Incomplete p information: Perfect Bayesian equilibrium Lectures in Game Theory Fall 2011, Part 7 24.07.2011 G.B. Asheim, ECON3/4200-7 A static Bayesian game U 1 A ( 12 ) Nature B ( 12 ) D U 1 D Equivalent representation: 2 L R 2 L R 2 L R 2 L 1 L 2 A ( 12 ) Nature B ( 12 ) R L 2 R R 1 U D U 1 D 1 U D 1 U D What if the uninformed gets to What if the informed gets to observe the informed choice? observe the uninformed choice? 24.07.2011 G.B. Asheim, ECON3/4200-7 2 1 A dynamic Bayesian game: Screening A ( 12 ) Nature 1 L B (2) 2 R A ( 12 ) Nature B ( 12 ) Equivalent representation: The informed gets to observe the uninformed choice. 1 U 1 U D 1 U D 1 U D 1 U L 2 A ( 12 ) Nature B ( 12 ) D 1 U R D 1 U L D U 1 2 R D D Use subgame perfect Nash equilibrium! 24.07.2011 3 G.B. Asheim, ECON3/4200-7 The uninformed gets to observe the informed choice. U 1 A ( 12 ) Nature B ( 12 ) D U 1 D 2 L R 2 L A dynamic Bayesian game: Signaling Equivalent representation: L 2 U1 R R 2 L R 2 L D 2 L R A ( 12 ) Nature B ( 12 ) L R U 1 R R Requires a new equilibrium concept: Perfect Bayesian equilibrium 24.07.2011 D L G.B. Asheim, ECON3/4200-7 Why? 4 2 Gift game, version 1 F1 G 0 0 N 0, F ( 12 ) F 2 A 2, 2 R - 2, 0 Nature 1 2 E ( ) 0, 0 NE 1 GE 1 2 G FG E R A 2,, 1 - 2,, - 1 G F N E 1, 1 - 1, 0 N FG E 1, 0 - 1, - 1 A 2, 0 N F N E 0, 0 0, 0 R - 2, 2 -2 Are both Nash equilibria reasonable? Note that subgame perfection does not help. Why? 24.07.2011 G.B. Asheim, ECON3/4200-7 5 Remedy 1: Conditional beliefs about types F1 G F 2 A 2, 2 0, 0 N ( q ) R - 2, 0 F ( 12 ) Nature E ( 12 ) (1 q ) A 2, 0 0, 0 E N 1 GE R - 2, - 2 Let player 2 assign probabilities to the two types player 1: the belief of player 2. What will the players choose? R d 22: Sequential Remedy S ti l rationality ti lit Each player chooses rationally at all information sets, given his belief and the opponent’s strategy. 24.07.2011 G.B. Asheim, ECON3/4200-7 6 3 Gift game, version 2 2 A F1 G 0 0 N 0, (q) R F ( 12 ) F 2 2, 2 R A G FG E 0,, 0 0,, 0 2, 0 G F N E 1, 1 1, 0 Nature 1 N FG E - 1, - 1 - 1, 0 1 2 E( ) (1 q ) A - 2, - 2 N F N E 0, 0 0, 0 0, 0 E N 1 GE R - 2, 2 0 If q ½, then player 2 will choose R. If so, the outcome is not a Nash equil. outcome. 24.07.2011 G.B. Asheim, ECON3/4200-7 7 Remedy 3: Consistency of beliefs F1 GF 2 A 0, 0 N (1) R F ( 12 ) Nature E ( 12 ) 0, 0 2, 2 2,, 0 Player 2 should find his belief by means of Bayes’ rule, when-ever possible. (0) A - 2, - 2 Pr[ G | F] Pr[ F] q Pr[ F | G ] NE 1 GE R - 2, 0 Pr[ G ] An example of a separating equilibrium. An equilibrium is separating if the types of a player behave differently. 24.07.2011 G.B. Asheim, ECON3/4200-7 8 4 Gift game, version 3 If q ½, then player 2 will choose R. Bayes’ rule cannot be used. 2 A 2, 2 F1 G 0 0 N 0, ( q ) R - 2, 0 F ( 12 ) F Nature 1 2 E( ) 2 R A G FG E 2,, 0 - 2,, 0 1 G F N E 1, 1 - 1, 0 N FG E 1, - 1 - 1, 0 (1 q ) A 2, - 2 N F N E 0, 0 0, 0 0, 0 E N 1 GE R - 2, 2 0 An example of a pooling equilibrium. An equilibrium is pooling if the types behave the same. 24.07.2011 G.B. Asheim, ECON3/4200-7 9 Perfect Bayesian equilibrium Definition: Consider a strategy profile for the players, as well as beliefs over the nodes at all information sets. These are called a perfect Bayesian equilibrium (PBE) if: (a) (b) 24.07.2011 each player’s strategy specifies optimal actions given his beliefs and the strategies of the other players players. the beliefs are consistent with Bayes’ rule whenever possible. G.B. Asheim, ECON3/4200-7 10 5 Algorithm for finding perfect Bayesian equilibria in a signaling game: posit a strategy for player 1 (either pooling or separating), calculate restrictions on conditional beliefs, calculate optimal actions for player 2 given his beliefs, beliefs check whether player 1’s strategy is a best response to player 2’s strategy. 24.07.2011 11 G.B. Asheim, ECON3/4200-7 Applying the algorithm in a signaling game Player 1 has four pure strategies. 1, 3 U 2 L 1 R 2 U 2, 1 (r ) (q ) D 0, 0 PBE w/(LL w/(LL)? )? YES 4,, 0 D A ( 12 ) [(LL), (DU), q , r 1 / 2] where q 2 / 3. Nature B ( 12 ) 2, 4 U (1 r ) (1 q ) U 1, 0 L 1 R D 1, 2 0, 1 D PBE w/(RR)? NO PBE w/(LR w/(LR)? )? NO [(RL), (UU), q 1, r 0] PBE w/(RL)? YES [(RL), (UU), q 1, r 0] 24.07.2011 is a separating equilibrium. [(LL), (DU), q , r 1 / 2 ] is a pooling equilibrium. G.B. Asheim, ECON3/4200-7 12 6