version of the
Noteboom Variation in which
White has pushed to e4. The pawn
is more aggressive there but also
more vulnerable. Both sides have
chances after 7 .te2 ttJf6 8 'Yi'c2
.tb7 9 0-0 a6 10 eS ttJdS 11 ttJe4
ttJd7.
Then S•.. bS 6 a4 b4 7 ttJa2 e6
8 .txc4 reaches the diagram, e.g.
8... ttJbd7 9 0-0 .tb7 10 .td2 as
11 ttJc1 .td6 12 ttJb3 cS 13 'i'e2
0-0 141Hdl 'i'b6. Black has the last
laugh. He not only got into the
Open Slav that White tried to avoid
but even reached a quite equal
version of it.
The thematic idea is 6 axbS cxbS
7 b3. But again Black should attack
e4, e.g. 7 ... ttJf6! 8 eS ttJe4 9 bxc4
.tb4+ 10 ttJbd2 ttJc6 11 .td3 ttJcS!,
with good chances for Black in
Belyavsky-Kharlov,
Ljubljana
2002.
More common after 1 d4 dS 2 c4
c6 3 ttJf3 is 3.•.ttJf6. Then 4 'Yi'c2!?
allows White to sidestep the Meran
SLAV: 3 ttJf3
by meeting 4... e6 with S g3. Black
Even though 3 ttJf3 dxc4 has less finds his main choice is between a
punch than the 3 ttJc3 dxc4 version, closed Catalan, which few Meran
Black might like it because of players like, and a Stonewall Dutch
tricks. When White tries to punish with S... ttJe4 and 6... fS, which even
3... dxc4 with 4 e4 bS S a4, he can fewer do.
avoid transposition to the Slav
The queen move can also be a
Gambit, S... ttJf6, by means of good weapon against an Open Slav
S.••e6!.
player because of 4•.. dxc4 S e4 bS
6 b3!. White has good compensation after 6 ... cxb3 7 axb3 e6
8 .td2, e.g. 8... aS 9 .td3 i..b7
10 0-0 .te7 11 .tc3 (NakamuraHansen, Malmo 200S).
There's no major drawback to
4 'i'c2. But 4.•. g6 is a better-thanusual version of the Schlechter Slav
because ... iHS will gain a te~po.
147
Double QP Openings
If White passes up the various
evasions of the Open Slav and
continues 4 ttJc3, Black has an
alternative to 4... dxc4 5 a4 it.f5 in
the form of 5... e6!?
The natural temptation is to
punish such a passive move with
6 e4. But 6 ...it.b4 7 e5 ttJd5 8 it.d2
b5 9 axb5 transposes to a line of the
Slav Gambit that's not at all bad for
Black (5 e4 b5 6 e5 ttJd5 7 a4 e6
8 axb5 it.b4 9 it.d2). This has a
psychological benefit since White's
preference for 5 a4 rather than 5 e4
indicates he's not as confident in
gambit lines.
The more common reply in the
diagram is 6 e3. Then 6•••c5 7 it.xc4
gives us a QGA in which White has
an extra tempo, a2-a4. But
Kramnik, among others, likes
Black's position because after
7 ... ttJc6 8 0-0 cxd4 9 exd4 it.e7 he
gets some control of d5 by
occupying the hole at b4 with his
knight, e.g. 10 'ii'e2 0-0 II ':'dl
ttJb4. Black seems to stand well ,
e.g. 10 it.g5 0-0 11 ':'e 1 it.d7,
Navarra-Svidler, Turin 2006.
SLAV: THREE PAWNS
When Black uses his first three
moves to put pawns at d5, c6 and e6
he sacrifices development for
transpositions. He can drag a
reluctant White into the Botvinnik,
Noteboom, and other sharp
variations or even a better-thanusual Dutch, as well as into some
Meran lines, while avoiding more
dangerous Meran lines.
Black may prefer 1 d4 d5 2 c4 c6
followed by 3 ... e6 to the other
order, 2 ... e6 and 3 ... c6, because
2 ... c6 is more likely to prompt
3 ttJf3. In this way he won't have to
face the Marshall gambit, 3 ttJc3 e6
4 e4. But another Black may prefer
2 ... e6 3 ttJc3 because it rules out the
ttJbd2 option in the Semi-Slav.
The 3 ttJc3 order deprives White
of 4 it.g5 and discourages 4 cxd5
because 4 ... exd5! and ... .tf5
equalizes.
Black risks Marshall's 4 e4 but
can try to bail out with 4 •.. it.b4 and
then 5 exd5 cxd5 6 ttJf3 into a
148
Double QP Openings
relatively benign Panov Caro-Kann.
However, 5 e5 or 5 'i'g4 are more of
a challenge, e.g. (5 'i'g4 tt:Jf6
6 'i'xg7 l:tg8 7 'i'h6 dxe4 8 a3
i.xc3+ 9 bxc3 e5 as in GelashviliIzoria, Athens 2005).
The chief alternatives at move
four are a little-explored Catalan,
4 g3, and keeping the situation fluid
with 4 tt:Jf3 or 4 e3. The flexibility
of the Three Pawns Variation is
illustrated by Black's options after
4 tt:JO:
(a) 4... fS creates a Stonewall.
Black may feel happy because he
won't face the tt:Jh3-f4 and tt:Jbd2
maneuvers that are dangerous in
normal Dutch orders. But White
hasn't ruled out 5 i.f4!, which
favors him.
This is White's way of saying,
'I'm guessing you don't want a
Botvinnik because if you did you
would have played 4 ... tt:Jf6 and then
5 i.g5 dxc4. But now you've got a
problem if you want to avoid the
Botvinnik. '
He means S..•f6 when 6 i.d2 cS
7 e3! is a dangerous gambit based
on the weakened e6, e.g, 7... tt:Jc6 8
i.xc4! cxd4 9 exd4 tt:Jxd4 10 O-O!.
For example, 1O ... tt:Jxf3+ 11 'i'xf3
'i'xd2
12 l:tadl 'i'xb2 13 tt:Jb5 or
(b) 4•..tt:Jf6 invites a Semi-Slav
(5 e3 tt:Jbd7) or a QGD (5 i.g5). 12 ... 'i'g5 13 .l:.fel i.e7 14 tt:Jb5
Black might expect the latter 'i'c5 15 .:tel, Glek-An. Bykhovsky,
because if White wanted a Semi- Porto San Giorgio 1999.
Slav he could have sought it
A more conservative White will
directly with 4 e3, without having prefer 4 e3. That rules out the
to deal with 4 tt:Jf3 dxc4.
Noteboom, Botvinnik, et al and
invites a Semi-Slav, 4 ... tt:Jf6 5 tt:Jc3.
(c) 4.•. dxc4 steers White towards Black can keep his cards concealed
a Noteboom (5 a4 i..b4). The a bit longer with 4...tt:Jd7.
Noteboom is so theory-heavy that
He is hinting that on S tt:JO he'll
only the brave and the booked up
will play 4 tt:Jf3 confidently build an excellent Stonewall
(S .•. fS!). Then 6 i.e2 tt:Jgf6 7 0-0
unless ...
i.d6 is the kind of aggressive
Unless White intends to meet position Black can only dream
about in a Dutch Defense move
4 tt:JO dxc4 with S i..gS!?:
149
Double QP Openings
order, e.g. 8 a3 a5 9 l::tb1 fie7
10 i.d2 0-0 11 b4 ttJe4 12 bxa5
':f6! 13 i.e1 l1h6 13 g3 g5 with a
fine game, Polak-Stangl, Brunn
1993.
White has extra options, such as
8 .lif3 and ttJge2/a2-a4 in that line,
or 7 .lid3 .lib7 8 ttJf3 a6 9 0-0 c5
10 .lie4 but they aren't very
impressive.
White can counter with the
finesse S i.d3, so that S... fS 6 cxdS!
exdS? 7 i.xfS hangs a pawn and
6... cxd5 is an Exchange Slav in
which Black's QN is misplaced.
As a result, Black usually plays
5 ... ttJgf6 instead and 6 ttJf3
gets White to the Semi-Slav he
wanted.
His best in the diagram may be
simply to transpose to a Reynolds
Variation of the Meran with 7 .lid3
.lib7 8 ttJf3 a69 e4! and then 9... c5
10 d5 c4 11 .lic2 ttJgf6!' But the
super-sharp Reynolds isn't for
everyone.
Alexey Dreev and Yevgeny
Sveshnikov have met 5 .lid3 with a
counter-finesse, S... dxc4 6 i.xc4
sidesteps. The other version occurs
when White's KN comes out first,
1 d4 dS 2 c4 c6 3 ttJf3 e6.
bS.
Clearly the Three Pawns holds a
huge number of feints and
After 4 e3 ttJd7 Black can again
playa quasi-Meran without fear of
e3-e4-e5, e.g. S .lid3 dxc4! 6 .lixc4
bS 7 .lid3 a6 and 8 0-0 ttJgf6 is a
harmless line of the Meran.
The thematic idea is 8 ttJc3 i..b7
9 e4. But 9... cS 10 eS has less point
when there's no knight on f6. And
once again a Reynolds-like 10 d5
offers little following 10 ... exd5
This is an accelerated Meran 11 ttJxd5 ttJgf6, e.g. 12 ttJxf6+
with little experience to judge it. fixf6 13 0-0 .lid6 14 a4 c4 15 .lic2
For example, 7 i.e2 .lib7 8 ttJf3 a6 0-016 i.g5 fie617 ttJd4 fie5 18 f4
9 0-0 cS is a fine version of the fixd4+! 19 fixd4 .lic5, as in JelenMeran (10 d5 exd5 11 ttJxd5 ttJgf6 Sveshnikov, Nova Gorica 1998.
12 e4!? ttJxd5 13 exd5 i.d6,
The major extra option White
Scherbakov-Sveshnikov, Rostov
enjoys in this order is 4 .ligS.
1993).
150
Double QP Openings
But the S liJbd2 order has two
advantages based on being able to
retake on c4 with the knight. First,
he doesn't allow a Meran since
S... liJbd7 6 ~d3 dxc4 7 liJxc4 bS?!
8 liJceS liJxeS 9 liJxeS ~b7 10 0-0
~d6 11 f4 0-0 12 'iVf3, LitinskayaArakhamia, Womens Candidates
1988, is excellent.
Then 4...liJf6 is a QGD. Black
could prefer 4•.. ~e7 5 ~xe7 'ifxe7.
Trading off the good bishop seems
dubious but some might be happy to
face 6 e3 liJf6 7 liJc3 0-0 8 :c 1
liJbd7 9 .te2 dxc4 10 .txc4 eS.
SEMI-SLAV DEFENSE
Second, he doesn't allow Black's
best version of Mikhail Tchigorin's
system, 6... ~d6 7 e4 dxc4 (8 ..ixc4
eSt) because 8liJxc4! stops ... eS.
Alekhine thought this trick was
so strong that Black had to meet an
early S liJbd2 with S... cS!? That's
why White often inverts the order
today - S ~d3 and if S... liJbd7 then
6 liJbd2. His reasoning is that
Black's QN is misplaced at d7 in
case of a later ... cS.
The Semi-Slav arises via the
QGD, Slav or Three Pawns orders.
White's first finesse is developing
his QN on d2 instead of c3, as in
Then Black's best may be to fall
1 d4 d5 2 c4 c6 3 liJf3 liJf6 4 e3 e6
back
on the QGA in which White is
and now 5 liJbd2.
committed to a quiet liJbd2, as
This allows him to reach Kramnik did in the 2006 world
favorable book positions if Black championship match, S... dxc4
continues S... liJbd7 6 .td3 ~d6 6 ~xc4 cS.
7 e4! dxe4 8liJxe4. That's the same
position he gets from S liJc3 liJbd7
Slightly different is another anti6 ~d3 .td6 7 e4!.
Meran order, 5 ~d3liJbd7 6 0-0.
151
Double QP Openings
This employs psychology: Meran
players usually don't want a QGA
position, such as 6 ... dxc4 7 .Jixc4
c5, even if it's their best option.
Instead, they will be attracted to the
Meran-like 7... b5 8 .Jid3 a6 and
take their chances after 9 a4.
But this trickery backfires if
Black adopts the Tchigorin plan of
6..•.Jid6!. Then 7 tDbd2 0-0 8 e4 e5
is equal.
And on 7 tDc3 0-0 we get a
that
occurs
more
position
commonly via 1 d4 d5 2 c4 e6
3 tDc3 tDf6 4 tDfJ c6 5 e3 tDbd7
6 .Jid3 .Jid6 when White plays the
inexact 7 O-O?! and Black replies
7 ... 0-0.
The difference is that now 8 e4
dxc4! 9 .Jixc4 e5 is good because
Black is castled. However, if White
had played 7 e4! In the 5 .Jid3 Nbd7
6 tDc3 .Jid6 order Black cannot
respond 7 ... dxc4 8 .Jixc4 e5?
9 dxe5 tDxe5 10 tDxe5 .Jixe5 in
view of 11 'i'xd8+ and .Jixfl with
advantage.
13 gxh8('i'). His only good
alternative is a highly complex
variation (10 tDa4 c5 11 e5 tDd5)
that may require him to sack at least
a pawn to have any hope of
advantage.
For that reason many a White
will avoid the issue by playing
9 0-0. Then 9 ... b4 10 tDe4 gets all
the attention in books and the
verdict is mixed. But 9.•• a6!? is the
trickster's choice.
What has happened is Black
reached a normal Meran (8 ... a6) in
which the second-best 9 0-0 has
been played and met by 9 ... .Jib7.
This is a good weapon against a
White who doesn't like the
Reynolds Variation (10 e4 c5
11 d5). The point is that 10 e4 c5
11 e5 is like the old main Meran
line but with 0-0 and ....Jib7 added.
In the main Meran line, 1 d4 d5
2 c4 c6 3 tDf3 tDf6 4 e3 e6 5 tDc3
tDbd7 6 .Jid3 dxc4 7 .Jixc4 b5
8 .Jid3, there is a deceptive order
beginning with 8....Jib7 instead of
This is a fairly new position and
8... a6. The natural 9 e4, which is the
so
far chances appear mixed after
right move against 8 ... a6, is
11...cxd4
12 tDxb5 .JixfJ or
answered by 9 ...b4!.
11...tDd5 12 a4 b4 13 tDe4 cxd4
Then White can enter a four-'ii 14 .Jig 5 'i'b8 (Dambachermiddlegame (!) after 10 e5 bxc3 Stellwagen, Dutch Team Champ11 exf6 cxb2 12 fxg7 bxal('ii) ionship 2006).
152
Chapter Six:
Indians
The most popular answer to 1 d4
today is 1...tbf6 and the most
popular reply is 2 c4, opening the
door to the Hypermodern and
related defenses we call Indians. We
tend to overlook 2 tbf3 because it
usually 'just transposes.'
makes sense for a White who only
likes the Tromp when he's allowed
to double pawns, i.xf6, or to build
a center with f2-f3, after 2... tbe4.
For the Tromp player who has no
interest in transposing to another
opening, this is a critical position:
But it allows White to sidestep
the Benko Gambit, Budapest
Defense and a few other systems. In
exchange, Black can enter a King's
Indian without fear of the Samisch
Variation or a Modern Benoni
without concern about f2-f4
systems. Or he can offer a QGD
(2 ... d5) in which White is
committed to tbf3.
Books also underestimate 2 ..tg5
but it becomes a significantly
stronger weapon in the hands of a
flexible thinker who can shift into a
Torre Attack or a QGD when
appropriate.
For example, after 2 ... e6 White
can back out of the normal
Trompowsky (3 e3 or 3 e4) and
enter the Torre with 3 tbf3. This
Theory claims Black can
equalize with 4.. :iVb6. If that's true,
how can White avoid this and how
can Black obtain it?
One route is 1 d4 tbf6 2 ..tg5 c5
and then 3 d5 tbe4 4 i.f4. But
Black has to be concerned about
3 i.xf6 and 3 tbc3!? Both are
dangerous and have their own body
of must-memorize book.
153
Indians
More precise is 1 d4liJf6 2 i.g5
liJe4 and then 3 i.f4 c5 4 d5.
Black's chances of reaching the last
diagram are much greater because
the B retreats to h4less often in this
order.
Moving on, let's suppose
4.. :ilVb6 is met by 5 i.el!. (Tromp
players do things like that, they
really do.) White is hoping to build
a center with f2-f3 and e2-e4 that
justifies his loss of time. But after
5 ••. e6 he either has to play a
doubtful gambit, 6 c4 "ifb4+, or go
into 6 f3 liJf6. The latter often leads
to 7 c4 exd5 8 cxd5 c4:
5 dxc5!? Then 5.. :iia5+ 6 'i'd2!?
'i'xc5 7liJc3 favors him, e.g. 7... d5
8 0-0-0 i.e6 9 e4 or 7... d6 8 e4 a6
9 0-0-0, Berg-Ilincic, Budapest
2006.
But Black has a course correction
in 4•••'i'a5+! 5 c3liJf6. Then 6 dxc5
'i'xc5 7 e4 e5! is fine for him.
White's best is 6 d5, after which
6 ... 'i'b6! 7 i.el (not 7 'i'd2liJxd5!)
e6 8 c4 exd5 9 cxd5 c4 leaves us
back at the diagram.
One last point about the
Trompowsky: After 1 d4 liJf6
2 i.g5 liJe4 3 i.f4 d5, rather than
3... c5, White can land in a favorable
version of the Blackmar-Diemer
Gambit, of all things.
The Blackmar is dubious in its
basic form, 1 d4 d5 2 e4 dxe4
3 liJc3 liJf6 4 f3 exf3 5 liJxf3. But
with an extra tempo, 1 d4 liJf6
2 i.g5 liJe4 3 i.f4 d5 4 f3 liJf6
5 e4!? dxe4 6liJc3! exf3 7liJxf3:
This is significant for two
reasons. First it's considered by
some Trompmeisters to be Black's
best defense. He gets good play
after 9 e4 i.c5 10 liJh3 d6 11 liJd2
i.xh3 or 9 e3 i.c5 10 ~f2 0-0
11 i.xc4 l1e8. Second, it's another
position that Black can bring about
through misdirection.
Suppose White tries to avoid it
by means of 1 d4 liJf6 2 i.g5 liJe4
3 i.f4 c5 and now 4 f3 liJf6
It becomes downright respect-
able. Even Garry Kasparov played
this (in a simul) - 7... i.g4
154
Indians
8 h3 i.xf3 9 'iixf3 c6 10 0-0-0 e6
11 i.c4 lLlbd7? 12 d5! cxd5
13 lLlxd5!, with a terrific attack in
Kasparov-Cameiro, Sao Paolo
2004.
TORRE ATTACK
When Black plays ... d5 in the
Tromp or Torre, White usually has
the option of shifting to a QGD with
c2-c4, and that's often good
psychology against a 1...lLlf6
player.
3 f4 h6 4 i.xf6 'iVxf6 5 lLlc3 d6,
White gets his best play from f2-f4,
e.g. 6 'iVd2 c6 7 f4!. This means the
Torre doesn't transpose well into a
Tromp and that shift should be
reserved for use against a Black
who knows only the 1 d4 lLlf6 2
i.g5 lLle4 and 2 ... c5 lines of the
Tromp.
To minimize the danger of ... h6
White can try the delayed Torre,
1 d4 lLlf6 2 lLlf3 e6 3 c3, a favorite
of Tony Miles.
For example, when Wolfgang
Unzicker first faced the normal
Torre, 1 d4 lLlf6 2 lLlf3 e6 3 i.g5 he
responded 3 ... d5. His opponent,
Tigran Petrosian, knew that
Unzicker preferred the Slav to a
normal QGD, so he shot back 4 c4!.
Unzicker was in an unfamiliar
position and it showed after 4 ... c6
5 1rc2 i.e7 6 e3 0-0 7 lLlc3 h6?!
8 i.f4lLlbd7 9 cxd5 cxd5?! 10 i.d3.
White waits for Black to commit
Petrosian won one of his greatest
himself to a system of development
games.
(3 ... b6, 3 ... d5, 3 ... c5) before he
In the main lines of the Torre,
decides on whether to play i.g5.
Black doesn't occupy the center.
White has a reasonable Torre after
Instead he goes after White's QB,
3... d5 4 i.g5 c5 5 e3. The GM
such as via 1 d4 lLlf6 2 lLlf3 e6
preference, 3 ... b6, is also fairly
3 i.g5 h6. Then 4 i.xf6 'iixf6 5 e4
good for him after 4 i.g5.
d6 transposes to a Trompowsky in
which White has played lLlf3
For example, 4 ... i.b7 5lLlbd2 d5
instead of the more enterprising 6 e3 c5 7 i.d3 lLlbd7 8 0-0 ~c7
lLlc3.
9 i.xf6 lLlxf6 10 lLle5 and 11 f4
That's important because in the
Tromp version, 1 d4lLlf6 2 i.g5 e6
gave White an aggressive setup in
Timman-Tiviakov, Moscow 1993.
155
Indians
BENKO GAMBIT
can ignore 6 ... c5. Instead of 7 d5
The ... b5 sacrifice in a Benoni
pawn formation existed long before
the Benko Gambit, and there are
some players who offer it only after
starting with a King's Indian
Defense or Benoni. In this way they
benefit by knowing White's setup
before risking a pawn. For example,
1 d4 l2Jf6 2 c4 g6 3 l2Jc3 JL.g7 4 e4
d6 S h3 0-0 6 l2Jf3 and 6... cS 7 dS.
b5!? White should castle since
7 0-0 cxd4?! lands Black in an
inferior version of the Maroczy
Bind.
Another idea is to play ... a6
before ... c5, as in 1 d4 l2Jf6 2 c4
a6!? and 3 l2Jc3 cS. Then 4 dS bS!
S cxbS g6 is a book Benko.
There's an added benefit to this
order compared with the normal
Benko, 1 d4 l2Jf6 2 c4 c5. Then
White can obtain a promising
English Opening after 3 l2Jf3 cxd4
4 l2Jxd4 and if 4 ... a6 then 5 g3! d5
6 JL.g2.
But after 1 d4l2Jf6 2 c4 a6 3l2Jc3
cS:
Now 7... bS!? 8 cxbS a6 is a nice
version of the Benko for Black, e.g.
9 bxa6 'ifa5 10 l2Jd2 JL.xa6 11 JL.xa6
'ifxa6 12 'iVe2 e6, GheorghiuNemet, Suhr 1990.
A similar case is I d4 l2Jf6 2 c4
c5 3 d5 g6 4 l2Jc3 d6 5 e4 JL.g7
6 l2Jf3 0-0 7 h3 a6 8 JL.d3 b5!, e.g.
9 cxb5 axb5 10 l2Jxb5 l2Ja6! and
11 ... l2Jb4.
The English that arises after
4 l2Jf3 cxd4 S l2Jxd4 is excellent for
But in many other sequences Black in view of S•.. dS!. For
White has a favorable way of example, 6 cxd5 l2Jxd5 7 l2Jdb5
declining or evading the gambit. In l2Jb4! or 7 'ifb3?! l2Jxc3 8 bxc3 e5!
the Classical Variation King's 9 l2Jc2 l2Jd7 and ... l2Jc5 favored
Indian, 1 d4 l2Jf6 2 c4 g6 3 l2Jc3 Black in Filgueira-D.Gurevich,
JL.g7 4 e4 d6 5 l2Jf3 0-0 6 JL.e2 he Buenos Aires 2003.
156
Indians
However, there's always a dxc5
problem when ... a6 is played before
... c5. White gets the upper hand
after 1 d4 tiJf6 2 c4 a6 3 tiJc3 c5
4 dxc5! and then 4 ... 'i!Va5 5 a3 'i!Vxc5
6 e4 d6 7 .lte3 or 4 ... e6 5 e4 .ltxc5
6 e5.
A slightly different version is
1 d4 tiJf6 2 tiJf3 a6, used by Miles
and Lev Alburt. Then 3 c4 c5 4 d5
b5 is a gambit-less Benko. This is
mainly a psychological weapon,
designed to upset a White who
played 2 tiJf3 specifically to avoid
the Benko.
BENKO PROPER
'Benko doesn't know how to
play the Benko,' said another expert
on the opening, John Fedorowicz.
He was referring to Pal Benko's
reliance on the original move order
(1 d4 tiJf6 2 c4 c5 3 d5 b5 4 cxb5
a6 5 bxa6 and now 5....ltxa6) long
after the merits of 5 .•. g6! were
recognized.
Black can retake the pawn later,
when he knows whether it's better
to have a bishop or knight on a6.
That's important when, for
example, White fianchettos his QB,
5 ....ltxa6 6 g3 d6 7 .ltg2 g6 8 b3!.
He has good chances of making the
extra pawn count, e.g. 8 ... .ltg7
9 .ltb2 lLlbd7 10 lLlh3 0-0 11 0-0
tiJb6 12 a4.
But after 5 ... g6! White can't play
6 g3 d6 7 .ltg2 .ltg7 8 b3?? lLlfd7.
Fedorowicz pointed out that even
if White puts his queenside house in
order first, 6 b3 .ltg7 7 .ltb2 0-0
8 g3, Black can use the flexibility of
his fifth move with 8 ••• lLlxa6!
9 .ltg2 .ltb7 and then 10 lLlh3 e6! to
undermine d5. Then 11 tiJf4 tiJxd5
12 .ltxg7 cJi;xg7 13 tiJxd5 .ltxd5
14 .ltxd5 exd5 15 'i!Vxd5?? "iff6
costs material.
The trickster's answer to the
Benko begins with 4 tiJf3. He
knows that a devoted Benko-ite
doesn't want the queenside tension
to be resolved by 4 ... bxc4 or 4 ... h4.
White has a tiny positional edge,
rather than a material one, in those
lines, but it's often enough, e.g.
4 ... bxc4 5 lLlc3 d6 6 tiJd2! g6
7 tiJxc4 .ltg7 8 e4 0-0 9 .lte2 .lta6
10 0-0 lLlbd7 11 .ltd2 lLlb6 12 b3
tiJfd7 13 tiJa5! and lLlc6 as in
Rashkovsky-Meshkov, Moscow
2002.
The point of 4 tiJf3 is revealed
when White accepts the gambit on
the next move, 4 ••. g6 5 cxb5! a6
and 6 b6!?
157
Indians
12 ttJxe5 ltxb7 13 \i'a4+! ttJbd7
14 ttJc6 or 13 ... ttJfd7 14 ttJc6 'ifb6
15 i.f4!, Milov-De Vreugt, Ohrid
2001.
Then 6...'iWxb6 7 ttJc3 d6 8 e4
i.g7 is promising after 9 i.e2 i.g4
10 e5! or 9... 0-0 10 ttJd2 ttJbd7
11 ttJc4 'iic7 12 0-0 ttJb6 13 ltbl,
Burmakin-Ostertag, Bad Worishofen 2006.
If White has doubts about 6 b6 he
can still use 4 ttJf3 g6 5 cxb5 a6
with the idea of 6 e3!? This looks
like a line of the Benko that was all
the rage in the 1980s, 1 d4 ttJf6 2 c4
c5 3 d5 b5 4 cxb5 a6 5 e3, until
Black found a good antidote in an
attack on the d-pawn, 5... axb5
6 i.xb5 \i'a5+ 7 ttJc3 i.b7!.
But in the 4 ttJf3 g6 5 cxb5 a6
6 e3 order:
White's point is that these
positions more commonly come
about via 4 cxb5 a6 5 b6 and then
5.. .'iixb6 6 ttJc3 d6 7 e4 g6 8 ttJf3
iLg7. But that order gives Black a
very promising alternative in
5... e6!? with good center play. This
option isn't possible after 4 ttJf3 g6
5 cxb5 a6 6 b6.
Black gets little out of 6... axb5
Another benefit is that Black
i.xb5 \i'a5+ 8 ttJc3 i.b7?
7
often replies 6... a5?!. He does this
out of habit - because it's a good because White can simply castle
move in the comparable 4 cxb5 a6 and keep his extra pawn (9 0-0 i.g7
5 b6 a5 position. Then Black can 10 a4 0-0 11 e4).
playa quick ... iLa6 and decide later
Instead, Black should play
whether to retake on b6 with the 8...i.a6! 9 iLxa6 \i'xa6. Then 10 e4
queen or QN.
d6 11 e5 needs more testing
(11...dxe5 12 ttJxe5 i.g7 13 \i'd3,
But in the delayed form White
Balgojevic-S.Kasparov,
Curto
gets the upper hand from 7 ttJc3
2006).
iLa6 and now 8 b7! i.xb7 9 e4 or
8... %:ta7 9 e4 iLxfl 10 ~xfl d6
And what happens if Black meets
11 e5!. For example, 11...dxe5 4 ttJf3 with 4... a6 instead of 4... g6 ?
158
Indians
This transposes to 1 d4 liJf6 2 liJf3
a6 3 c4 cS 4 dS bS. Then on 5 e3
two thematic ideas of the Benko,
S... b4 and S... dxc4 make little sense
because he's played ... a6.
And then 3 c4 exd5 4 cxd5 liJf6.
The benefit of 1 d4 cS is this time
2 liJf3 allows Black to seize center
ground with 2 ... cxd4 3 liJxd4 dS!,
as Alexander Alekhine showed.
For example, S... bxc4 6 liJc3 d6
7 e4 g6 8 .ltxc4. Black has an extra
. tempo compared with 4 liJf3 bxc4
S liJc3 d6 6 e4 g6 7 .ltxc4. But the
extra move is ... a6, which is more
harmful than useful.
But this order has a major minus
since White hasn't committed
himself to c2-c4. He can play a
more open game with 3 liJc3. Then
3...liJf64 e4 exd5? 5 e5! favors him
considerably and 4 ... d6 S .ltbS+ and
6 dxe6 less so.
A better irregular route to the
Modern Benoni is 1 d4 e6 2 c4 c5
A Benoni is distinguished by a 3 d5. This avoids the Trompowsky
White pawn on dS and a Black one and makes sure c2-c4 is played
at cS. There are several kinds of before Black commits to ... cS.
Benoni depending on whether Another benefit is evident after
White supports dS with c2-c4 and 3 ... exd5 4 cxd5 d6 5 liJc3 g6 6 e4
whether Black tries to liquidate it .ltg7:
with ... e6.
THE BENONI FAMILY
Among the irregular versions is
one that arose in the 1970s when
White frequently met I d4 liJf6 2 c4
cS with 3 liJf3. Black began to use
an anti-English order, 1 d4 c5, so
that he can transpose to a Modern
Benoni after 2 d5 e6.
Black can transpose into main
lines after 7... liJf6.But in case of
7 f4 it's a bad decision, as 7...liJf6
8 .ltb5+! is known to favor White.
Instead 7.. Ji'h4+! 8 g3 'ile7 has
stood up well in tests (9 .ltg2 liJf6
10 liJge2 0-0 11 0-0 liJe8 12 .lte3
159
Indians
tiJd7 13 'iVd2 :b8 14 a4 a6 15 :abl
f5, Skembris-Hamdouchi, Bastia
2002).
such as 1 d4 tiJf6 2 tiJo c5 3 d5 d6
and 1 d4 c5 2 d5 d6 3 e4:
Another virtue of this order is
that on 7 tiJo Black has time for
7•.• a68 a4 .tg4. This is a common
Benoni theme designed to gain
control of e5. If Black had
transposed with 7 ...tiJf6 White
could avoid ....tg4xf3 by means of
8 tiJd2.
But there is an obvious demerit
to 1 d4 e6 2 c4 c5. Black has to have
an answer to 2 e4.
Black often continues 3... g6
followed by ... i..g7 and ... tiJf6. But
3 ...tiJf6 4 tiJc3 g6 is more accurate.
SCHMID BENONI
This avoids 3... g6 4 tiJf3 .tg7
5 .tb5+ tiJd7 6 0-0 tiJf6 7 l1e 1
This became fashionable, in the when White can profit from
order 1 d4 c5 2 d5 d6 3 tiJc3 g6, delaying tiJc3, e.g. 7... 0-0 8 a4! a6
after Lothar Schmid scored 9 .tf! b6 10 h3 tiJe8 11 tiJa3! and
impressive wins with it in a 1954 tiJc4/e4-e5 with advantage, as
correspondence tournament. For Mikhail Tal showed.
example, 4 e4 .tg7 5 tiJf3 tiJf6
The bishop check is a recurring
6 .te2 tiJa6 7 0-0 tiJc7 8 tiJd2. By
problem in Schmid lines. To avoid
delaying ... 0-0 he had time for 8... a6
it Black can tweak the order, 1 d4
9 a4 .td7 10 tiJc4 b5 11 tiJb6 b4!
c5 2 d5 tiJf6 3 tiJc3 g6 and then
12 tiJxa8 'iixa8 13 tiJbl tiJxe4 with 4 tiJo .tg7 5 e4 0-0.
advantage in Hayes-Schmid.
This also enables him to
A benefit of the Schmid order is transpose into more normal Schmid
that if White chooses 3 c4 instead of lines with 6 ... d6. Or he could go
3 tiJc3, Black has his choice of into sharp lines such as 6 i..e2 b5!?
playing a KIDlBenoni hybrid or (7 e5 tiJg4 8 i..f4 b4! 9 tiJe4 d6).
transposing to a Modem Benoni
The dangers in this order come
with 3 ... g6 4 tiJc3 .tg7 5 e4 tiJf6
from an advance of White pawns
6 i..e2 0-07 tiJf3 e6 and ... exd5, as
such as the immediate 4 e4 i..g7
Schmid did.
5 e5!? or 4 tiJo i..g7 5 e4 0-0 6 e5
Another benefit is that Black can tiJg4 7 tiJg5!? and 6 d6 'iWb6 7 i..f4
experiment with different orders, 'iixb2.
160
Indians
KID/BENONI
This hybrid is similar to the
Schmid Benoni except that White is
committed to c2-c4. This is usually
due to the order 1 d4 ttJf6 2 c4 c5
3 d5 d6. The finesses generally
concern what Black does with his
e-pawn. For example, 4 ttJc3 g6
S e4 ..tg7 and 6 ttJf3 0-0 7 ..te2 e6
80-0.
Now 8... exd5 9 cxd5 transposes
into a very familiar Modem Benoni
position that is rated somewhere
between equal and plus-over-equal.
But Black can delay further with
8...:e8!?
That leaves 9 ttJd2, which
defends e4 and can reach a Modem
Benoni tabia (9 ... exd5 10 cxd5).
Black may feel happy that in this
way he avoids the ~c2 line. That is,
1 d4 ttJf6 2 c4 c5 3 d5 e6 4 ttJc3
exd5 5 cxd5 d6 6 e4 g6 7 ttJf3 ..tg7
8 ..te2 0-0 9 0-0 lte8 and now
10 ~c2 rather than 10 ttJd2.
But Black can delay further,
9 ttJd2 ttJa6!. He can spend the next
few moves preparing ... b5 with
... ttJc7, ... a6, ... l1b8 while White
doesn't have equally useful moves.
Once again 10 dxe6? is dubious.
After 10... ..txe6 11 f4 ..td7!
White's
vulnerable
e-pawn
becomes the main issue. e.g.
12 ..tf3 ..tc6 13 ttJd5 ttJb4 and
13 l1e 1 'iVb6 14 ttJb3 l:!ad8 15 ..td2
ttJb4 (Ardiansyah-Schmidt, Indonesia 1983) favors Black.
The delayed Modem Benoni is
still largely unexplored and the
indications are that it's worth
exploring.
This is more than a waiting move
because he is preparing 9... exd5
10 cxd5? ttJxe4. On quiet moves
such as 9 h3, Black gets a fine game
with 9... exdS 9 exdS ..tfS.
And if White tries to exploit the
delay in ... exd5, by means of9 dxe6
..txe6 10 ..tf4, Black has adequate
compensation for the lost pawn
after 10...ttJc6! 11 ..txd6 ttJd4. Or
so experience indicates.
MODERN BENONI
The Modem Benoni comes about
from two main orders, 1 d4 ttJf6
2 c4 c5 3 d5 e6 followed by
... exd5/... d6 and ... g6, and 2 ... e6 and
3... c5 (4 d5 exd5 5 cxd5 d6).
In both cases White can shift into
the English Opening with ttJf3
instead of pushing d4-d5. A virtue
of 1 d4 ttJf6 2 c4 cS is that it grants
161
Indians
Black the extra option of meeting
3 liJf3 with 3... cxd4 4 liJxd5 e5!?,
one of the most enterprising antiEnglish lines.
i.a6 15 'iVb3! exf4 16 i.xf4 i.xc3
17 'i!Vxc3 'i!Vxe2 18 ltae1 with a
powerful attack.
Objectively, a good solution to
To avoid that White may prefer White's order is 6... d5!, a better3 e3. A BenkolBenoni specialist can than-usual Gruenfeld. But to play it
become confused after 3... g6 4 liJc3 Black has to know complex theory
i.g7 5 liJf3 0-0 6 i.e2.
such as 7 dxc5 'ifa5 8 cxd5 ttJxd5!.
That helps make 3 e3 a more
credible weapon against a Benko/
Benoni-booked player.
Today you're more likely to see
the other Modem Benoni order,
1 d4 liJf6 2 c4 e6 and 3 ... c5
because it contains the Nimzothreat (3 liJc3 i.b4!). That often
prompts 3 ttJf3 and enables Black,
after 3 ... c5 4 d5, to avoid the
Now 6... cxd4 7 exd4 d5 is a
aggressive Modem Benoni lines in
reversed Tarrasch QGD with an
extra move for White. That sounds which White plays f2-f4 or f2-f3.
more promising than it really If Black is also happy with
is, e.g. 8 0-0 liJc6 9 i.e3 dxc4 Symmetrical English positions such
10 i.xc4 liJa5 with equal chances.
as 3 liJf3 c5 4 liJc3 cxd4 5 ttJxd4,
this is his most pleasant route to the
Instead, a more Indian-like Modem Benoni.
approach, 6... d6 7 0-0 b6?! 8 e4!,
If White tries 1 d4 liJf6 2 c4 e6
drifts toward another poor Maroczy
Bind (8 ... cxd4 9 liJxd4 and 10 f3) 3liJf3 c5 4 e3:
or a Benoni or KID in which
... b6/ ... i.b7 only blocks Black's
queenside counterplay.
For example, 8... i.b7 9 d5! e5
10 liJe 1 liJe8 11 liJd3 suddenly
becomes a dubious King's Indian.
This was evident in BisguierKavalek, Tel Aviv 1964 after 11...f5
12 exf5 gxf5 13 f4 ~e7 14 i.e3
162
Indians
This is a tacit invitation to a
Semi-Tarrasch QGD (4 ... d5). A true
Benoni player will prefer Mihai
Suba's 4... a6 5 tiJc3 "iic7 followed
by ... d6/ ... tiJbd7 and eventually
... g6/ ... iLg7.
All of those moves are useful to
7... iLd7 8 'Yib3 is nice for him and
7 ... tiJbd7 8 tiJfd2! imperils the
g4-bishop, e.g. 8... a6 9 h3 b5
10 "iic2 iLh5 11 a4! b4 12 tiJc4 and
iLf4. Or 11...iLg6 12 axb5 tiJxd5
13 bxa6 and White won in
Dokhoian-Lukov, Plovdiv 1988.
Black in a Benoni. That means he'll
be in great shape when White
recognizes that his only chance for
an edge lies in d4-d5, e.g. 6 iLe2 d6
70-0 tiJbd7 8 d5 exd5 9 cxd5 b5.
Or 8 a4 g6 9 d5 exd5 10 cxd5
iLg7 11 e4 0-0 (Tukmakov-Suba,
Erevan 1980) - when Black has
actually gained a move on a normal
Benoni.
White can try to exploit the
Nirnzo-threat order, 1 d4 tiJf6 2 c4
e6 3 tiJf3 c5 4 d5 exd5 5 cxd5 d6,
with 6 e4!?
Of course, Black has an easier
time engineering ... b5 in this order.
However, that push doesn't solve
his problems in 6.•. g6 7 h3 iLg7
8 iLd3 0-0 9 0-0, e.g. 9 ... b5 10 :el!
a6 11 a4 b4 12 tiJbd2 or 10 .. J~e8
11 iLxb5 :'xe4 12 tiJc3 :'xe1+
13 "iixe 1 a6 14 iLa4 and iLf4, as in
I.Sokolov - Papaioannou, Turin
2006.
There's a simple counter to a
delayed tiJc3. After 1 d4 tiJf6 2 c4
e6 3 tiJf3 c5 4 d5 Black should play
4... d6! rather than 4 ... exd5.
This is more than a one-move
trap (6 ...tiJxe4?? 7 'iVa4+). White is
holding his QN back until he knows
whether it belongs on c3, a3 or d2.
White can't play 5 e4? tiJxe4 this
time and 5 tiJc3 allows Black to
transpose, with 5... exd5 6 cxd5, to
normal lines.
He also benefits from tactics
such as 6...iLg4 7 'iVa4+!. Then
This also gives White reason to
think about 6 tiJxd5 tiJxd5 7 'iVxd5.
163
Indians
8 tiJd2) in which Black has been
finessed into a ... tiJbd7 line, rather
than one of the more popular ... tiJa6
or ... i.g4 variations. There's no
downside to White's order - or at
least we haven't discovered it yet.
But the backward d-pawn is not the
weakling it seems after 7•..tiJc6, e.g.
8 e4 i.e6 9 "ii'dl i.e7 or 8 tiJg5
'ike7 8 'iVe4 h6 as in the forgotten
game Bandal-Fischer, Meralco
1967.
In the Modem Benoni's main
OlD-in to-KID
line, 1 d4 tiJf6 2 c4 cS 3 dS e6
Some people actually like the
4 tiJc3 exdS S cxdS d6 6 tiJf3 g6,
Old
Indian. After 1 d4 tiJf6 2 c4
the move 7 tiJd2 was considered a
virtual refutation when the opening they voluntarily play 2... d6 and
was introduced in Nimzovich- 3... eS or 3••.tiJbd7 and 4... eS. But
Marshall, New York 1927, e.g. others use the Old Indian order as a
7 ... i.g7 8 tiJc4 and 9 i.f4 or safe path into a King's Indian. The
7... tiJa6? 8 tiJc4 tiJc7 9 a4 i.g7 early ... e5 usually discourages
White from sharper lines, such as
10 i.f4!.
the Samisch Variation. That shrinks
But the Benoni was rescued by
the amount of theory Black has to
the realization that 7 ... tiJbd7 8 tiJc4
know.
tiJb6 is a good response. Some
This finesse is a century old.
annotators gave '7 tiJd2?!' and
Janowsky-Tchigorin,
Ostende 1907
'7 ... tiJbd7!' when those moves
occurred in the 1972 Spassky- went 1 d4 tiJf6 2 c4 d6 3 tiJc3
Fischer match. Nevertheless, 7 tiJd2 tiJbd7 4 tiJf3 eS S e3.
has transpositional value because
after 7...tiJbd7:
White can play 8 e4!. Then he's
reached the Classical Variation
(usually occurring via 7 e4 i.g7
There followed S.•.i.e7 6 i.d3
and Black soon got a bad game
(6 ... c6 7 0-0 'ikc7 8 i.d2 0-0 9 l:tc1
l:te8 10 tiJg5 tiJf8?! 11 'iVc2 h6
12 tiJge4 i.e6 13 d5! i.c8 14 f4!).
164
Indians
But when the diagram arose in
another Ostende 1907 game, Black
(Aron Nimzovich) deviated with
5... g6!. He tricked White into one of
the most innocuous of KID
variations and equalized after
6 .te2 .tg7 7 0-0 0-0 8 b3 :e8.
For instance, 7... exd4 8 tiJxd4
0-0 9 'iYd2 and O-O-O! or 7... h6
8 .te3 0-0 9 'i'id2 'it>h7 10 d5! a5
11 g4! tiJg8 12 h4 tiJc5? 13 h5 g5
14 tiJg3 with a big positional
plus (Kohlmeyer-Lymar, Balaguer
2006).
White can insist on playing a
Samisch setup, after say 4 e4 e5
5 d5 tiJc5 6 fJ. But the point of this
order is revealed when Black
changes his mind about the KID.
After 6••• a5 7 .te3 .te7! was played
in the 1954 USSR Championship,
the tournament book said this was
'an openmg accuracy even
M.l. Tchigorin knew. '
A sly move like 5 tiJge2 deserves
a sly reply, 5... c6. Then 6 f3 .te7!
transposes to a pretty good Old
Indian, since 7 .te3 0-0 8 'iYd2
allows 8... d5! (9 cxd5 cxd5 10 exd5
exd4 11 .txd4 tiJb6). Or 8 tiJg3 d5!
9 cxd5 cxd5 10 exd5? .tb4 with an
edge.
The traditional counter to 1 d4
tiJf6 2 c4 d6 3 tiJc3 tiJbd7 is 4 tiJfJ
but few players realize why. It stops
4... e5 5 .tg5 g6? because of 6 dxe5
dxe5 7 tiJxe5!. If instead Black
chooses an Old Indian (5 ... .te7),
it's an OlD that is quite good for
White after 6 e3 and .te2!'iic2.
Black's counter-finesse to 4 tiJf3
is simply 4.•. g6.
By avoiding ... g6 there is no
object for the h2-h4-h5 attack.
Black drew after 8 'ifd2 0-0 9 tiJge2
tiJe8 10 g4 h6! 11 0-0-0 .tg5!
12 .txg5 'ifxg5.
White can be subtle, too. After
1 d4 tiJf6 2 c4 d6 3 tiJc3 tiJbd7 4 e4
e5 he can try 5 tiJge2 and wait for
Black to commit his KB. Then
5... g6 6 f3 .tg7 7 .tg5! is a betterthan-usual Samisch.
Then if White continues in OlD
mode (5 .tg5 .tg7 6 e3) he finds
165
Indians
himself in an offbeat KID, 6•.• h6
7 i.h4 g5 S i.g3 ttJh5, that is fairly
equal.
The downside of 4... g6 is that
Black usually ends up, after S e4 or
S g3, in a Classical or Fianchetto
Variation of the KID in which he's
relinquished some of his most
aggressive weapons, involving
For example, 7 i.e3 bxc5
... ttJc6, and has made ... cS less
S i.xc5 ttJc6 9 J:[c1 J:[bS 10 'iWd2
effective. But the upside of
J:[e8! 11 ttJge2 d6 12 i.e3 e6
avoiding the Samisch, et aI, may be
13 ttJf4 d5!? 14 cxd5 exd515 exd5
more important.
ttJb4 with a fierce initiative in
On the other hand, if Black really Lomaya-Gufeld, Tbilisi 1961.
wants to play the Old Indian, 1 d4
If White declines the gambit with
ttJf6 2 c4 d6 3 ttJc3 ttJbd7 and 6 dS d6 7 i.e3 he transposes into
4 ttJf3 c6! is most precise. In that another Samisch gambit, 4... d6 S D
way he can transpose to the main 0-06 i.e3 cS!? (7 dS). The counterOlD lines after S e4 eS and 6... i.e7 finesse to 4... 0-0 is S i.e3, which
but he avoids 4 ... eS S i.gS! and
allows White to transpose into a
6 e3.
normal Samisch after S... d6 6 D.
KING'S INDIAN DEFENSE
After the traditional KID order,
1 d4 ttJf6 2 c4 g6 3 ttJc3 i.g7 4 e4,
Black is not required to play 4... d6.
With 4... 0-0 he tries to lure White
into overextending himself, 5 e5
ttJeS. The pawn push hasn't
recovered from Letelier-Fischer,
Leipzig 1960 (6 f4 d6 7 i.e3 cS
8 dxcS ttJc6!).
Aside from provocation, 4... 0-0
serves Black by creating a gambit,
5 f3 c5!? and then 6 dxc5 b6.
Black has to be careful about
S i.gS because of the temptation in
KIDs to kick the bishop as soon as
it lands on gS. If he had played
4 ... d6 instead of castling, S... h6!
(6 i.e3 ttJg4) is recommended
before White can stop it with
6 'iWd2.
However on 4... 0-0 5 i.g5 h6?!
6 i.e3 Black lacks 6... ttJg4. The
result is an inferior Samisch, e.g.
6... d6 7 D! eS 8 dS c6 9 'iWd2 cxdS
10 cxdS ttJe8 11 0-0-0 ~h7 12 g4!,
Ushenina-Siti, Kochin 2004.
166
Indians
11 0-0, e.g.ll ... a5 12 'ilVc2 "ikc7
13 l'itadl tiJc5 14 %:lfel tiJfd7 15 f4
tiJe6 16 "ikd2 tiJxd4 17 .i.xd4,
Rogozenko - Dragomirescu, Timisoara 2006.
On the opposite side of the coin,
Black can delay rather than
accelerate castling. This gives him
an extra tempo for development and
denies White a kingside target, e.g.
1 d4 tiJf6 2 c4 g6 3 tiJc3 .i.g7 4 e4
d6 S f3 tiJc6!? and 6 .i.e3 a6
7 tiJge2 %:lbS.
What has happened is White
reached a favorable position that
usually comes about from 6 0-0
tiJbd7 7 tiJc3 e5 8 e4 c6 9 h3 %:le8
and ... exd4, a line Black gave up on
decades ago. In the 6 tiJc3 order
Black lost his chance to play better
lines.
That tempo enables Black to
meet S tiJc1 eS 9 dS tiJd4 10 tiJb3
with 10... cS! 11 dxc6 bxc6 with
b-file play, e.g. 12 tiJxd4 exd4
13 .i.xd4 %:lxb2. After 9 tiJb3 Black
can transpose into normal lines with
9... exd4 10 tiJxd4 0-0.
White can also profit from
delayed castling. In the Fianchetto
Variation, 1 d4 tiJf6 2 c4 g6 3 g3
i.g7 4 .i.g2 0-0 sllJf3 d6 he can try
Viktor Korchnoi's favorite 6 tiJc3
(instead of 6 0-0).
Then 6...tiJbd7 7 e4 eS S h3
offers to transpose into the
conventional ... c6 lines, such as
8... c6 9 .i.e3 exd4 10 tiJxd4 %:le8
The main drawback to tiJc3 in
place of 0-0 is it may make ... c5
stronger, as we'll see in a few
pages. But if White doesn't mind
... c5 lines, this order makes sense.
KID: SAMISCH
After 1 d4 tiJf6 2 c4 g6 3 tiJc3
.i.g7 4 e4 d6 S f3 0-0 Black usually
chooses between breaking in the
center with ... c5 or with ... e5. The
first seemed harder to achieve until
it was discovered in the 1980s that
6 .i.e3 c5 is a sound gambit. That
told White he should consider other
sixth moves, such as 6 tiJge2.
It's a good waiting move since
there are only a few Samisch lines
in which White develops the knight
elsewhere but there are several in
which .i.g5 serves better than .i.e3,
e.g. 6 tiJge2 eS?! 7.i.gS!.
167
Indians
On 7... c6 8 'iVd2 tiJbd7 9 dS!
Black cannot easily play the
thematic ... ~e8/... f5, e.g. 9 ... cxd5
10 ~xd5 or 9 ... ~b6 10 ~g3 'iVc7
11 l:.cl is awful (Dreev-Mork,
Linares 1999).
Well, then, how good is 6 ~ge2
cS 7 dS ? Black can seek a doubleedged Modern Benoni, 7 ... e6 8..tg5
exd5 9 cxd5.
However, 8 ~g3 retains more
options for White, since on 8 ... exd5
9 cxd5 his bishop may belong on f4
(e.g. 9 ... ~a6 10 ..te2 ~c7
11 0-0..td7 12 a4 a6 13 ..tf4! ~fe8
14 'ilVd2, with advantage in
Khismatullin- N echepurenko,
Serpukhov 2004).
Now 7 ..td3 can be answered by
7 ... cS!. Then 8 dxc5 dxc5 9 ..txc5
~c6 is an improved version of the
6 ..te3 c5 gambit and 8 d5 e6! is a
good Benoni.
Black's main idea is that 7 'iVd2
can be met by 7 ... c6!. ECO
considered 8 0-0-0 b5 and 8 ..th6 b5
to be equal - and among the few
Byrne Variation lines that don't
favor White.
A similar idea was used by Tal in
the 1961 world championship when
he was trailing badly. He wanted to
play an ... e5/... c6 line but was afraid
of 5 ... 0-0 6 ..te3 e5 7 dxe5! dxe5
8 'iVxd8!, when Black has the worst
of a likely draw.
Tal refined the order with 6... c6,
feinting in the direction of
... a6/ ... b5. But after Botvinnik
Against Robert Byrne's idea of replied 7 ..td3, he continued 7 ... e5!
... c6/ ... a6/ ... b5, a conventional view and got the opening he wanted .
holds that ..td3 and kingside
Before leaving the Samisch, we
castling is a better policy than should mention S tiJge2. Books tell
preparing a mating attack with 'iVd2 you that White's idea is 6 tiJg3 and
and 0-0-0. Black can try to exploit they add that it only equalizes after
S... eS 6 dS tiJa6.
that via S f3 0-0 6 ..te3 a6.
168
Indians
However 6... ttJbd7 can be tested
by 7 e5!? The key line was long
regarded to be 7... dxe5 8 dxe5 ttJg4
9 e6 fxe6 10 0-0 ttJde5, and while
experience indicates Black is equal
(11 'iib3 ttJxf3+ 12 .txf3 ttJe5
13 .te2 ttJc6! and ... ttJd4), the
positions that arise can be difficult
to handle.
That may be true after 7 ttJg3
h5 and 8... h4. But 7 f3! and
8 .tg5 reaches a better-than-usual
Samisch.
The younger - and wiser sibling of 6... ttJbd7 is 6...ttJa6! and
then 7 0-0 e5.
Black should ignore the books
and play 5... 0-0 and then perhaps
6 .tg5 ttJbd7 7 'iid2 c5 since 8 d5
b5 9 cxb5 a6 is a good Benko
(Damljanovic - Martinovic, Smederevska Palanka 1999).
KID: CLASSICAL
One database shows only eight
examples of this played before
1988. In most cases Black will be
able to transpose into 6 ... e5 7 0-0
ttJbd7 positions that are quite good
for him, e.g. 8 l:te 1 exd4 9 ttJxd4
ttJc5 or 8... c6 9 d5 ttJc5! and 8 .te3
c6 9 d5 ttJg4 10 .tg5 f6 11 .td2 f5
That's a vote in favor of followed by ... ttJc5.
6...ttJbd7 and then 7 0-0 e5. This
The benefits of having a knight at
order also denies White the
Petrosian Variation (6 ... e5 7 d5 and a6 show up when we don't
.tg5 and ttJd2). White can't reach it transpose. White gets little from the
after 6... ttJbd7 7 0-0 e5 because endgame, 8 dxe5 dxe5 9 'iixd8
8 d5 ttJc5 attacks the e-pawn l:.xd8, because 10 ttJd5 just loses
the e-pawn and because the
(9 .tg5 h6! 10 .th4? g5).
The most common Classical
Variation order is 1 d4 ttJf6 2 c4 g6
3 ttJc3 .tg7 4 e4 d6 5 ttJf3 0-0
6 .te2. Then 6... e5 allows White to
force a riskless, slightly favorable
endgame, 7 dxe5 dxe5 8 'iixd8
l:.xd8 9 .tg5.
169
Indians
preparatory 10 ii..g5 is adequately
Black can counter 6 ii..e3 with
met by 10... c6 or 1O ... h6. In all of 6... c5. This is based on 7 dxc5 'i'a5
the endgame lines, tbxe5 is and 7 .te2? tbg4 8 ii..g5 cxd4
answered by ... tbxe4!.
9 tbxd4 'i'b6!. White's best may be
7 d5, reaching a double-edged
Nevertheless, books recommend Benoni after 7... e6 8 h3 exd5
the orthodox 6 ... e5 7 0-0 tbc6 9 exd5. If that's the worst case,
instead, with analysis that can run 6 ii..e3 may be worth White's while.
25 moves deep. If White doesn't
like the mega-theory, he should
KID: FIANCHETTO
consider a Bent Larsen idea, 1 d4
The fianchetto order that has
tbf6 2 c4 g6 3 tbc3 ii..g7 4 e4 d6
earned the greatest currency is 1 d4
5 tbf3 0-0 6 ii..e3.
tbf6 2 c4 g6 3 g3 ii..g7 4 ii..g2 0-0
5 tbf3. There is more room for
finesse when White delays pushing
the c- or d-pawns, as in 1 d4 tbf6
2 tbf3 g6 3 g3 ii..g7 4 ii..g2 0-0
5 0-0:
Larsen's intent was to escape
book (6 ... tbbd7 7 h3). But 6 ii..e3
can be used as a way to stay in
book, but on another page.
The reason is that when Black
White can decide, after 5...d6,
sees 6 .te3 he often figures 6 ... e5
whether to make it a KID, with 6 c4
7 dxe5 is a superior form of the
or jump into the Pirc Defense
6 ii..e2 e5 7 dxe5 endgame. As a
(6 tbc3 and 7 e4). Against a player
result he prepares ... e5 with
who rarely if ever adopts the Pirc,
6... tbbd7 or 6... tba6 instead.
6 tbc3 can be strong.
But then 7 .te2! e5 8 0-0 allows
White to transpose into normal
... tbbd7 or ... tba6 lines - having
dodged the ... tbc6 mega-theory.
That was the case in LangewegVasiukov, Wijk aan Zee 1973:
6... tbbd7 7 e4 e5 8 a4 exd4 9 tbxd4
l:te8 10 a5! tbc5 11 l:tel tbg4?!
170
Indians
12 h3 4JeS 13 f44Jed7 14 .lte3 4Jf8
IS 'ti'd2 a6 16 b4! 4Jcd7 17 4Jb3
with a clear edge.
But Black can hide his intentions
with 5... c6!? Then on 6 4Jc3 d5! he
should equalize (7 b3 .ltfS 8 iLb2
4Je4!). The ... dS idea is naturally
stronger when Black hasn't spent a
tempo on S... d6.
If White has advanced his cpawn and not the d-pawn, as in 1 c4
4Jf62 4Jf3 g6 3 g3 .ltg7 4 .ltg2 0-0
5 0-0, then 5•.. c6 is what Tal called
'a slight psychological diversion.'
Black gets a good version of the
Neo-Gruenfe1d with 6 4Jc3 dS!,
threatening ... d4, and then 7 cxdS
cxdS 8 d4 4Je4!.
White can try for a reversed
version of that with 6 d4 so that his
knight is the one in the center after
6 ... dS 7 cxdS cxdS 8 4JeS!. But then
Black can go back to the KID with
6•••d6, transposing into a position
pioneered by Lubos Kavalek.
comes about via 1 d4 4Jf6 2 c4 g6
3 g3 .ltg7 4 iLg2 0-0 5 4Jf3 d6
6 0-0 4Jbd7 7 4Jc3 eS 8 e4 c6
9 h3 'it'aS. The major difference is
that Black has avoided a heavily
analyzed alternative in the second
order, 9 .lte3!? 4Jg4 10 .ltgS.
This order also enables Black,
after 7 4Jc3 'ii'aS, to attack the
c-pawn (8 dS 'ii'b4). In recent years
8 h3 'ii'a6 has had some vogue. Joel
Benjamin proposed a more accurate
move order beginning with
7...'iib6.
This transposes after 8 d5 'iVb4 or
8 h3 'it'a6. But 8 e4, which is a
central line in the Kavalek order, is
easier for Black to meet here with
8.•.iLg4!. Then 9 h3 .ltxf3 10 .ltxf3
4Jfd7 11 .lte3, which is played
against the Kavalek, would hang
the b-pawn. And 9 b3 cS! equalizes,
e.g. 10 dxcS dxcS 11 .lte3 4Jc6
12 h3 :ad8, Tregubov-Izoria,
Moscow 2006.
Black can insert ... cS at a number
of points in the first half dozen or so
moves of a Fianchetto Variation.
Some KID experts do it as soon as
White declares his preference for
g2-g3, e.g. 1 d4 4Jf6 2 c4 g6 3 g3
cS.
Then 7 4Jc3 'it'a5 and 8 h3 e5
9 e4 4Jbd7 is a main line that also
But others wait for 4Jc3, such as
1 d4 4Jf6 2 c4 g6 3 g3 iLg7 4 .ltg2
0-0 5 4Jf3 d6 6 4Jc3 and now 6.•. c5
7 d5 e6!.
171
Indians
White can try to punish Black's
last move with 8 dxe6 because his
pawns are loose after 8 ... fxe6 and
because 8 ... .1i.xe6 is a sacrifice
(9 l2Jg5 .1i.xc4 10 iLxb7).
Here 7 ... b5!? 8 cxb5 a6
transposes into the Fianchetto
Variation of the Benko. That's a
solid line for White but no threat to
refute the Benko.
But after Korchnoi and Rafael
Vaganian lost celebrated games to
the sacrifice, theory said White
should avoid this and merely
transpose to a quiet Benoni (8 0-0
exd5 9 cxd5).
KID: .1i.g5
However, if White has played
6 0-0 instead of 6 l2Jc3, then 6 ... c5
7 d5 e6 8 dxe6! is strong because
8... iLxe6 9 l2Jg5 iLxc4 10 .1i.xb7
l2Jbd7 is rudely surprised by
11 l2Ja3!. This gave Korchnoi two
easy victories when he introduced it
(11...!:.b8 12 l2Jxc4 llxb7 13 l2Jxd6
with a clear extra pawn). This gives
White a major reason for delaying
l2Jc3.
The downside is allowing Black
a good version of the Benko
Gambit, e.g. 1 d4 l2Jf6 2 c4 g6 3 g3
iLg7 4 iLg2 0-0 5 l2Jf3 d6 6 0-0 c5
7 d5:
In Yuri Averbakh's variation,
1 d4 l2Jf6 2 c4 g6 3 l2Jc3 .1i.g7 4 e4
d6 5 .1i.e2 0-0 6 .1i.g5, Black cannot
play 6 ... e5? 7 dxe5 dxe5 8 'i'xd8
and 9 l2Jd5. His main choice is
between 6 ... c5 7 d5 h6 8 .1i.e3 e6
and 6 ... h6 7 .1i.e3 c5 8 d5 e6, which
transpose.
Both orders have merit. Svetozar
Gligoric preferred 6 ... h6 because
after 7 .1i.h4 c5 8 d5 Black has a
strong alternative in 8... a6 9 a4
'i'a5, since .1i.d2 is not possible. On
7 .1i.e3 e5 (or 7 ... l2Jbd7 8l2JO e5) he
reaches a Classical Variation
position in which the extra tempo,
... h6, is more asset than liability.
Other authorities prefer 6 ... c5
7 d5 h6. But to play it Black needs
to know a huge amount of theory
about the pawn sacks that arise after
172
Indians
KID: FOUR PAWNS
8 i.f4 e6 9 dxe6. In contrast, 6... h6
7 i.f4 can be met by 7...t2Jc6 (8 12Jf3
12Jxd4! and 8 d5 e5!). That makes
6... h6 the mental hygiene order.
In the Four Pawns variation (1 d4
12Jf6 2 c4 g6 3 12Jc3 i.g7 4 e4 d6
S f4) Black usually challenges the
center with ... c5 and White
typically responds d4-d5. The
timing of ... c5 doesn't matter
unless White allows ... cxd4. After
S••. cS 6 12Jf3!? cxd4 7 12Jxd4 12Jc6
8 i.e3:
Another i.g5 system is 5 h3 0-0
6 i.gS.
Then 6... h6 7 i.e3! and 'iVd2
favors White because Black cannot
play 7... l2Jg4 and because this time
... h6 is more liability than asset,
e.g. 7... c6 8 'iYd2 ~h7 9 i.d3 e5
10 d5 12Ja6 11 g4 with advantage,
Paunovic-Illic, Belgrade 2004.
Theory used to regard 6••• cS as
good because 7 dS e6 8 i.d3 exdS
9 exdS is a somewhat lifeless
Benoni. But 9 cxdS! transposes into
a good one, e.g. 9... l:.e8 10 12Jf3 a6
11 a4 'iYa5 12 0-0 12Jbd7 13 :e1
'iYb4 14 'iYc2 c4 15 i.fl, BareevPredescu, Kitchener 2006.
Note that since ... h6 and i.h4
haven't been played, all lines with
... 'iYa5 can favorably be met by
i.d2, e.g. 6... c5 7 d5 a6 8 a4 'iYa5
9 i.d2! e6 10 i.d3 or 7... b5 8 cxb5
a6 9 a4 'iYa5 10 i.d2 with
advantage.
Black can get dynamic play with
8 •.. l2Jg4!, e.g. 9 12Jxc6 bxc6
10 i.g1 e5! or 9 i.g1 12Jxd4
10 i.xd4 e5.
This is better than 8...0-0 9 i.e2!
which stops ... l2Jg4. White gets a
somewhat favorable middlegame or
a good endgame like 9 ... e5 10 12Jxc6
bxc6 11 fxe5 dxe5 12 i.c5 and
'iYxd8 and the related 9 ... l2Jxd4
10 i.xd4 e5 11 fxe5 dxe5 12 i.c5.
The reason this is significant is
that if Black plays 5... 0-0 instead of
5... c5, White can safely play 6 12Jf3
c5 7 i.e2!? cxd4 8 12Jxd4, forcing
his way into the good line of the last
paragraph.
173
Indians
Therefore 5 ... c5 is more precise.
White can avoid it with a counterfinesse, S Jie2!. Then if Black
assumes he's headed for a Classical
Variation, 5 ... 0-0 6 ttJf3, he'll be
surprised by 6 f4 c5 7 ttJf3, slipping
into the 5 f4 0-0 6 ttJf3 c5 7 Jie2
line.
But 8 ttJrJ! transposes into a
favorable line of the related
ttJf31h2-h3 variation. For instance,
8 ... cxd5 9 cxd5 ttJa6 8 Jie2 ttJc5
9 ttJd2 and 9 ... ttJe8 10 ttJc4 a6?!
11 b4 ttJd7 12 Qd2 as in
Aleksandrov-Zhelnin, Krasnodar
1997.
There is one further finesse.
Black can meet 5 Jie2 with S... cS,
so that 6 ttJf3 cxd4 7 ttJxd4 ttJc6
thwarts White.
But if White tries to reach this
through the ttJf31h3 order, S ttJrJ
0-0 6 h3, he can be stopped after
6.•. eS 7 dS, when Black favorably
opens the kingside, not the
queenside, with 7... ttJa6! 8 Jie3
ttJhS and ... fS.
Another curiosity of 5 ttJf3 0-0 6
h3 is that books say 6... cS is best,
after which 7 dS e6 8 Jid3.
But 6 dxcS! favors White a bit,
6•.:iVaS 7 Jid2! 'iVxcs 8 ttJrJ 0-0
90-0 and now 9 ... b6?! 10 ::'cl Jib7
11 ttJd5! as in Uhlmann-Bindrich,
Dresden 2001.
KID: OTHER VARIATIONS
An obscure variation that
deserves more attention, if only
because of its transpo possibilities,
is 1 d4 ttJf6 2 c4 g6 3 ttJc3 Jig7
4 e4 d6 S h3 0-0 6 Jie3. Some KID
experts say 6... eS is best and give
7 dS c6. Then 8 Jid3 b5!? leads to
equal chances, they say.
And now 8... exdS 9 exd5 ::'e8+
10 Jie3 Jih6 or 10 ... ttJh5 are
considered unclear. But 9 cxdS!
reaches a Benoni that many
authorities say favors White. He is
being rewarded for misdirection: if
White tries to get to the 9 cxd5
position through a Benoni order
there are various ways for Black to
thwart him, such as I d4 ttJf6 2 c4
e6 3 ttJf3 c5 4 d5 exd5 5 cxd5 d6
174
Indians
6 ttJc3 g6 7 h3 and now 7... a6 8 a4
'fie7.
MODERN DEFENSE
The Modem Defense after 1 d4
g6 2 c4 i.g7 becomes a King's
Indian after ... ttJf6. Black can use
this route to dodge the Samisch
among other KID lines. For
example, 1 d4 g6 2 c4 i.g7 3 ttJc3
d6 4 e4 and now 4.•. e5.
More common is 5 ttJf3, which
can become a Classical KID
following 5 ... ttJd7, as Tigran
Petrosian, among others, played.
But if Black doesn't want to playa
Classical he can exploit the 1 d4 g6
2 c4 i.g7 3 ttJc3 order with 3 ••• c5!?
That can become quite sharp
after 4 d5 i.xc3+!? 5 bxc3 f5 or
lead to a KIDlBenoni hybrid after
4 ... d6. Instead White usually plays
Black has to be willing to risk an 4 ttJf3. But that allows 4 •.•cxd4
endgame but there are benefits. If 5 ttJxd4 ttJc6!, a fine version for
White tries 5 i.e3, hoping for Black of the English because White
something like 5 ... ttJd7 6 d5 or 6 f3, doesn't get the time he needs to set
Black can misplace the bishop with up the Maroczy Bind.
5... exd4!? 6 i.xd4 ttJf6.
If White is interested m the
He will attack it with ... ttJc6, e.g.
7 ttJf3 ttJc6 8 i.e3 0-0 9 i.e2 l:te8 Classical, a safer route is 1 d4 g6
10 ttJd2 ttJd7! 11 0-0 ttJd4 12 i.d3 2 c4 i.g7 3 ttJf3 or 3 e4. Note that
ttJc5, Lputian-Todorcevic, Cannes in the latter case, Black has an extra
option in 3 ••. ttJc6.
1996.
If, instead, 5 d5 in the diagram,
Black gets a normal KID after
5... ttJf6. But another benefit of this
order is 5..•ttJd7 6 i.e3 ttJe7! and
.••f5, which theory smiles on.
The attack on the d-pawn turns
out well if Black gets to secure his
knight on d4, 4 d5 ttJd4 5 i.e3 with
5 ... c5 and then 6 dxc6 dxc6. This
recapture isn't possible after ... d6.
175
Indians
But 3... tbc6 can tum out very
badly after 4 tbf3! instead of 4 dS.
Then a typical continuation such as
4••• d6 5 d5! tbb8 6 h3 heads for a
Benoni (6 ... cS 7 i.d3) in which
White enjoys two extra tempi,
thanks to ... tbc6-b8. Life isn't easy
for the Modem player.
But today we know ltJf3 fits in
quite well with the center exchange
because after, say, 7... cS 8 :blltJc6
White can favorably allow 9 dS!
i.xc3+ 10 i.d2. And 4 ... dS no
longer deserves an exclamation
point.
A Black who can play both the
KID and Gruenfeld may prefer the
GRUENFELD DEFENSE
more flexible 4 ... 0-0. If White plays
When Bobby Fischer called S i.f4, which is designed to hold up
2 tbf3, after 1 d4 tbf6, 'a common ... eS after S... d6 6 e3, Black can
error,' he was thinking of how it then go ahead with S... d5, as
limits White's options in the Fischer did in his 'Game of the
Gruenfeld as well as in the King's Century,' six years before he tried
Indian. It makes the Exchange the normal Gruenfeld order.
Gruenfeld harmless because of a
In that order, 1 d4 ltJf6 2 c4 g6
pinning ... i.g4, he felt. That meant
after 1 d4 tbf6 2 tbf3 g6 3 c4 i.g7 3 tbc3 d5, the most fertile area for
traps lies in the Russian ('itb3)
4 tbc3:
Variation. One of the tabias comes
about after 4 tbf3 i.g7 S 'iVb3 dxc4
6 'Yi'xc4 0-0 7 e4 or alternatively,
4 'iVb3 dxc4 5 'Yi'xc4 i.g7 6 e4 0-0
7 tbf3.
In an earlier era 4 'Yi'b3 was
considered more accurate because
4 ... c6 S cxd5! favors White,
whereas in the second order, 4 ltJf3
i.g7 5 'iVb3 c6 is a reasonable
defense.
That seems trivial today
Black could play 4.•. d5! since
5 cxd5 tbxd5 6 e4 tbxc3 7 bxc3 but the case for 4 'iVb3 was
allows him to corrupt the White strengthened in the 1963 world
match
when
center with a timely ... i.g4. One of championship
Ernst Gruenfeld's first games with Petrosian demonstrated that after
the opening he created went 7 ... 0-0 4... dxc4 5 "iixc4 i.g7 6 e4 0-0
8 i.e2 cS 9 0-0 cxd4 10 cxd4 tbc6 White did not have to play 7 ltJf3
11 i.e3 i.g4! with advantage.
but could try 7 i.e2!?
176
Indians
Annotators pass over this
In practice Black usually replies
because 7... a6 8 ltJf3 lands in the 6....td7, because it's forcing and
Hungarian Variation and 7... ltJa6 because it leads to something
8 ltJf3 transposes to the Prins familiar, 7 'iib3 ltJb6 8 e4 0-0
Variation. Both have good 9 .te3 ..tg4. But this means he was
reputations. But the point of 7 .te2 denied the Hungarian and Prins
is White avoids Vasily Smyslov's lines.
tried-and-true 7 ltJf3 .tg4. If that
White also sets a trap, since the
matters to you, it may be worth
natural 6... .td7 7 'iib3 ltJb6 8 e4
risking 4 'iib3 dxc4 5 'iixc4 .te6!?,
.tg4? is punished by 9 ltJg5!,
Black's extra option.
threatening 10 'iVxf7+. That wins
time
for 9 ... 0-0 10 h3! .tc8 (or
But suppose you'd be happy in
the Smyslov line and not in the 1O ....td7) 11 ltJf3. As usual in the
alternatives. Then you should Gruenfeld, when Black has no
consider 4 ltJf3 i.g7 5 cxd5 ltJxd5 pressure on the center he is worse.
6 'iia4+!?, an idea of Andor
Lilienthal's.
This is both a ' something-tothink-about' order and a good way
to limit Black's options. White
wants his opponent to think about
6... ltJc6 and worry about the
consequences of7 e3 and 8 i.b5, as
well as of 6... c6 7 e4!, e.g. 7... ltJb6
8 'iic2 .txd4 9 i.h6 and :dl or
8... 0-0 9 i.e3 .tg4 10 ltJe5!.
A potential drawback to this
order is Black's extra option of
8... 0-09 i.e3 a5!? and ... a4. If you
don't mind that - and you want to
dodge the Prins or Hungarian 6 'iia4+!? suits you.
A bit dicier is Wolfgang
Uhlmann's similar 4 ltJf3 .tg7
5 'iVa4+. We again approach a
Russian tabia, 5....td7 6 "iVb3 dxc4
7 'iVxc4! 0-0 8 e4.
177
Indians
tZJxc5 as in Goletiani-Neubauer,
Port Erin 2005. Only tricksters
should rely on 5 'ifa4+.
GRUENFELD:EXCHANGE
In the old Exchange Variation, a
few issues stand out:
Black's Sixth Move. After 1 d4
ttJf6 2 c4 g6 3 ttJc3 d5 4 cxd5
Black has an extra tempo, ... .itd7. ttJxd5 5 e4 ttJxc3 6 bxc3 Black
But that limits his choices usually plays ... i.g7 and ... c5 in
somewhat since 8... a6 9 e5! is a some order. They typically
problem. His knight cannot retreat transpose.
to d7, as in the usual Hungarian.
But beginning in the 1990s a
Ditto 8... ttJc6 9 e5!.
difference was found in 7 .itb5+. If
Notice that in either line, Black
can clear d7 with 9....ite6. But
10 exf6! sacrifices the queen for
three pieces. Black has been
trapped in a line usually seen via
4 ttJf3 .itg7 5 ""3 dxc4 6 'ifxc4
0-0 7 e4 ttJc6 8 e5 .ite6? (8 ... ttJd7!)
9 exf6 .itxc4 10 fxg7 and 11 .itxc4.
In the last diagram Black usually
opts for 8 ... .itg4, reaching a
Smyslov position once more. But
that's probably a mistake because
he can try to punish White's order
with 8...ttJa6!? Then on 9 e5 ttJg4
10 i.f4 he gets to put his extra
... .itd7 to good use with 10...c5
11 d5 b5! as in Gerzhoy-Tiomin,
Tel Aviv 2002.
Another added option is 8... b5!,
e.g. 9 ttJxb5 ttJxe4 10 'ifxc7 ttJc6
with good play. Or 9 ~3 c5!
10 dxc5 ttJa6 11 e5 ttJg4 12 .itf4
Black has played 6... i.g7:
White can meet 7...i.d7 with
8 .ite2 and transpose to an ttJf3
Exchange Variation in which
Black's extra move, ... i.d7, blunts
his pressure on d4, e.g. 8... c5 9 ttJf3
ttJc6 10 0-0 0-0 11 l:.b 1 cxd4
12 cxd4 b6 13 d5! with advantage
(Almasi-Balusz, Gyula 1993).
Instead, Black should meet
7 .itb5+ with 7... c6! after which
8 .ita4 begins a controversial
178
Indians
line championed by Alexander
Belyavsky.
However if he had played 6... c5
then on 7 i.b5+ he has nothing
better than 7•.. i.d7, allowing White
to transpose to the Exchange line
with 8 i.e2!.
Then Lev Polugayevsky found a
finesse, 10 :tel, instead of 10 0-0.
His first point was that 10..:i!ic7
transposes into the Shamkovich
plan after 11 O-O!.
His second point was that the
rook will defend the c4-bishop after
Instead, 7...tiJc6? 8 d5 is an old
10 ... cxd4 11 cxd4 so Black can't
transpose into the ... ttJa5 line with
trap that has caught a number of
11...i.g4 12 f3 ttJa5?? and he is
GMs (8 .. .'ilfa5 9 'ifa4! 'ifxc3+
worse after 12 ... i..d7 13 0-0
I 0 ~e2 with a big edge in
(13 .. ttJa5 14 i.d3 and 'ifd2/i.h6).
Belyavsky-Mikhalchishin, Terme
Either way White benefits from
Zrece 2003, among others). This
10 l:kl.
makes 6•..i.g7! more precise.
The drawback is 10... cxd4
White Delays 0-0. A common
11 cxd4 'iVa5+. But White can make
sequence is 4 cxd5 ttJxd5 5 e4
a virtue of having not castled by
ttJxc3 6 bxc3 i.g7 7 i.c4 0-0
means of 12 ~f1 i..d7 13 h4! to
8 ttJe2 c5 9 i.e3 ttJc6.
open the h-file. This, too, has scored
some impressive victories and the
merit of 10 .l:i.c I is still in dispute.
Another wrinkle is to accelerate
castling with 9 0-0, rather than
9 i.e3.
In the 1950s, 10 0-0 cxd4
11 cxd4 i.g4 (12 f3 ttJa5) was
regarded as slightly favorable to
White. Black began to try a plan of
10 .. :ifc7 and 11...l:td8 favored by
Leonid Shamkovich and Isaac
Kashdan. But after some impressive
White is ready to transpose into
Spassky victories with II l:tc1!, this main lines after 9...ttJc6 10 i.e3. In
fell out of favor.
this way he avoids Peter Svidler's
179
Indians
idea of meeting 9 j"e3 with 9 ... cxd4
10 cxd4 'iia5+ 11 j"d2 'it'd8!.
But with the c-pawns on the
board, 12 j"d5 j"d7 13 ':bl 'it'c8
Experience with 12 j"c3 liJc6 is
inconclusive. If White doesn't have
anything better than 12 j"e3 'iia5+
he is headed towards a draw by
repetition before the game is five
minutes old. So far no downside to
9 0-0 has been found.
threatens the bishop with ... c4! and
... e6.
Another problem with ... cxd4
lies in the line made famous in the
1987 world championship match,
13 j"xf7+ ':xf7 14 fxg4 ':xfl+ 15
~xfl 'iib6 16 ~gl! 'it'e6.
Black delays ... cxd4. The old
main line of the Exchange
Gruenfeld was 4 cxd5 liJxd5 5 e4
liJxc3 6 bxc3 j"g7 7 i.c4 c5 8 liJe2
0-0 9 j"e3 liJc6 10 0-0 and then
1O ... cxd4 11 cxd4 j"g4 12 f3 liJa5.
But in the late 1970s Jan Timman
demonstrated the value of delaying
an exchange of c-pawns, with
10 ... i.g4. Play often transposes
following 11 f3 liJa5 and a later
... cxd4.
Then 17 "ii'd3 'iixg4 18 ':fl is
regarded as a slight edge for White.
But with the c-pawns on the board
we reach a slightly different
position after ~g1/ .. :ife6.
What's the point of 10... i.g4
11 f3 liJa5 ? Two stand out.
After 10 ... cxd4 11 cxd4 j"g4
12 f3liJa5 theory says White can try
13 j"d5 and get a nice game
following 13 ... i.d7 14 l:tbl 'it'c8
15 'iVd3 and :t.fc 1.
Then 16 'it'd3 'it'c4! gives Black
more vigorous queenside play, e.g.
17 'it'd2 'it'e6! 18 liJg3 liJc4 or
17 'it'xc4+ liJxc4 18 j"f2 cxd4
19 cxd4 e5. As a result, the more
accurate 10 ... j"g4 is replacing
1O... cxd4.
NEO-GRUENFELD
In the Neo-Gruenfeld, g2-g3 and
... d5 are played. If ... d5 comes first,
Black can capture on c4. That may
make recapturing hard, and this
explains why 1 d4 liJf6 2 c4 g6
3 liJc3 d5 4 g3?! is rarely seen, e.g.
4 ... dxc4! 5 "ii'a4+ liJfd7! 6 "ii'xc4
liJb6 7 'it'd3 liJc6 8 liJf3 j"f5 9 e4
j"g4 with an excellent game for
Black.
More common is an order such
as 1 d4 liJf6 2 c4 g6 3 g3 j"g7
4 j"g2.
180
Indians
7 ttJge2 eS 8 .te3? ttJg4 or 8 d5
ttJd4!.
There's
also
independent
territory to explore in 6 dS ttJaS
7 'iid3 cS, e.g 8 e4 d6 9 ttJge2 a6
and ... b5 as in Ilincic-Djuric,
Vrnjacka Banja 1998.
Black often delays ... dS because
he doesn't like 4 ... dS S cxdS ttJxd5
6 e4 ttJb6 7 ttJe2!. Instead, he waits,
with 4... 0-0, for the easier-to-handle
S ttJf3 dS! 6 cxdS ttJxdS.
White's counter-finesse is S ttJc3,
to rule out S... d5. Black can
continue in KID mode (5 ... d6 or
S... c5). But there is also S...ttJc6!?
It says, 'You tried to avoid the
Neo-Gruenfeld. But if you continue
6 ttJf3 now, I can transpose with
6... dS.'
The plot deepens if White
responds 6 e4 because 6 ...d6! gets
Black into one of those Fianchetto
KID lines with an early e2-e4 that
were discredited in the 19S0s, e.g.
Note that 8 ttJf3 d6 transposes to
a Black-friendly KID that usually
arises from 1 d4 ttJf6 2 c4 g6 3 g3
.tg7 4 .tg2 0-0 S ttJf3 d6 6 ttJc3
ttJc6 7 d5 ttJa5 when White plays
8 'iid3?! cS 9 0-0 instead of8 ttJd2!.
QUEEN'S INDIAN DEFENSE
The Queen's Indian and NimzoIndian are opposite sides of the
same coin, one designed if not
minted by Aron Nimzovich. Before
settling on his formula - 1 d4 ttJf6
2 c4 e6 followed by 3 ttJf3 b6 or
3 ttJc3 .tb4 - he experimented with
an order that didn't depend on
White's third move, 1 d4 ttJf6 2 c4
b6.
The benefit of this order is Black
has sharper options after 3 ttJf3
181
Indians
If White can conquer his Nimzo.Jib7 4 g3 such as 4 ... e5!?, 4 ....Jixf3
and 4 ... c5 5 d5 b5!? And the fear, he can play the more ambitious
downside? Books say White can 1 d4 tDf6 2 c4 e6 3 tDf3 b6 4 tDc3,
win control of e4 with tDc3 and allowing 4 ... .Jib4. He takes that risk
'iVc2 or f2-f3. For example, 3 tDc3 in the hope that Black will prefer
.Jib74 'iVc2 d5 5 cxd5 tDxd5 6 tDf3 4 ... .Jib7 and 5 'iVc2 .tb4 or 5 .tg5
e6 7 e4.
h6 6 .th4 .tb4.
But good players, including
But then 5 a3! reaches the 4 a3
Fischer and Larsen have used .Jib7 5 tDc3 Petrosian system. The
2 ... b6!? and others as White,
difference is that Black has lost his
including Anatoly Karpov, usually
chance for 4 a3 .ta6, a major
transposed into normal lines with
alternative.
3 tDf3, so there may be more
plusses than minuses to it.
That's why 4 tDc3 is more often
In the conventional order, 1 d4 met by 4 •.• .Jib4. White usually
tDf6 2 c4 e6. White is indicating his chooses between 5 .tg5 and 5 e3
feelings about the Nimzo-Indian and may transpose into one of the
when he chooses 3 tDf3. Black can 4 .Jig5 and 4 e3 lines. For example,
try to exploit that in lines such as 5 e3 .Jib7 6 .Jid3 tDe4 7 0-0 is that
gambit again.
3 ..•b64 e3 .Jib7 5 .Jid3 .Jib4+!?
In the old main line of the
Queen's Indian, 1 d4 tDf6 2 c4 e6
3 tDf3 b6 4 g3 .Jib7, White begins a
waiting game when he delays tDc3.
Books say 5 tDc3 .tb4! equalizes.
Play typically continues 5 .tg2
.Jie7 6 0-0 0-0. Then quiet moves
such as 7 b3 or 7 l:e 1 allow Black
to transpose into an equal Catalan
with 7 ... d5!.
White often replies 6 tDbd2 even
though it offers no more than
equality. His Nimzo-fear has hurt
him since 6 tDc3! transposes into a
line that runs 1 d4 tiJf6 2 c4 e6
3 tDc3 i..b4 4 e3 b6 5 .Jid3 .Jib7
6 tDf3 and offers excellent chances,
even in the 6 ... tDe4 7 0-0 gambit.
The critical move is 7 tDc3,
preparing to seize the center with
8 'ifc2 and 9 e4. Black usually
occupies e4, 7... tDe4, to mechanically stop e2-e4. But if ... .tb4
equalized earlier, why not 7.••.Jib4,
a tempo down?
182
Indians
a Queen's Gambit Declined with
4... d5 and then 5 a3 i.e7!:
This is rarely if ever mentioned
in books but it's been played by
Leonid Stein, David Bronstein and
Boris Spas sky. White may be better
after 8 i.d2 but not by a great deal
(8 ... i.xc3 9 i.xc3 liJe4).
Or 8... d6 9 :tel liJbd7 10 d5
liJc5 as in W.Schmidt-Gofshtein,
Debrecen 1992 (11 liJh4 a5 12 'ifc2
:e8 13 :fe 1 i.xc3 14 i.xc3 exd5
15 i.xf6 'i'xf6 16 cxd5 g6 with
equality).
It's as if the game began 1 d4 d5
2 c4 e6 3 liJf3 liJf6 and then 4 e3
i.e7 5 liJbd2?! and White got an
extra a2-a3 as compensation. It isn't
enough and 6 e3 0-0 7 i.d3 b6
8 0-0 i.b7 9 b3 liJbd7 10 i.b2 c5 is
easy equality.
The main stem of the Bogo is
4 i.d2, not 4 liJbd2. Black usually
protects the bishop (4 .. :Yi'e7, 4 ... a5,
BOGO-INDIAN DEFENSE
4... c5), thinking that by keeping the
According to Mark Taimanov's queenside situation in flux White
version of an oft-told story, a will be reluctant to allow Nimzowomen's tournament game began like positions with liJc3. For
1 d4 liJf6 2 c4 e6 3 liJf3 i.b4+. example, 4 .. .'iVe7 5 liJc3?! b6 6 e3
White was surprised by the check i.b7 7 i.d3 and now 7... i.xc3!
and asked, 'What opening are you 8 i.xc3 liJe4 is typical equalizing
playing?'
play.
The reply: 'I'm playing the
The 'something to think about'
Nimzovich! What are you doing?'
alternative is 4..•i.e7!? Black is
The Bogo-Indian deserves more offering another routine QGD
respect than that. After 4 liJbd2 (5 liJc3 d5 6 i.g5). Of course, he
White hopes to win the two bishops could have transposed more
with 5 a3. Experience indicates that directly with 3... d5. But he may
Black's best chances lie in creating want White to consider 5 g3 as well.
183
Indians
His point is that White would see
that S... b6 6 ii.g2 ii.b7 is a
favorable version of the QID. But
after S g3 Black has a good fallback
- 5... d5! transposes into a good
version of the Closed Catalan, one
that usually arises from 1 d4 llJf6
2 c4 e6 3 g3 dS 4llJf3 .tb4+ S .td2
.te7 - and one a 3 llJf3 player may
know little about.
NIMZO-INDIAN DEFENSE
The Nirnzo-Indian is related to
an opening without a name, 1 d4 d5
2 c4 e6 3llJc3 .tb4!?
But Black can avoid transposing into Botvinnik-Capablanca
(7 ... llJf6?!) with 7 ... llJc6 8 .td3
llJge7! and ... i.fS. For example,
9 'iVc2? c4 10 .te2 .tfS and 9 llJe2
c4 10 i.bl .tfS 11 a4 0-0 12 i.a3
l:.e8 with sufficient counterplay as
in Khurtsidze-Tsereteli, Tbilisi
2006 .
His KN can also be put to better
use if White tries 4 "ika4+. This
would transpose into a Ragozin
QGD after 4 ... llJc6 S llJf3 llJf6
which is double edged after 6llJeS.
But Black renders llJeS harmless
with S... llJe7!?
This move order is still very
young and that's why many players
sitting in White's chair will flee
towards the familiar with 4 llJf3.
But that limits him after 4 ... dxc4 to
allowing the Noteboom QGD (S e3
bS) or trying to reach a Vienna after
S e4 cS.
Black seems to be saying, 'If I'd
played 1.. .llJf6 2 c4 e6 you might
have avoided the Nirnzo with 3 llJf3
or 3 g3 or even 2 .tgS. But now
I've gotten into a Nirnzo after all.'
But this is not a normal Nirnzo
because if Black doubles the
c-pawns, White can undouble with
cxdS, as we saw in BotvinnikCapablanca in the Introduction.
That means the litmus test should
be 4 a3 .txc3+ 5 bxc3. For
instance, 5... c5 6 cxd5 exd5 7 e3.
NIMZO: LENINGRAD
Alexander Alekhine said 'the
most logical answer' to 1 d4 llJf6
2 c4 e6 3 llJc3 .tb4 was 4 .tgS
because 4 ... dS is 'a rather
unattractive variation' of the QGD.
That is, 1 d4 dS 2 c4 e6 3 llJc3 llJf6
4 .tgS ii.b4?!, which favors White
after S 'ii'a4+ or S e3.
But the real test of 4 i.gS, the
Leningrad Variation, is ... cS!. A
main line is 4 .tg5 h6 5 .th4 c5
6 d5 .txc3+ 7 bxc3.
184
Indians
NIMZO: QUEEN MOVES
After Gideon Stahlberg beat him
with 4 'ifb3 three times in their
1934 match, Nimzovich declared
the queen move 'almost a refutation
of my variation.' It's not even a
major line today. But it's so closely
related to 4 'iic2 that Black has to
be careful of being trapped in a
On 7... d6 Black prepares a 4 'ifc2 variation that is either
traditional dark-square blockade inferior or unfamiliar.
strategy ( ... e5, ... 'ife7 and perhaps
For example, 4 ... liJc6 is regarded
... g5). But that allows White to as sufficient to equalize against
choose piece play over pawn 4 'ifb3 but unfavorable against
structure with 8 dxe6!? since 4 'ifc2. Black can find himself in
8... fxe6 9 e3 and liJf3/i.e2 favors the wrong line after 4 'ifb3 liJc6
him. Better is 8... i.xe6 but 9 e3 sliJf3 dS.
liJbd7 10 i.d3 liJe5 is unclear.
Black can try to finesse with
7... eS and then 8... d6!, transposing
into the blockade line. But White
has a good extra option in 8 d6!.
Experience shows that even if
Black wins the advanced pawn
White stands well.
A superior order for Black is
4... h6 S i.h4 cS 6 dS d6!, delaying
... i.xc3+. This rules out d5-d6 and
reaches the desirable blockade
following 7 e3 i.xc3+! 8 bxc3 eS,
with equal chances.
For example, 9 'iVc2 'ife7 10 liJf3
g5 11 i.g3 liJh5 12 i.d3 liJd7
13 liJd2 liJdf6 14 0-0-0 ~d8!? and
... <3;c7, as in Guseinov-Ibragimov,
Baku 2006.
Now 6 a3 i.xc3+? 7 'ifxc3 favors
White. He's transposed into 4 'ifc2
liJc6?! 5 liJf3 d5 6 a3 i.xc3+. In
that order, 6... dxc4?? would lose a
piece. But with the queen on b3,
6... dxc4! 7 'iixc4 i.d6 is excellent
for Black.
A similar case arises after 4 'ifb3
c5 and 5 dxc5. Then 5... liJa6 6 a3
i.xc5 is an easy road to equality.
185
Indians
In contrast, 6....ltxc3+?! 7 'ifxc3 13 'ifxf5 g6 14 'tic2 tiJh5 with a
tiJxc5 8 f3! is poor. This transposes quick draw in Shipov-Rozentalis,
into 4 'iYc2 c5 5 dxc5 when Black Tromso 2006.
plays the inferior 5... tiJa6?! (6 a3
That means an aggressive Black
.txc3+ 7 'iYxc3 tiJxc5 8 f3!) rather
can use this move order as a form of
than 5... 0-0! 6 a3 .ltxc5.
reconnaissance: He begins with
The main difference between the 1 d4 tiJf6 2 c4 e6 3 tiJc3 ~b4 in
two queen moves is that 4 'ifc2 hopes of reaching one of the
doesn't attack the bishop but does sharp lines. If White adopts the
prepare e2-e4. If Black stops that conservative 4 'ifc2 instead, Black
push with 4... d5 what kind of QGD can retreat into a solid QGD
do we have ? The answer is a good position with 4... d5 5 a3 i.e7 at
one for Black after 5 .ltg5 dxc4 and little cost.
even better after 5 tiJf3?! dxc4!.
The newest finesses in 4 'tie2
Moreover, the cautious 5 e3 occur after 4... d5 5 exd5 e5!? This
doesn't fit in well with 4 'ifc2. After is Veselin Topalov's gambit, based
5... c5, Black reaches comfortable on lines such as 6 dxe6 cxd4
4 e3 Nimzo territory following 7 exf7+ cJ;xf7 or 6....txe6 7 dxc5
6l2Jf3 tiJc6 7 a3 .ltxc3+ 8 bxc3 0-0. 0-0. Whether Black gets enough
The real test of ... d5 in a Nimzo compensation is controversial but
is usually a2-a3 so we have to our concern is whether White can
avoid the messy issue entirely.
evaluate 5 a3 .te7!?
One try is 6 .ltg5, since 6... exd5
reaches a quiet line that more
commonly arises after 4 'ifc2 d5
5 cxd5 exd5 6 .ltg5 c5.
White is better but only
somewhat, e.g. 6 cxd5 exd5 7 .ltg5
c6 8 e3 .ltg4!? 9 tiJge2 tiJbd7 10 h3
.lth5 11 tiJg3 .ltg6 12 tiJf5 .ltxf5!
The resulting position may be
good for White after 7 a3 ~xc3+
8 bxc3, e.g. 8... h6 9 i.xf6 'tixf6
10 e3 0-0 11 tiJf3 tiJc6 12 .te2 ~e6
13 0-0 l:.fc8 14 'i'a4 with a modest
edge in Rowson-Pert, British
Championship 2006.
There is, however, a complication in 6 i.g5 'i'xd5!?
186
Indians
White's goal is again to trap Black
in an unfamiliar 4 'iVc2 line.
'What's this?' Black thinks to
himself. 'Does he really believe
4 ... c5 5 d5 is good? Or is his idea to
meet 4 ... d5 with 5 cxd5 exd5
6 'ifb5+?'
What happened is Black
inveigled his opponent into a line
that runs 4 'ifc2 d5 5 ~g5 c5 6 cxd5
'ifxd5, when he threatens 7... 'ifxg5,
7... cxd4 and 7... ttJe4. This usually
leads to equal chances after 7 ~xf6
gxf68 dxc5 'ifxc5.
Another way for White to deal
'I can't figure this all out so I'll
with the new gambit is to answer
4 'ifc2 d5 5 cxd5 c5 with 6 dxc5. He just castle,' he may conclude. That's
gets the upper hand after 6... 'fixd5 not a bad decision. There's nothing
7 a3! ~xc3+ 8 'ifxc3 0-0 9 f3!. To wrong with 4 ••• 0-0 5 a3 and
avoid this Black can take his 5...~xc3+ 6 'fixc3 - provided that
chances with 6... ttJxd5.
Black knows the theory.
But watch out when a trickster
tries 6...exd5. Then on 7 ~g5 we
transpose into a position that comes
from 4 'ifc2 d5 5 cxd5 exd5 6 ~g5
when Black tries the ambitious
6... c5 and White fails to find 7 a3!.
Instead he's played 7 dxc5,
giving Black a fine game after
7••• h6! 8 ~h4 g5 9 ~g3 ttJe4 and
then 10 .txb8 'iff6! 11 .tg3 ttJxc3
as in the sensational game
I.Sokolov-Aronian, Turin 2006.
Finally, few books even mention
the third queen move, 4 'fid3.
What theory? The considerable
amount of book that's been built up
around 4 'i'c2 0-0 5 a3 i.xc3+
6 'fixc3, into which he's
unwittingly transposed. This is a
problem if Black normally meets
4 'ifc2 with something other than
4 ... 0-0.
Similarly, ifhe is unfamiliar with
4 'ifc2 d5 5 a3 ~xc3+, because he
plays some other fourth move as
Black, he will be a stranger in a
strange land after 4 'iVd3 d5 5 a3!
~xc3+? 6 'fixc3.
187
Indians
NIMZO: SAMISCH
line, 4 a3 i.xc3+ 5 bxc3 0-0 6 e3 c5
After Fritz Samisch introduced
4 a3 attempts were made to refine it
by delaying a2-a3 in favor of
Nikolai Riumin's 4 D and Akiba
Rubinstein's 4 e3, which are
supposed to be more flexible. But a
case can be made that 4 a3 i.xc3+
S bxc3 is the most flexible, because
it allows White to choose between
f2-D and e2-e3 once Black
commits himself.
7 i.d3 tbc6 8 tbe2, with a quick e3e4 and f2-f4-f5 attack.
But there's a better transposition
than Samisch-into-Riumin or
Samisch-into-Rubinstein. It is
Rubinstein-into-Samisch, 4 e3
followed by a2-a3. It arises in
several orders including 4 e3 cS
S i.d3 0-0 and 6 a3 i.xc3+ 7 bxc3.
For example, after S...b6:
This is more promising than 4 a3
because Black's castled king is
vulnerable, e.g. 7... b6 8 e4! i.b7
Black would be happy to see 6 e3 9 i.gS, as in the celebrated Keressince 6... i.b7 7 D c5! renders good Reshevsky, 1953 Candidates game
counterplay. But White is better (9 ... h6 10 h4! d6 11 e5 dxe5
after 6 f3! and 7 e4 (or 6 ... d5 12 dxe5 i.e4 13 1:th3! with a
ferocious attack).
7 i.g5!).
Black often gets confused in the
On the other hand, if Black
chooses S... cS he would have good delayed Samisch, as in that game.
play after 6 D d5 7 cxd5 tbxd5!. He thought he was transposing into
This transposes into an unclear 4 D an earlier game from the same
variation.
tournament. That one went 4 e3 cS
S a3, rather than 5 i.d3, and then
White may do better with 6 e3!?
S...i.xc3+ 6 bxc3 b6 7 i.d3 i.b7.
For instance, 6 ... tbc6 7 i.d3 d6
8 tbe2 0-0 or 8 ... e5 9 e4 0-0
Black obtained excellent play
transposes to another promising following 8 D tbc6 9 tbe2 0-0
188
Indians
10 0-0 lba5 11 e4 lbe8! with the
idea of ...lbd6. Reshevsky would
have been able to transpose into that
game after 8lbe2 .tb7 or 8 f3 .tb7.
But White crossed him up with
8 e4!.
NIMZO: RUBINSTEIN
The best known Nimzo-Indian
tabia arises three moves after 4 e3.
commitments to pawn structure and
can't be forced to trade off his
bishop (as 4 ... c5 5 lbe2 and 6 a3
does). But 4 ... 0-0 does commit his
king which is important if White
shifts to the delayed Samisch, 5 a3
.txc3+ 6 bxc3.
Castling also costs Black a tempo
and prevents him from reaching
Robert Huebner's dark-square
blockade strategy. After 5 .td3 c5
6lbf3lbc6 he doesn't get a chance
to see 7 0-0 .txc3! because White
interrupts him with 7 d5!.
A shrewd Black may be tempted
to avoid all of this by means of
4...lbe4!? He is trying to trick
White into one of the best versions
of 4 ... b6.
'This
position
has
been
investigated by some masters,
lovers of chess analysis, to the 3035 th (!) moves,' Vladimir Simagin
wrote way back in 1954. Black's
defenses have stood up well in this
and similar positions and his main
concern is move order.
Some Nimzo experts reach the
diagram via 4 ... 0-0, 5... c5, 6 ... d5
and 7... lbc6. Others prefer 5... d5
and 6 ... c5. Still others began with
4 ... c5 and then 5 ... lbc6 and 6 ... d5.
And others have faith in 4 ... d5,
5 ... 0-0 and 6 ... c5.
Today 4 ... 0-0 has the most
supporters because Black makes no
Black succeeds after 5 "ifc2 f5
6lbe2 b6 or 6lbf3 b6 7 .td3 .tb7,
both excellent 4 ... b6 lines for him.
The main benefit in the 4 ... lbe4
order is that he doesn't have to deal
with the enormous amount of 4 e3
b6 theory, as well as some specific
dangers, such as the 5 iL.d3 iL.b7
6lbf3 lbe4 7 O-O!? gambit.
189
Indians
White can also meet the threat to
c3 in the last diagram with 5 tbe2
instead of 5 'iYc2. If Black doesn't
like the looks of 5... 'iYh4 6 g3 'iif6
he can try to reach a good 4... b6line
once more with 5... f5 6 'iYc2 b6.
TANGO
In a remarkably short time, 1 d4
tbf6 2 c4 tbc6 has spawned a lot
of theory, especially in sharp
continuations beginning with 3 tbc3
e5 and 3 d5 tbe5. But in most
games White flees towards the
familiar with 3 tbf3. After 3... e6:
White can try 4 g3 after which
4... d5 makes it a Catalan. Such
positions, with a blocked c-pawn,
were once considered heretical
but now have not-too-shabby
reputations. After 5 i..g2 dxc4 the
position is considered unclear
following 6 'iia4 tbd7! or 6 0-0
l:tb8!?
Moreover the Black QN is well
placed ifhe manages to transpose to
a Bogo-Indian with 4 ... i.b4+
5 i.d2 'iie7. But White should
avoid that with 5 tbbd2! and 6 a3.
This is a transposition from a
normal Bogo, 1 d4 tbf6 2 c4 e6
3 tbf3 i.b4+ when 4 tbbd2 has
been met by the weak 4...tbc6?!
5 g3 'iie7.
If, instead, White plays 4 tbc3 in
the diagram, Black can transpose
into a rare Nimzo-Indian with
4... i.b4 5 'iic2. White can be
reasonably happy with his
prospects after, say, 5... 'iie7 6 a3
i.xc3+ 7 'iixc3 d6 8 b4 e5 8 dxe5
tbxe5 9 i.b2, Baburin-Scannell,
Feugen 2006.
But the main question is who has
tricked whom? Did Black drag his
opponent into a 4 'iic2 Nirnzo that
he didn't want to play?
Or did White get into a 4 'iic2
line in which Black has been denied
the more popular 4... d5 or 4... c5 ?
Once again, both players may have
been tricked.
190
Chapter Seven:
Dutch Defense and Flank Openings
Everyone knows the first finesse
of the Dutch Defense: Black can
avoid the Staunton Gambit, 1 d4 f5
2 e4, by means of 1...e6 and 2 c4 f5.
In the last few decades 1...e6 has
also been used to reach a NimzoIndian or Queen's Indian (2 c4 liJf6)
or a Benoni (2 c4 c5) without risk of
facing a Trompowsky.
White hints he is amenable to a
Tarrasch French (2 ... d5 3 e4). But
his real intent may be a Colle (2 ... d5
3 liJgf3!) in which Black has
blocked in his QB. Black can
also find himself in an inferior
Stonewall Attack after 2... c5 3 e3
liJc6 4 c3 liJf6 5 i..d3 d5 and 6 f4!.
Of course, there's another
hazard. After 2 e4 Black has to be
ready to defend a French (2 ... d5) or
the more dangerous version of the
English Defense (2 ... b6) or try
2 ... c5, as Bent Larsen did on
occasion. And he has to know what
to do after 2 liJd2!?
White can only dream of getting
that position after 1 d4 d5 2 e3 and
it's quite good, e.g. 6.•.i..d6 7 'iff3
i..d7 8 liJh3 'iib6 9 liJf2 0-0-0
10 0-0 ~b8 11 e4!, MarshallRubinstein, Vienna 1908.
Today the threat of the Staunton
Gambit doesn't scare many Dutch
players. But some prefer 1.. .e6
anyway, to avoid other l...f5 lines,
such as 2 liJc3 or 2 i..g5. They are
an attempt to improve on the
normal Staunton, 2 e4 fxe4 3 liJc3
liJf6 4 i..g5.
After 1 d4 f5 2 liJc3:
191
Dutch Defense and Flank Openings
DUTCH: MAIN LINES
Black often delays his biggest
decision, ... d6 or ... d5, and looks for
a good time for ... .tb4(+). For
example, 1 d4 f5 2 c4 lLlf6 3 g3 e6
4 .tg2 .tb4+:
Black's choice is limited since
2... lLl f6 3 ~g5 threatens a
~xf6/e2-e3/
Trompowsky-like
~d3riVf3 setup that is not to the
liking of many Dutch players. For
example, 3... c6 4 ~xf6! exf6 5 e3
d5 6 ~d3 g6 7 'it'f3 ~e6 8 h3 ~g7
9 lLlge2 lLld7 10 g4! fxg4 11 hxg4
'it'e7? 12 l:txh7! as in KorchnoiKostjoerin, Havana 1966.
The most common response to
2 lLlc3 is 2•.. d5, when 3 e4!?
transposes to a good version of the
Staunton after 3... fxe4? 4 f3 exf3
5 lLlxf3 but 4 'iWh5+! is better.
That means 3... dxe4 is best. But
this often transposes (4 f3 lLlf6 5
fxe4 fxe4) into 3... fxe4 4 f3 lLlf6
5 fxe4 dxe4. Then 6 ~c4 e5!?
7 ~g5! and 6... lLlc6 7lLlge2 e5 8 d5
are unclear.
Bottom line: If Black can also
play the French, avoiding all this
with I d4 e6 may be his best route
to the Dutch. Moving on to 1...f5:
Now 5 lLlc3?! is fine for Black,
e.g. 5... 0-0 6 lLlf3 lLle4 7 'iic2 d6
8 0-0 .txc3! 9 bxc3 lLld7 10 l:tdl
~e8 11 lLld2 lLldf6!? as in HortShort, Prague 1990.
Better is 5 ~d2! since an
exchange of bishops tends to favor
White. This is particularly true if
Black adopts a Stonewall pawn
structure, 5... ~xd2+ 6 lLlxd2 d5?,
instead of 6... d6! and ... e5.
Black has more success delaying
... .tb4+ in the hopes that White will
voluntarily develop his knight on c3
and allow the pin. But that's only
good when White is committed to
lLlf3 or ~g2. The pin is premature
after 1 d4 f5 2 c4 e6 3 lLlc3 .tb4?!,
for example, because White hasn't
ruled out Akiba Rubinstein's plan of
e2-e3/lLlge2/a2-a3, which he used
192
Dutch Defense and Flank Openings
just as effectively against the Dutch
as against the Nimzo-Indian.
control of e5 with ltJd3. But after
... d6 a knight maneuver to f4 runs
into ... e5.
In practice an order such as 1 d4
fS 2 g3 ltJf6 3 iLg2 e6 4 ltJf3 iLe7
S 0-0 makes S... dS attractive
because White has already showed
his hand with ltJf3. In contrast, after
4 ltJh3 iLe7 S 0-0 his best is S... d6!.
That means 4 'iic2 ltJf6 S e3! 0-0
6 iLd3 followed by ltJe2 and a2-a3,
favors White, e.g. 6... d6 7 ltJe2 ltJc6
8 0-0 iLxc3 9 'iixc3 a5 10 b3 e5
11 iLa3! J:te8 12 d5 ltJe7 13 f4!,
Yakovich-Gleizerov, Stockholm
2000.
Consequently, if Black commits
his d-pawn too early, White knows
what to do with his KN - 1 d4 fS
2 g3 e6 3 iLg2 ltJf6 4 c4 dS S ltJh3!.
The same applies if White develops
his knight at e2 rather than h3. Two
of the 1951 world championship
games reached this position.
The more accurate order, after
1 d4 f5 2 c4 e6 3 ltJc3, is 3••. ltJf6.
This waits for White to make a
decision, such as 4 g3 or 4 ltJf3,
which would justify 4 ... iLh4!. If
White chooses 4 e3?! instead he is
starting the Rubinstein plan too
early and Black should switch
to the ... d6/ ... e5 pawn structure, e.g.
Mikhail
Botvinnik
played
4 ...iLe7 5 iLd3 0-06 ltJge2 d6! with
6
...
d5?!,
got
the
worst
of7
ltJge2
c6
a good game.
8 b3 ltJe4 9 0-0 ltJd7 10 iLb2 ltJdf6
Once White decides to fianchetto
11 f3! and eventually lost.
his KB, the focus turns to his KN. If
it goes to e2 or h3, the conventional
When David Bronstein was
wisdom holds that Black is better Black he chose 6... d6!. White's
off with ... d6 than with ... d5. A knight had little to do following
knight stands nicely on f4 in a 7 ltJge2 c6 8 0-0 eS. In fact it never
Stonewall because it watches the moved again and he was soon
backward e-pawn and can reinforce worse.
193
Dutch Defense and Flank Openings
The situation becomes more
difficult when Black seeks a
Stonewall with ... ~d6. Then he
can't use ... ~e7 as a waiting move.
But its place as a 'pass' can be taken
by ... c6.
Mark Taimanov, who praised
6 ... c6,
explained
that
the
Stonewall's defects are best
exploited when White can play
ttJf3-e5-d3 and ttJbl-d2-f3. He
cited a game that went 6... d5,
After 1 d4 f5 2 g3 ttJf6 3 ~g2 e6
4 c4 d5 White stands well with
5 ttJh3! c6 6 0-0 ~d6 7 ~f4. But
instead of 6 ... c6, and turned in
White's favor following 7 ttJbd2!
c6 8 ~c2 ~e8 9 ttJe5 ttJbd7
10 ttJd3! ttJe4 11 ttJt3 g5 12 b4.
Black has 4 ... c6!?, which he
believes will be a useful move after
both 5 ttJh3 d6 and 5 ttJO d5.
However,
once White is
committed to ttJc3, Black has an
easier game, Taimanov said.
For example, 5 ttJf3 d5! 6 0-0
~d6 and 7 b3 'fie7 stops the
unfavorable trade of Black's good
bishop (8 i..a3).
If White insists on the trade, with
8 a4, Black gets adequate play from
8... a5! 9 ~a3 i..xa3 10 ttJxa3 0-0
11 ttJc2 b6 12 ttJce 1 i..b 7 13 ttJd3
ttJbd7 and ... c5 as in TenikashvilliRadjabov, Moscow 1998.
A further bit of subtlety involves
White's QN, as in 1 d4 f5 2 g3 ttJf6
3 ~g2 e6 4 ttJf3 ~e7 5 0-0 0-0 6 c4
and then 6... c6!. Black waits for
7 ttJc3, which makes 7 ... d5 more
appropriate.
For example, 7 ttJc3 d5 8 b3 ttJe4
9 ~b2 ttJd7 10 e3 ~d6 11 ttJd2
ttJdf6. If White avoids 7 ttJc3 by
means of 7 b3 Black has a good
reply in 7 ... a5. He can answer
8 ~a3 with 8... ~xa3 9 ttJxa3 d6!
or 9 ... ~e7 10 ttJc2 d6!. His bad
bishop is no longer so bad.
TRANSPO-DUTCH
Black can set up a Dutch on his
side of the board after 1 c4, 1 ttJO
or 1 g3 by playing 1...f5. He can
also make a good Dutch out of a
194
Dutch Defense and Flank Openings
doubtful English Defense, as in 1 c4
b6 2 d4 ~b7 3 liJc3 e6 4 a3 and
now 4..• f5!'
pioneered by Carlos Guimard, 1 d4
d5 2 c4 e6 3 liJc3 ~e7 4 liJf3 and
now 4••• f5.
White's a-pawn plays a minor
role, e.g. 5 d5liJf6 6 g3liJa6 7 ~g2
ltJc5 8 liJh3 ~d6 9 ~e3 ~e5!,
Karpov-Morozevich, Cannes 2002.
This sequence rules out the
Staunton Gambit as well as the 1 d4
fS 2 liJc3 and 2 ~gS lines, and gets
White to commit his knights, so that
the liJbd2, liJh3 and b2-b3/~a3
options are lost.
But after I c4 fS or 1 liJf3 fS
White can play d2-d3 rather than
d2-d4. This takes e4 away from
Black's KN and may prepare a
powerful push of the e-pawn, e.g.
1 liJf3 fS 2 g3 liJf6 3 ~g2 e6 4 c4
dS S 0-0 c6 and now 6 d3 ~d6
7 liJc3 0-0 8 e4!.
But White can make a favorable
course correction with 5 ~f4! liJf6
6 e3 c6 7 ~d3 0-0 8 'iic2. For
example, 8... liJe4 9 g4! liJa6 10 a3
'iVaS 11 ~e2 liJxc3+ 12 bxc3 liJc7
13 cS liJe8 14 l:thg 1 ~h8 IS l:tg2
and White crashed through in LimaGiardelli, Sao Paolo 1993.
Black was victimized by his
insistence on 3... e6 and ... dS. Ifhe's
ENGLISH DEFENSE
willing to handle a ... d6 Dutch, he
should play 3... d6, preparing 4... eS.
Most lines of the English
If White deters him with 4 d4 Black
can transpose to normal ... d6 lines Defense favor White at least
with 4... e6.
slightly and Black should be alert to
transpositions to the Dutch, QID or
Before leaving the Dutch we other openings. After 1 d4 e6 2 c4
should mention a little-known order b6 3liJc3 ~b7 4 e3:
195
Dutch Defense and Flank Openings
... i.b4 pin if he inserts a2-a3 first,
such as after 1 d4 e6 2 c4 b6 3 a3.
Black can reach the Petrosian QID
with 3... ltJf6 4 ltJc3 i.b7 5 ltJf3.
Another route to the same place is
3... i.b7 4ltJc3 ltJf6 5 ltJf3.
The Petrosian QID is hardly a
bad line. Why then would Black use
one of these orders (and risk the
Now 4... i.b4?! 5 ltJe2! ltJf6 is a complications of 5 d5)? The usual
White-friendly Nimzo-Indian, e.g. reasons are to avoid the
6 a3 i.xc3+ 7ltJxc3 d5 8 b3.
Trompowsky and the Fianchetto
But 4... f5! is a good Dutch. The
lines of the QID.
right time for the pin would be
5 ltJf3 i.b4 since White can no
longer break it favorably with ltJe2
and a2-a3. Better for White is
5ltJge2 ltJf6 6 d5 but after 6...ltJa6
this is one of the better lines of the
English Defense for Black.
But a more adventurous Black
will try 3...f5!? 4 ltJc3 ltJf6. White
is better off here than he was in the
position considered earlier, I d4 e6
2 c4 b6 3ltJc3 i.b7 4 e3 f5!. In this
case he can benefit from a
fianchetto, 5 g3 i.b7 6 d5!.
An irony is that many English
Defenders rarely - if ever - begin
with the characteristic ... b6. The
reason is that I d4 b6 2 e4! i.b7 and
its twin, I e4 b6 2 d4 i.b7, are
considered better for White than
when c2-c4 is played.
However, there is room for
subterfuge after 5 g3 c6! 6 i.g2 d5:
The more precise order is 1 d4
e6, waiting for 2 c4 b6, they say.
Against 1 e4 they avoid 1... b6 and
go into something else entirely.
Only in reply to I c4 or I ltJf3 do
they begin 1... b6.
Another aspect of the English
Defense is that White can avoid the
Each side has an extra move
compared with a normal Stonewall.
But ... b6 is more useful than
196
Dutch Defense and Flank Openings
White's extra a2-a3. That's shown
by 7 cxdS cxdS 8 .tf4 .td6 9 .txd6
'iixd6 10 ':'c 1 ttJc6 11 ttJf3 0-0
12 0-0 .tb7 13 e3 ':'ac8 with
balanced play in RogozenkoMorozevich, Bundesliga 2000.
Note how late Black waited to
put his bishop on b7. He was
holding back because after S dS it
makes more sense to play S... .ta6!.
Then 6 e3 exdS 7 cxdS .txfl and
7 ttJxdS .td6! are quite nice for
him.
If Black insists on the English
Defense against 1 e4, a key
question is how White defends the
e-pawn after 1 e4 b6 2 d4 .tb7. On
3 ttJc3 e6 Black is ready for
4....tb4. Experience with 4 ttJf3
.tb4 S .td3 ttJf6 has varied widely
but the results are better for Black
than other 1 e4 b6 lines.
This retains the options of ttJc3,
c2-c3 and even c2-c4, as in S... dS
6 eS ttJfd7 7 0-0 cS 8 c3 and 7... .te7
8 c4!, e.g. 8... dxc4 9 i.xc4 a6
10 ttJc3 bS 11 i.d3 ttJb6 12 i.e4
with an edge in Khalifinan-Bauer,
internet 2004.
CATALAN OPENING
A Catalan is i.g2 attached to a
Queen's Gambit. Curiously, the
Catalan comes about more often
these days via 1 d4 ttJf6 than 1 d4
dS, at least at the master level.
After 1 d4 dS 2 c4 e6 3 g3 Black
More exact is 3 .td3! and 3... e6 equalizes faster, with 3 ... dxc4!
4 ttJf3. Then 4 ... cS S ttJc3 cxd4 4 ttJf3 cS, e.g. S 'i'a4+ i.d7 6 'i'xc4
6 ttJxd4 is an excellent Sicilian .tc6 7 .tg2 ttJd7 8 0-0 ttJgf6 9 ttJc3
Defense for him.
bS 10 'i'd3 'iYb6, OpocenskyKotov, Prague-Moscow 1946.
Black's best is 3... ttJf6!. The
More commonly, a Catalan arises
attacked e-pawn can't move (4 eS
after 1 d4 ttJf6 2 c4 e6. But if White
.txg2) and on 4 ttJc3 e6 Black has
is concerned about the Benko,
... .tb4 on tap.
Benoni, Budapest, et aI, he may try
2 g3 instead.
White's most precise order may
This transposes to a normal
be 3 i.d3 ttJf6 4 'iie2! and then
Catalan after 2 ... e6 3 c4. This order
4..• e6 5 ttJf3.
197
Dutch Defense and Flank Openings
allows Black to develop his QB
outside his pawn center (2 ... d5
3 tbf3 c6 4 .lig2 .lif5). But that is no
longer considered an easy road to
equality after, say, 5 0-0 e6 6 b3
h6 7 .lib2 tbbd7 8 c4 .lid6 9 tbbd2
0-0 10 tbe5 and £2-f4, S.PolgarUshenina, Dresden 2006.
There are two common I d4 tbf6
2 c4 e6 Catalan tabias and they
often appear after 3 g3 d5 4 .lig2
and 4 tbf3 (or 3 tbf3 d5 4 g3). There
are very minor differences between
the tWo. But the third order, 1 d4
tbf6 2 c4 e6 3 tbc3 d5 4 g3, has
plusses and minuses.
superior form of the Tarrasch QGD,
4... c5 5 cxd5 cxd4! 6 'iVxd4 tbc6
that worked well for Boris Spassky.
Note that White can limit enemy
resources by holding back the c- or
d-pawn. For example, 1 d4 tbf6
2 tbf3 e6 3 g3 d5 4 .lig2 .lie7 5 0-0
and a later c2-c4 rules out ... .lib4+
and early ... dxc4 lines.
He can also delay d2-d4, as with
1 c4 e6 2 g3 d5 3 tbf3 and then
3... tbf6 4 .lig2 .lie7 5 0-0 0-0 6 d4
is the closed Catalan. Of course,
there is less chance of reaching a
Catalan middlegame via these
orders than after 1 d4, 2 c4 and
3 g3.
Black's main decision in the
Catalan is whether (and when) to
capture on c4. If he wants to open
the position the careful way is 1 d4
tbf6 2 c4 e6 3 g3 d5 4 .lig2 and now
4•••.lie7 5 tbfJ 0-0 6 0-0 dxc4.
For example, on 4 ... dxc4 5 'iWa4+
.lid7 6 'iWxc4 .lic6 7 tbf3 Black can
unbalance matters with 7... .lixf3
and obtain good play against the
isolated d-pawn. Better is 7 f3!,
which has been considered
favorable since Tinsley-Schlechter,
Hastings 1895.
But this order is imprecise
because Black can transpose into a
In this way he avoids 'iVa4+ lines
such 1 d4 tbf6 2 c4 e6 3 g3 d5
4 .lig2 dxc4 5 'iWa4+. This makes
198
Dutch Defense and Flank Openings
sense if he has no interest in
4 ... dxc4 S tbf3 lines such as S... a6,
S... tbc6, S... i.d7 and S... i.b4+.
The queen check can set a trap in
delayed form, 4... i.e7 5 tbf3 dxc4
6 'i'a4+.
But this is a splendid time for
2•.. b6!, transposing to an English
Defense in which White is
committed to c2-c4 yet has denied
himself the ambitious e2-e4/f2-f3
setups. A player who usually meets
1 d4 e6 2 c4 b6 with a setup such as
3 tbc3 .ltb7 4 a3 or 4 e3, can find
Then 6... tbbd7?! doesn't seem himself in a suspect position after
like an error. But what has 1 c4 e6 2 tbf3 b6 and then 3 e4 i.b7
happened is that Black has been 4 tbc3 i.b4 5 .ltd3 tbe7 6 a3?!
tricked into S 'i'a4+ tbbd7 6 tbf3 i.xc3 7 dxc3 0-0 8 0-0 f5!, as in
i.e7? (instead of6 ... cS! or 6 .. ..l:t.b8!) Ibragimov-Shabalov, U.S. Open
which is a slight disadvantage for 2003.
him.
Quite a different story is 1 c4 g6
ENGLISH OPENING
which retains a wide range of
An amateur who plays 1 c4 options including the King's Indian,
is often at a psychological Modern Defense and Dutch Indian.
disadvantage when the game Black can wait for 2 tbc3 so that
transposes into a I d4 opening 2... c5 gets him into a good line after
because he typically chose 1 c4 to 3 tbf3 .ltg7 4 d4 cxd4 5 tbxd4
avoid main 1 d4 lines such as in the i.g7. If he plays 1 c4 cS instead,
Slav, Gruenfeld and Nirnzo-Indian. White has the more promising
It pays for him to be able to handle 2 tbf3!? and 3 d4.
a QGD because 1... e6 2 tbc3 d5
Another family of finesses
3 b3? d4!, for example, is bad.
White can try to stay in Reti-range begins with 1 c4 tbf6. White
with 2 tbf3 and then 2 ... dS typically responds 2 tbc3 so that he
can meet 2 ... e6 with the well3 b3.
199
Dutch Defense and Flank Openings
regarded 3 e4. This also enables
him to dodge the Gruenfeld with
2 ... g6 3 e4 and transpose into the
King's Indian after 3 ... d6 4 d4.
Books devote a lot of attention to
2 12Jc3 d5. But perhaps the best
aspect of the 1 c4 12Jf6 2 12Jc3 order
for Black is he can transpose into a
1...c5 line with 2 ... c5! and then
break in the center with 3 ... d5. As
noted earlier, if he had played the
more common order such as 1 c4 c5
White can break first with 2 12Jf3!?
and 3 d4.
ENGLISH: 1... e5
After 1 c4 e5 Black typically
chooses between open systems
( ... l2Jf6/ ... d5) and closed ones
( ... l2Jc6 and possibly ....tb4).
The finesses begin with 2 g3.
This will transpose into main lines
after, say, 2 ... l2Jf6 3 .tg2 d5 4 cxd5
12Jxd5 5 12Jc3 or 2 ... l2Jc6 3 .tg2 g6
4 12Jc3. But in this way White
avoids all the nasty ... .tb4 business.
After Anatoly Karpov made good
use of 2 12Jc3 12Jf6 3 12Jf3 12Jc6 4 g3
Another virtue of 1 c4l2Jf6 is that .tb4 as Black in his 1987 world
match,
Vasser
2 12Jc3 e5 gets Black to a common championship
position without having to concern Seirawan played 2 g3 against him.
himself with 1 c4 e5 2 g3!? If Karpov was out of his repertoire so
White insists on 2 g3 (after 1 c4 he played 2... g6?!, which he had
12Jf6) he has to deal with 2... c6! used only once before, when he was
3 .tg2 d5.
11. That game went 3 .tg2 .tg7, a
reversed closed Sicilian in which
the absence of an early ... l2Jc6 takes
the sting out ofb2-b4-b5.
He either has to risk a gambit
(4 12Jf3 dxc4/...b5) or defend the cpawn with 4 'ilVc2 or 4 b3, which
allows Black to solve his opening
problems with 4 ... e5!, e.g. 4 'ii'c2 e5
5 d3 h6! 6 12Jf3 .td6 7 0-0 0-0
8 12Jc3 d4, Socko-Belov, Hastings
2004.
But Seirawan's 3 d4! gave him
either a distinct edge in space
(3 ... exd4 4 'ii'xd4) or a superior
endgame (3 ... d6 4 dxe5 dxe5
200
Dutch Defense and Flank Openings
5 'i'xd8+). Karpov chose the latter
and lost.
against the OlD. For example, 8 e4
:e8 9 h3 i.f8 10 l:tel a6 11 d5?!
The drawback to 2 g3 is 2...tbf6
3 i.g2 c6 since 4 tbfJ e4 5 tbd4 d5
6 cxd5 can be met by Mikhail Tal's
6.••'i'xd5!. By attacking the knight
cxd5 12 cxd5 b5! 13 a3 tbb6 14 b3
i.d7 15 i.e3 'i'b8 with a fine game
for Black (A.Sokolov-Sherwin,
Swiss Team Championship 2001).
Black obtains excellent chances,
e.g. 7 tbc2 'iih5 and ... i.h3.
In the main closed variation of
the Reversed Sicilian, 1 c4 e5
The comparable 2 tbc3 tbf6 3 g3
c6 line never returned to fashion
2 tbc3 tbc6 3 g3 g6 4 i.g2 i.g7,
White usually chooses among 5 d3,
after it was met in the Tal-Botvinnik 5 e3 and 5 :b 1. But the finesses
rematch by 4 tbfJ! e4 5 tbd4 d5 start earlier if Black plays 1 c4 g6
6 cxd5. Transposing to the other 2 tbc3 i.g7, as Kasparov did. This
line with 6 ... 'i'xd5?? is clearly enabled him to shift to his favorite
King's Indian after 3 d4 tbf6.
impossible and 6 ... cxd5 7 d3!
favors White.
But what about 3 g3 ? That seems
to narrow Black's options since the
However, this order IS quite
Gruenfeld is discouraged and the
playable after 4... d6!.
Neo-Gruenfeld, 3... c6 and 4 ... d5,
can be stopped by 4 i.g2/5 e4.
Instead, Kasparov responded
3...tbc6!?
White's only real chance for
advantage lies in d2-d4. After
5 i.g2 Black can choose between a
book KID position with 5... g6 6 d4
and a well-known OlD position
after 5•••.te7 6 0-0 0-0 7 d4 tbbd7.
The latter is a good choice because
the fianchetto is one of the least
ambitious weapons White has
White has missed his chance for
d2-d4. Moreover, 4 i.g2 d6 5 tbfJ
e5! puts him in a closed variation
with tbf3 and d2-d3, which theory
calls fairly even.
201
Dutch Defense and Flank Openings
ENGLISH: ..•ii.b4
The reversed version of 1 e4 c5
2 tLlf3 tLlc6 3 ii.b5 has more nuance
in the English because Black can
play ... ii.b4 at move two, three, four
or five. An instructive example is
1 c4 e5 2 tLlc3 tLlc6. White can try
to punish Black's last move with
3 e3 and d2-d4-d5. Then 3 ...tLlf6 is
an excellent waiting move.
Now 4 tLlf3 ii.b4! is a wellguarded line for Black. He has
transposed into 1 c4 e5 2 tLlc3 tLlf6
3 tLlf3 tLlc6 and then 4 e3 ii.b4.
Theory indicates that after 3 tLlf3
tLlc6 White's only serious chance
for an edge lies instead in 4 g3!.
That was a case of White tipping
his center plans with his third move.
He does better with 1 c4 e5 2 tLlc3
tLlf6 3 g3 or 3 tLlf3. But Black can
anticipate that with 2 ... .tb4. He
wants to know how White will
retake after 3 ... .txc3.
That's important information
because the conservative dxc3 is
recommended in some 2 tLlc3 tLlf6
3 tLlf3 tLlc6 4 g3 ii.b4 lines. But it's
harmless after 1 c4 e5 2 tLlc3 ii.b4
3 g3 ii.xc3 4 dxc3?! because Black
can make better use of his KN than
transposing with ... tLlf6, e.g. 4••• d6
5 ii.g2 tLle7! 6 tLlf3 tLlbc6 with
good chances.
There are other English lines in
which bxc3 is preferable. But in this
case, on 4 bxc3 Black takes
advantage of the absence of ... tLlf6
by playing ... f5 or ... f6.
For example, 4•.• d6 5 ii.g2 tLlc6
6 d3 f5 or 6 d4 tLlge7 7 tLlf3 0-0
8 0-0 f6! (9 ii.a3 ~h8 and now
10 d5? tLla5 11 'ti'a4 b6 12 c5 tLlxd5
13 cxd6 cxd6 14 tLlxe5 ii.b7
favored Black in Akesson-Sisatto ,
Stockholm 2006).
The drawback to 1 c4 e5 2 tLlc3
ii.b4 is, of course, 3 tLld5 which
both avoids ... ii.xc3 and gains time.
But if you believe recent experience
indicating that 3 ....td6 or 3 ....tc5
equalize, then 2 ... ii.b4! is an
improved order for a Black who
likes the positions that arise after
1 c4 e5 2 tLlc3 tLlf6 3 g3 ii.b4 or
3 tLlf3 tLlc6 4 g3 ii.b4.
202
Dutch Defense and Flank Openings
It may also pay to accelerate
... i.xc3 in that 3 tiJf3 tiJc6 4 g3
i.b4 line. Play usually continues
5 i.g2 0-0 6 0-0. Since 6... .l:te8
7 tiJd5! favors White Black's task is
to time ... i.xc3 and ... e4 accurately.
Books consider 6... e4 at length
and hardly mention 6..•i.xc3!?
After 7 bxc3 .l:te8 Black obtains a
good version of 6....l:te8, in which
he has dodged 7 tiJd5. And on 8 d3,
which threatens to gain a positional
edge with 9 e4, Black has 8 •.. e4!.
If White retakes with the d-pawn
Black may do best to avoid ... e4,
which could transpose to a difficult
6... e4 7 tiJel i.xc3 8 dxc3 line.
Instead he should maintain his darksquare pawn structure, e.g. 7 dxc3
d6 and then 8 tiJel i.e6 9 b3 'iVd7
and ..•i.h3.
REVERSED DRAGON
The mirror image of the
Accelerated Dragon arises after
1 c4 e5 2 tiJc3 tiJf6 3 tiJf3 tiJc6 4 g3
d5 5 cxd5 tiJxd5 as well as after
3 g3 d5 4 cxd5 tiJxd5. Once more
White's choice between the two
depends on what he most wants to
avoid.
The first order allows Black
alternatives such as 4 ... i.c5 and
4 ... tiJd4 that are quite reasonable. In
contrast, after 1 c4 e5 2 tiJc3 tiJf6
3 g3, there is no ... tiJd4 at all, and
3•.. i.c5?! is premature because
White has not committed himself to
tiJf3.
This means he can get the upper
hand with 4 i.g2 tiJc6 5 e3!
followed by tiJge2 and d2-d4. For
example, 5... a6 6 tiJge2 d6 7 d4
exd4 8 exd4 i.a7 9 h3 0-0 10 g4,
as in Moskalenko-Diaz, Badalona
2000, among others.
The main objection to the second
order is 3 ... i.b4, which some
English players regard as more
promising for Black than 4 ... i.h4 is
in the four-knights order.
When the Dragon occurs in the
four-knights order, White decides
whether to delay d2-d3, as in
203
Dutch Defense and Flank Openings
1 c4 e5 2 liJc3 liJf6 3 liJf3 liJc6 4 g3
d5 5 cxd5liJxd5:
ENGLISH: l ... c5
Many 1 c4 players consider l...c5
the most unwelcome reply - not
because of its fighting qualities but
rather because of the drawishness of
symmetry. If Black simply imitates
his opponent, 2 liJc3 liJc6 3 g3 g6 4
.tg2 i..g7, White has a hard time
developing a significant advantage.
In fact 5 e3 has long been
regarded by GMs as 'a tacit draw
offer,' as Andras Adorjan put it,
There used to be a debate over because of the liquidating 5... e6
whether 6 d3 is necessary to avoid 6 liJge2 liJge7 7 d4 cxd4 8 liJxd4
6 .tg2 liJxc3 7 bxc3 e4. The issue d5! and then 9 cxd5 liJxd5
seemed to hinge on the soundness 10 liJxd5 exd5. Another potential
of 8 liJd4. That question has not problem with 5 e3 is the
been resolved but it now seems unbalancing and highly doub1e8 liJgl! f5 9 f3 favors White, edged 5... i..xc3 6 bxc3 f5.
as does 8....tf5 9 'iVa4 'iVd5 10 liJh3
White can avoid both lines with
(Yrjo1a-Westerinen, Finnish Cham- 2 g3!?, delaying liJc3. If Black
pionship 1995).
copies him, 2... g6 3 .tg2 .tg7, he
That suggests 6 .tg2 is right provided White wants to quickly
push his b-pawn, such as 6 ... liJb6
7 0-0 .te7 8 a3 or 8 .l:lbl. If he
doesn't, 6 d3 is easier to handle.
can try 4 e3 and if 4... e6 then 5 d4
d5 6 dxc5! or 5... cxd4 6 exd4liJe7
7 d5! and 6... d5 7 c5.
More
5liJe2!:
Note also that 1 c4 e5 2liJc3liJc6
would transpose into the fourknights lines following 3 liJf3 liJf6.
But this doesn't happen often
because theory says Black is doing
well after 3 liJf3 f5 and White is
doing better after 3 g3. If Black
wants the four-knights, 2 ... liJf6 is
the way to go.
204
common
IS
4... liJc6
Dutch Defense and Flank Openings
apparent in Seirawan-Dlugy, World
Open 1989, which went 6 'ilVb3 tbc6
7 .txc6+ bxc6 8 tbf3 f6 9 'i'a4.
Black was lost soon after 9 ... 'ilVd7
10 d3 e5 11 i.e3 tbb5? 12l::i.cl tbd4
13 0-0 h5? 14 h4 .
This enables White to transpose
safely into normal lines (5 ... e5
6 tbbc3 or 5...tbf6 6 tbbc3) as ifhe
had played 2 tbc3 and Black had
passed up both the symmetrical
... e6/ ... d5 and the unbalancing
... .txc3.
If Black wants the reversed
And if Black tries to restore Maroczy setup he has a sneaky way
symmetry with S•.• e6, White can to reach it beginning with 1 c4 tbf6
break out with 6 d4 and then and then 2 tbc3 dS 3 cxdS tbxdS:
6... cxd4 7 exd4 dS 8 cSt?
(Grischuk-Malakhov,
Moscow
2006).
That modest 2 g3 can pose
problems for a Black who likes to
play the 'improved Tarrasch
Defense,' 1 c4 c5 2 tbf3 tbf6 3 tbc3
e6 4 g3 tbc6 5 .tg2 d5 6 cxd5
tbxd5!. In the 1 c4 cS 2 g3 order, he
discovers that 2...tbf6 3 .tg2 dS
4 cxdS tbxdS S tbc3 puts
immediate pressure on d5.
This variation is almost always
treated as some offbeat Gruenfeld,
with analysis of 4 g3 met by
4 ... tbxc3. However, Black also has
4... cS! and then 5 i.g2 tbc7.
If Black responds with the
natural S... e6 he reaches his goal
after 6 tbf3 b6 or 6...tbc6, as has
This setup is more often created
happened in dozens of master by 1 c4 c5 2 tbc3 tbf6 3 tbf3 d5
games. But 5... e6?? 6 tbf3?? is a 4 cxd5 tbxd5 and 5... c5. But in that
double blunder since 6 tbxdS exdS order White can avoid the Maroczy
7 'ilVb3! wins a pawn, as in the with 2 tbf3!? and 3 d4.
Vladimir Kramnik debacle cited in
In the last diagram White has
the Introduction.
another chance to avoid the bind,
If an 'improved Tarrasch' player with 4 tbf3. But he has to know
realizes the problem in time, he can what he's doing after 4...cS S d4 or
recover with S...tbc7! and seek a S e4.
reversed Maroczy Bind. But that's
Closely related to those positions
not an easy system to play for the are the ones arising from 1 c4 cS
first time as Black. That was 2 tbf3 tbf6 3 g3. This innocent205
Dutch Defense and Flank Openings
looking order was prepared by
When White focuses on d5 with
Kramnik for his world champion- 2 tbc3, rather than 2 g3 or 2 tbf3,
ship match with Kasparov. On the character of the game changes
3...tbc6 he could respond 4 tbc3 slightly. Black can fall into the
and reach a Symmetrical Four Kramnik trap through a back door,
Knights. That would be a moral 2 ... tbf6 3 g3 d5 4 cxd5 tbxd5
victory for Kramnik because he 5 i.g2 and now 5... e6?? 6 tbxd5
liked the Four Knights, while cxd5 7 'iib3.
Kasparov, like many other players,
One of the drawbacks to 2 tbc3 is
avoided it (by means of 1 c4 c5
that
2 ... g6!? allows Black to
2 tbc3 tbf6 3 tbf3 d5 or 2 ...tbc6
concentrate firepower on d4. His
3 tbf3 tbd4).
point is that 3 tbf3 i.g7 4 d4 cxd4
A key position arises after 3 ... d5 5 tbxd4 tbc6 stops the Maroczy
4 cxd5 tbxd5 5 i.g2.
Bind:
The natural 5... g6? is faulty
because 6 d4! cxd4 7 "iVxd4 attacks
Black's KR.
Black is also worse after 6 ... i.g7
7 e4 tbc7 8 d5. What has happened
is we've reached a Neo-Gruenfeld
(1 d4 tbf6 2 c4 g6 3 g3 d5 4 i.g2
i.g7 5 tbf3) in which Black has
played 5... c5?! (and 6 cxd5 tbxd5
7 e4!) instead of the superior 5... 0-0
or 5... dxc4.
The d4-knight is under attack
before White can smooth out his
development with 6 e4 and 7 i.e3.
If the knight retreats to b3 or c2
then 6 ... i.xc3+! puts White's
queenside under severe pressure.
Trading that bishop was once
thought to be foolhardy but that
view was overthrown by examples
like 6 tbc2 i.xc3+ 7 bxc3 tbf6 8 f3
But in the diagram Black can "iVa5 9 i.d2 "iVa4 10 e4 d6 and
avoid all of this with the simple ...tbd7-c5, with excellent chances
5•..tbc6!.
in Agzamov-Taimanov, Riga 1975.
206
Dutch Defense and Flank Openings
There is a possible drawback to
1 c4 c5 2 liJc3 g6 in the rarely
played 3 d4. Then 3•.. cxd4 4 'iixd4
attacks the KR. This leads to
4 ... liJf6 5 i.g5 liJc6 6 'iid2, which
is hard to evaluate, when compared
with vaguely similar positions, such
as 1 e4 c5 2 liJf3 liJc6 3 d4 cxd4
4 liJxd4 g6 5 c4 liJf6 6 liJc3 liJxd4
7 'iixd4 d6 8 i.g5 i.g7 9 'iVd2.
prefer 6 ... liJa5, e.g. 7 e4 liJf6 8 h3
0-0 9 .td3 a6 10 0-0 b5 11 cxb5 c4
12 i.c2 axb5 13 liJxb5 liJb3!
(lzoria-Savchenko, internet 2006).
Despite that, and support for
6 ... i.xc3+ 7 bxc3 liJa5 as well,
theory doesn't trust Black's system.
But if there is no knight on c6,
the pawn push is weak, as BisguierKarpov, Skopje 1972 showed. After
There are two other ways of 1 c4 c5 2 liJc3 g6 3 liJf3 i.g7 4 e3
trying to reach the last diagram. liJf6 5 d4 Black continued 5•.• 0-0
Bobby Fischer met 1 c4 on occasion 6 .te2 cxd4! 7 exd4 d5! with an
with 1...g6 2 liJc3 i.g7 3 liJf3 c5. excellent game, as we saw with
But Black has to be willing to play 1 d4liJf6 2 c4 c5 3 e3.
a KID or Modem (3 d4).
RETI OPENING
Fischer also met 1 c4 c5 2 liJc3
with 2.•.liJc6 and then 3 liJf3 g6.
This rules out a 'iixd4 recapture
after d2-d4/ ... cxd4. Black gets
where he wants to go following
4 d4 cxd4 5 liJxd4 i.g7.
But White can play 4 e3! and
threaten 5 d4 and 6 d5, which is
stronger when it attacks a knight on
c6.
Kramnik made 1 liJf3 famous as
a purely transpositional device.
After 1...d5 he reaches normal
Queen's Gambits with 2 d4 and
3 c4. But after 1...liJf6 he almost
always replies 2 c4.
In this way he avoids the NimzoIndian by meeting 2 ... e6 3 liJc3
?-.
\lUI
Ab4 with 4 wc2!?, rather than 4 d4.
He also dodges normal Gruenfeld
positions, meeting 2 ... g6 with
3 liJc3 d5 4 cxd5 liJxd5 5 'iVa4+ or
3 ... i.g74 e4!. That saves gallons of
midnight oil.
Note that 1 c4liJf6 2 liJf3 would
get White to the same position as
1 liJf3 liJf6 2 c4. But 1 c4 runs the
risk of 1...e5!?
Fischer was willing to play
4 ... i.g7 5 d4 d6 6 d5 liJe5. Others
Of all the major first moves,
1 liJf3 grants Black his best
207
Dutch Defense and Flank Openings
- 1...f5 tries to exploit White's
opportunity for 1... b6 since White
cannot build a center with £2-f3 and early commitment of his KN.
cannot easily transpose to an But as we saw in 1 c4 f5 White can
e2-e4/d2-d4 position. After 1 1LJf3 try for e2-e4, even starting with
2 d3.
b62 e4 ~b7 31LJc31LJf6!:
The attack on the e-pawn
interrupts White's center-building,
4 e5 lLJd5 5 lLJxd5 .txd5 6 d4 e6,
e.g. 7 c4 ~b4+ or 7 .te2 c5 8 c4
~xf3! 9 ~xf3 lLJc6 10 d5 lLJxe5
11 ~e4 f5 with a fine game
in Shchekachev-Bauer, ClermontFerrand 2003.
Black's other options are more
elastic:
- l...c5 is a favorite of
Sicilianistas (2 e4). Since 1 1LJf3
players generally prefer closed
lines, the typical result is a
Symmetrical English (2 c4) or a
King's Indian Reversed (2 g3).
- 1. .. d6 seems to say, 'I'm
offering a KID this way (2 d4 lLJf6
and 3 ... g6) because I don't want to
end up in a Pirc after 1...g6 2 e4.' If
White tries to trick him into a Pirc,
1... d6 2 e4, he can find himself in a
Sicilian,2 ... c5.
White's goal is to transpose into
a favorable version of the Lisitsyn
(2 e4) Gambit after 2.••lLJf6 3 e4
fxe4 4 fxe4 lLJxe4 5 ~d3 lLJf6
6 lLJg5 and lLJxh7! or ~xh7!. The
original Lisitsyn runs 2 e4 fxe4
3 lLJg5 lLJf6 4 d3 exd3 5 ~xd3 but
Black gets the edge with 3... d5! in
that version.
If Black is willing to playa ... d6
Dutch, he should reply 2 ... d6,
which avoids the gambit as well as
2 ... lLJf6 3 ~g5. But if Black wants a
Stonewall, then 2 d3 is a major
problem for him.
- 1. .. g6 keeps more Black
options open. Against 2 e4 he can
shift to a hyper accelerated Dragon
Sicilian with 2 ... c5 or head toward a
safe Pirc, with 2... d6 or 2 ... ~g7,
because White has ruled out £2-f4
and £2-f3 lines.
208
Dutch Defense and Flank Openings
After 2 d4 or 2 c4 Black retains
the option of various kinds of
Modem Defenses, Englishes and
the Dutch Indian (2 ... f5). If he
meets 2 d4 with 2... c6 the results
include a Neo-Gruenfeld (3 g3 d5),
a Modem (3 e4 d5), or some
nondescript Indian (3 c4 d6
4 tbc3 i.g7 or 4 ... i.g4).
- 1... e6 is, next to 1. .. tbf6, the
most elastic response to 1 tbS. It's
often used by a Dutch loyalist (2 d4
f5) who is trying to trick White into
a French (2 e4 d5!) or a Sicilian
(2 e4 c5).
RETI: 1...d5
After 1...d5 the Hyperrnodemist
will play 2 c4 or 2 g3. But the
transpo-minded will look at 2 e3.
Taimanov used this to get into a
Slav (2 ... c6 3 c4 tbf6 4 tbc3 e6
5 d4) or QGA (2 ... tbf6 3 c4 dxc4
4 i.xc4 e6 5 d4, dodging ... i.g4
lines). White enjoys many of the
same opportunities as he would
after 2 c4 c6 3 d4 tbf6 but without
having to find an answer to 2 c4
d4!?
More common after 1 tbf3 d5 is
2 c4. The simplest defense is
2... dxc4. Then 3 e3 becomes a book
QGA in the vast majority of games
(3 ... tbf6 4 i.xc4).
Can White do better? Richard
Reti liked 3 tba3. But 3 ... c5 4 tbxc4
tbc6 and ... e5 transposes into a kind
of English (1 c4 c5 2 tbs tbf6 3 g3
d5 4 cxd5 tbxd5 5 i.g2 tbc6 6 0-0
e5 7 d3 i.e7) in which White
has chosen a suspect knight
development, 8 tbbd2 and tbc4.
After 3 tba3 fell out of fashion,
White tried 3 'i'a4+. The simplest
response to that is 3...tbd7 and
... a6/ ... b5, seeking a Catalan.
For example, 4 'i'xc4 e6 5 g3 a6 6
i.g2 b5. And on 4 g3 a6 5 tbc3 e6
6 i.g2 ltb8!.
The threat of 7 ... b5 leads to
7 'i'xc4 b5 8 'i'b3 i.b7. Then White
has nothing better than to transpose
into an equal Catalan with d2-d4.
That is, 1 d4 tbf6 2 c4 e6 3 g3 d5
4 i.g2 dxc4 5 'i'a4+ tbbd7 6 tbS
a6 7 tbc3 ':b8 8 'i'xc4 b5 etc.
209
Dutch Defense and Flank Openings
If Black doesn't like these
positions he should consider
meeting 2 c4 with 2•••c6 since the
Slav Defense (3 d4) is holding up
well these days. White can remain
true to the spirit of Reti with 3 b3.
Today's theory says Black can
equalize whether he adopts
Lasker's defense ( ... i.f5) or
Capablanca's ( ... i.g4). A major
reason is that 3 b3 rules out 'ifb3,
the attack on b7 that is so
troublesome to Black in ... d5/ ... i.f5
openings.
4lbf3 e4 5lbd4 i.c5! 6lbxc6 dxc6
7 e3 i.f5 8 'ifc2 'ife7 9 i.e2 0-0-0
and White was virtually lost after
10 f4?
Theory concluded that White
should play lbe2 and f2-f4 instead.
But another point of view says
Larsen's real error was playing c2c4 before lbf3. Instead of 3 e3 and
3 c4 lbf6 4 e3, which are played
most often today, White may prefer
3lbf3.
The modem treatment of the Reti
is to delay both b2-b3 and c2-c4.
White hints he is willing to play a
King's Indian Reversed instead,
with 1 lbf3 d5 2 g3 lbf6 3 i.g2.
Then on 3•.•c5 4 0-0 e6 he can act in
the center with 5 d3 and lbbd2/e2e4. The KIR enjoys a pretty good
reputation against ... e6.
If, however, Black brings out his
QB, with 3•••i.f5. White can switch
from the KIR back to the Reti with
4 c4. In this way he avoids 2 c4 d4
and 2 ... dxc4.
NIMZO-LARSEN OPENING
Bent Larsen helped make 1 b3 e5
2 i.b2 lbc6 famous by attacking
the e-pawn with lbf3. Then after
... e4 he occupied the central dark
squares with pieces and assaulted
... d5 with c2-c4.
But this strategy was almost
retired in 1970 by a Larsen-Spassky
miniature, which went 3 c4 lbf6
Black defends the e-pawn with
... f6 in some 1 b3 lines but here
3••. f6?! 4 e4! and 5 i.c4 can't be
good for him. Moreover, 3... d6,
which is fine following 3 e3, allows
White to punch at e5 with 4 d4, e.g.
4•••e4 5 d5! exf3 6 dxc6 bxc6
7 gxf3 lbf6 8 'ifd4! with advantage.
For example, 8... i.f5? 9 e4 i.g6
10 lbd2 i.e7 11 'ifa4! 'ifd7
12 i.h3! resigns, Hansen-Bergfalk,
correspondence 1992, or 8... i.e7
9 l:tgl c5 10 'ife3 0-0 11 lbc3 g6
12 0-0-0 l:te8 13 'iff4 and 14 e4.
Note also that in similar positions
... i.d6 defends the e-pawn well -
210
Dutch Defense and Flank Openings
1 b3 e5 2 ~b2 ~c6 3 e3 ~f6 4 ~b5
~d6!' But it fails here, 3 ~f3 ~d6
4 ~a3! and ~c4 or ~b5.
without reaching the Spassky
position (5 ... .i.xc5?? 6 ~xc6 and
Black most natural response to
3 ~f3 is 3 .••e4 and then 4 ~d4 ~f6.
The absence of c2-c4 plays a role in
S e3 ~cS 6 ~xc6 dxc6:
If White likes such positions he
should be happy (after 1 b3 e5
2 i.b2 ~c6) with 3 e3 d6 because
of 4 ~f3, rather than the bookendorsed 4 .i.b5.
~xg7).
This is better for White than
Larsen-Spassky after 7 d4!? ~e7
8 c4 and perhaps after 7... exd3
8 ~xd3 as well. If instead S...~xd4
6 ~xd4 dS then 7 c4 is timed well
and transposes into positions
Larsen had shown to be good. For
instance, 7 ... ~f5 8 ~c3 dxc4
9 ~xc4 ~d6 10 'ifc2 0-0 11 f4! c6
12 0-0 l:e8 13 ~dl 'ife7 14 ~f2,
Welling-Minasian, Capelle la
Grande 1996.
Nevertheless books say that after
1 b3 eS 2 ~b2 ~c6 the best move
is 3 e3. They often give 3... dS
followed by 4 ~b5 ~d6. Yet here
again ~f3 is an attractive idea 4 ~f3 e4 S ~d4 ~xd4 6 ~xd4
transposes into the above line, again
Now 4... e4 will bring about the
desired position, a tempo ahead,
after S ~d4 ~xd4 6 .i.xd4 when
Black inevitably plays ... dS.
Instead 4 ... ~f6 S d4! favors
White after 5... exd4 6 ~xd4 i.e7
7 ~b5, e.g. 7... i.d7 8 .i.xc6 bxc6
9 'iif3! d5 10 ~f5 as in HodgsonPercerias, Breda 1989. White has
transposed into a good 3 e3 d6
4 ~e2 .i.e7 5 d4 exd4?! 6 ~xd4
line.
And if White is happier facing
1 b3 d5 than 1...e5, he should
consider 1 ~f3 so that 1.. .d5 2 b3!
gets him closer to the middlegame
he wants.
211
212
Index of Opening Variations
(numbers refer to pages)
Albin Counter Gambit
123
Alekhine's Defense
94
Alekhine's Four Pawn Attack
94
Alekhine's 4 lDf3
95
Benko Gambit
156
Benoni Defense
159
Bishop's Gambit
16
Bishop's Opening
19
Black Knights Tango
190
Blackmar-Diemer Gambit
154
Bogo-Indian Defense
183
14
Breyer Gambit
Caro-Kann Defense
105
Caro-Kann, Panov Variation
108
Caro-Kann, Panov Accelerated
110
Caro-Kann, Two Knights Variation
106
Catalan Opening
197
Center Counter Defense
97
Center Counter, 2 ... lDf6
99
213
Index of Opening Variations
Center Game
18
Colle System
121
Danish Gambit
19
Dutch Defense
191
Dutch by transposition
194
English Defense
199
English Opening
199
English, Reversed Dragon
203
English, 1.. .c5
204
English, 1.. .e5
200
English, ... i.b4
202
Evans Gambit
31
Four Knights Game
28
French Defense
111
French Defense, Advance Variation
117
French Defense, Burn Variation
111
French Defense, Rubinstein Variation
111
French Defense, Steinitz Variation
112
French Defense, Tarrasch Variation
115
French Defense, Winawer Variation
113
Giuoco Piano
29
Goring Gambit
32
Gruenfeld Defense
176
Gruenfeld, Exchange Variation
178
Gruenfeld, Russian Variation
176
214
Index of Opening Variations
Hungarian Defense
17
King's Gambit
14
King's Gambit Declined
15
King's Indian, Classical Variation
169
King's Indian, Fianchetto Variation
170
King's Indian, Four Pawns Attack
173
King's Indian, Samisch Variation
167
King's Indian, h3
174
King's Indian, iog5
172
King's IndianlBenoni
161
Maroczy Bind
64
Max Lange Attack
29
Modem Benoni Defense
161
Modem Defense, 1 d4
175
Modem Defense, 1 e4
99
Modem, Gurgendize Variation
101
Neo-Gruenfeld Defense
180
Nimzo-Indian Defense
184
Nimzo-Indian, Leningrad Variation
184
Nimzo-Indian, Queen-move variations
185
Nimzo-Indian, Rubinstein Variation
189
Nimzo-Indian, Samisch Variation
188
Nimzo-Larsen Opening
210
Old Indian Defense
164
Petroff Defense, 3 d4
22
215
Index of Opening Variations
Petroff Defense, 3 lbxe5
22
Philidor Defense
20
Pirc Defense
102
Ponziani Opening
24
Pseudo-Lopez
20
Queen's Gambit Accepted
124
Queen's Gambit Accepted, 3 ... lbf6
126
Queen's Gambit Declined
129
QGD, Botvinnik Variation
142
QGD, Cambridge Springs Defense
140
QGD, Exchange Variation
134
QGD, Lasker Defense
140
QGD, Orthodox Variation
138
QGD, TarraschlSemi-Tarrasch
144
QGD, Vienna Variation
143
Queen's Indian Defense
181
Queen's Pawn Game
120
Reti Opening
207
Ruy Lopez, Anti-Marshall Variation
46
Ruy Lopez, Arkhangel Variation
44
Ruy Lopez, Berlin Defense
34
Ruy Lopez, Bird Defense
37
Ruy Lopez, Classical Defense
34
Ruy Lopez, Cozio Defense
35
Ruy Lopez, Exchange Variations
42
216
Index of Opening Variations
Ruy Lopez, Fianchetto Defense
35
Ruy Lopez, Marshall Gambit
46
Ruy Lopez, Main Line
49
Ruy Lopez, Modern Steinitz
39
Ruy Lopez, Moeller Defense
44
Ruy Lopez, Old Main Line
51
Ruy Lopez, Open Defense
42
Ruy Lopez, Steinitz Defense
37
Schmid Benoni
160
Scotch Gambit
32
Scotch Game
25
Semi-Slav Defense
151
Semi-Slav, Meran Variation
152
Sicilian Defense, Alapin Variation
60
Sicilian, Chameleon Variation
55
Sicilian, Classical Variation
76
Sicilian, Closed Variation
55
Sicilian, Dragon Variation, Accelerated
72
Sicilian, Dragon Variation, Classical
69
Sicilian, Dragon Variation, Fianchetto
65
Sicilian, Dragon Variation, Levenfish
69
Sicilian, Dragon Variation, Yugoslav Attack
65
Sicilian, English Attack
74
Sicilian, Four Knights Variation
92
Sicilian, Grand Prix Attack
58
217
Index of Opening Variations
Sicilian, Kan Variation
87
Sicilian, Keres Attack
74
Sicilian, Moscow Variation
63
Sicilian, Najdorf Variation
84
Sicilian, Rauzer Variation
78
Sicilian, Rossolimo Variation
63
Sicilian, Scheveningen Variation
73
Sicilian, Sozin Variation
81
Sicilian, Sveshnikov Variation
92
Sicilian, Taimanov Variation
87
Sicilian, .. :"b6
90
Sicilian, 4 'ii'xd4
56
Sicilian, 5... J.d7
84
Sicilian, 5 ... lLlbd7
83
Slav Defense
144
Slav, Exchange Variation
145
Slav, Three Pawns Variation
148
Staunton Gambit
192
Stonewall Attack
191
Torre Attack
155
Trompowsky Attack
153
Two Knights Defense
29
Vienna Game
16
218