Extensive form games - contd

Similar documents
Finitely repeated simultaneous move game.

Sequential Rationality and Weak Perfect Bayesian Equilibrium

Beliefs and Sequential Rationality

Extensive-Form Games with Imperfect Information

The Ohio State University Department of Economics Econ 601 Prof. James Peck Extra Practice Problems Answers (for final)

Lecture 6 Dynamic games with imperfect information

G5212: Game Theory. Mark Dean. Spring 2017

Answers to Problem Set 4

M.Phil. Game theory: Problem set II. These problems are designed for discussions in the classes of Week 8 of Michaelmas term. 1

FDPE Microeconomics 3 Spring 2017 Pauli Murto TA: Tsz-Ning Wong (These solution hints are based on Julia Salmi s solution hints for Spring 2015.

ECONS 424 STRATEGY AND GAME THEORY HANDOUT ON PERFECT BAYESIAN EQUILIBRIUM- III Semi-Separating equilibrium

In reality; some cases of prisoner s dilemma end in cooperation. Game Theory Dr. F. Fatemi Page 219

Game Theory. Important Instructions

Problem 3 Solutions. l 3 r, 1

Lecture Notes on Adverse Selection and Signaling

CUR 412: Game Theory and its Applications Final Exam Ronaldo Carpio Jan. 13, 2015

BAYESIAN GAMES: GAMES OF INCOMPLETE INFORMATION

ECONS 424 STRATEGY AND GAME THEORY HOMEWORK #7 ANSWER KEY

Advanced Micro 1 Lecture 14: Dynamic Games Equilibrium Concepts

Spring 2017 Final Exam

4. Beliefs at all info sets off the equilibrium path are determined by Bayes' Rule & the players' equilibrium strategies where possible.

Simon Fraser University Spring 2014

Agenda. Game Theory Matrix Form of a Game Dominant Strategy and Dominated Strategy Nash Equilibrium Game Trees Subgame Perfection

Out of equilibrium beliefs and Refinements of PBE

GAME THEORY: DYNAMIC. MICROECONOMICS Principles and Analysis Frank Cowell. Frank Cowell: Dynamic Game Theory

Rationalizable Strategies

Almost essential MICROECONOMICS

Introduction to Game Theory Lecture Note 5: Repeated Games

ECONS STRATEGY AND GAME THEORY QUIZ #3 (SIGNALING GAMES) ANSWER KEY

Dynamic games with incomplete information

Not 0,4 2,1. i. Show there is a perfect Bayesian equilibrium where player A chooses to play, player A chooses L, and player B chooses L.

Game Theory: Additional Exercises

CUR 412: Game Theory and its Applications, Lecture 11

Microeconomic Theory II Preliminary Examination Solutions

In the Name of God. Sharif University of Technology. Microeconomics 2. Graduate School of Management and Economics. Dr. S.

6.254 : Game Theory with Engineering Applications Lecture 3: Strategic Form Games - Solution Concepts

Microeconomics II. CIDE, MsC Economics. List of Problems

CUR 412: Game Theory and its Applications, Lecture 12

6 Dynamic Games with Incomplete Information

G5212: Game Theory. Mark Dean. Spring 2017

Microeconomics of Banking: Lecture 5

1 Solutions to Homework 4

Microeconomic Theory II Spring 2016 Final Exam Solutions

Preliminary Notions in Game Theory

CMSC 474, Introduction to Game Theory 16. Behavioral vs. Mixed Strategies

Answer Key: Problem Set 4

University of Hong Kong ECON6036 Stephen Chiu. Extensive Games with Perfect Information II. Outline

CHAPTER 15 Sequential rationality 1-1

HW Consider the following game:

Finish what s been left... CS286r Fall 08 Finish what s been left... 1

ECON 803: MICROECONOMIC THEORY II Arthur J. Robson Fall 2016 Assignment 9 (due in class on November 22)

Simon Fraser University Fall Econ 302 D200 Final Exam Solution Instructor: Songzi Du Wednesday December 16, 2015, 8:30 11:30 AM

Game Theory. Wolfgang Frimmel. Repeated Games

1 Solutions to Homework 3

1 R. 2 l r 1 1 l2 r 2

The Intuitive and Divinity Criterion: Explanation and Step-by-step examples

Games of Incomplete Information

Economics 171: Final Exam

Microeconomics Comprehensive Exam

(a) (5 points) Suppose p = 1. Calculate all the Nash Equilibria of the game. Do/es the equilibrium/a that you have found maximize social utility?

G5212: Game Theory. Mark Dean. Spring 2017

FDPE Microeconomics 3 Spring 2017 Pauli Murto TA: Tsz-Ning Wong (These solution hints are based on Julia Salmi s solution hints for Spring 2015.

Credibility and Commitment in Crisis Bargaining

Copyright 2008, Yan Chen

Game Theory with Applications to Finance and Marketing, I

Economics 209A Theory and Application of Non-Cooperative Games (Fall 2013) Repeated games OR 8 and 9, and FT 5

SI 563 Homework 3 Oct 5, Determine the set of rationalizable strategies for each of the following games. a) X Y X Y Z

An introduction on game theory for wireless networking [1]

Lecture 1: Normal Form Games: Refinements and Correlated Equilibrium

In Class Exercises. Problem 1

Topics in Contract Theory Lecture 1

Chapter 6. Game Theory

CUR 412: Game Theory and its Applications, Lecture 9

Lecture 3 Representation of Games

EconS Signalling Games II

National Security Strategy: Perfect Bayesian Equilibrium

Introduction to Industrial Organization Professor: Caixia Shen Fall 2014 Lecture Note 5 Games and Strategy (Ch. 4)

Credibility and Commitment in Crisis Bargaining

Games with incomplete information about players. be symmetric or asymmetric.

ECONS 424 STRATEGY AND GAME THEORY MIDTERM EXAM #2 ANSWER KEY

GAME THEORY. Department of Economics, MIT, Follow Muhamet s slides. We need the following result for future reference.

Corporate Finance: Asymmetric information and capital structure signaling. Yossi Spiegel Recanati School of Business

Advanced Microeconomics

Outline for Dynamic Games of Complete Information

Economics 502 April 3, 2008

The Ohio State University Department of Economics Second Midterm Examination Answers

Ph.D. Preliminary Examination MICROECONOMIC THEORY Applied Economics Graduate Program August 2017

Subgame Perfect Cooperation in an Extensive Game

Dynamic Games. Econ 400. University of Notre Dame. Econ 400 (ND) Dynamic Games 1 / 18

Comparing Allocations under Asymmetric Information: Coase Theorem Revisited

Notes for Section: Week 4

MIDTERM ANSWER KEY GAME THEORY, ECON 395

MA200.2 Game Theory II, LSE

Repeated Games. September 3, Definitions: Discounting, Individual Rationality. Finitely Repeated Games. Infinitely Repeated Games

Supplementary Material for: Belief Updating in Sequential Games of Two-Sided Incomplete Information: An Experimental Study of a Crisis Bargaining

Stochastic Games and Bayesian Games

ECE 586BH: Problem Set 5: Problems and Solutions Multistage games, including repeated games, with observed moves

Duopoly models Multistage games with observed actions Subgame perfect equilibrium Extensive form of a game Two-stage prisoner s dilemma

Signaling Games. Farhad Ghassemi

Best-Reply Sets. Jonathan Weinstein Washington University in St. Louis. This version: May 2015

Transcription:

Extensive form games - contd

Proposition: Every finite game of perfect information Γ E has a pure-strategy SPNE. Moreover, if no player has the same payoffs in any two terminal nodes, then there is a unique SPNE. Idea of sequential rationality can be extended to games with imperfect information. 1

Firm E Out In U(E) 0 U(I) 2 Firm E Fight Accommodate Firm I Fight Accomm Fight Accomm -3 1-2 3-1 -2-1 1 Firm E Firm I Accommodat Fight Accommodat 3,1-2,-1 Fight 1,-2-3,-1 I Accomm if E Fight if E plays In plays In Out, Accomm if In 0,2 0,2 E Out, Fight if In 0,2 0,2 In, Accomm if In 3,1-2,-1 In, Fight if In 1,-2-3,-1 2

Normal form version has 3 pure strategy NE: ((out, accomm if in),(fight if firm E plays in)); ((out, fight if in),(fight if firm E plays in)); ((in, accomm if in),(accomm if firm E plays in)) However (accomm, accomm) is the sole NE in the simultaneous move game that follows entry. Logic of sequential rationality implies that only the last equilibrium is a SPNE. 3

Above game had unique NE in post-entry subgame. In such games, the SPNE concept rules out history dependence of strategies. With more than 1 NE in post-entry game, behaviour earlier could depend on which equilibrium resulted after entry. 4

Firm E Out In U(E) 0 U(I) 2 Firm E Small niche Large niche Firm I Small Large Small Large niche niche niche niche -6-1 1-3 -6 1-1 -3 Firm E Firm I Small Large Small -6,-6-1,1 Large 1,-1-3,-3 E E Out In Out In 0 1 0-1 2-1 2 1 5

Two pure strategy NE in post-entry game: (large niche, small niche); (small niche, large niche) Suppose that firms will play (large niche, small niche). Then entrant optimally chooses to enter. So one SPNE is (σ E, σ I ) =((in, large niche if in),(small niche if firm E plays in)). Other SPNE is (σ E, σ I ) =((out, small niche if in),(large niche if firm E plays in)). 6

Beliefs and sequential rationality Firm E Out In1 In2 Firm I U(E) 0 U(I) 2 Fight Accomm Fight Accomm -1 3-1 2-1 0-1 1 Fight if Accomm if entry occurs entry occurs Out 0,2 0,2 In1-1,-1 3,0 In2-1,-1 2,1 7

Two pure-strategy NE in above game: (out, fight if entry occurs), (in1, accomm if entry occurs) But again, (out, fight if entry occurs) does not seem reasonable; regardless of type of entry (In1, In2), incumbent prefers to accommodate once entry occurs. But criterion of subgame perfection cannot be applied here to eliminate non-sensible equilibrium, since only subgame is entire game! 8

New equilibrium concept - weak perfect Bayesian equilibrium (weak PBE) or weak sequential equilibrium Requires that at any point in game, a player s strategy prescribe optimal actions from that point on, given her opponents strategies and her beliefs about what has happened so far, and that her beliefs be consistent with the strategies being played. 9

Definition: A system of beliefs µ in extensive form game Γ E is a specification of a probability µ(x) [0, 1] for each decision node x in Γ E such that Σ x H µ(x) = 1 for all information sets H. 10

Definition: A strategy profile σ = (σ 1,.., σ I ) in extensive form Γ E is sequentially rational at information set H given a system of beliefs µ if, denoting by ι(µ) the player who moves at information set H, we have E[u ι(h) H, µ, σ ι(h), σ ι(h) ] E[u ι(h) H, µ, σ ι(h), σ ι(h) ] for all σ ι(h) (S ι(h) ). If strategy profile σ satisfies this condition for all information sets H, then we say that σ is sequentially rational given belief system µ. 11

Suppose each player s equilibrium strategy assigns a strictly positive probability to each possible action at every one of her information sets. i.e., every information set is reached with positive probability. For each node x H i, i should compute probability of reaching that node given play of strategies σ, P r(x σ) and she should assign conditional probabilities using Baye s rule: P r(x H, σ) = P r(x σ) Σ x H P r(x σ) 12

Suppose in game above, firm E is using the completely mixed strategy such that P r( out ) = 1 4, P r( In1 ) = 1 2 and P r( In2 ) = 1 4. Probability of reaching firm I s information set given this strategy is 3 4. Using Baye s rule, probability of being at the left node of I s information set conditional on reaching it is 2 3 and that of being at the right node is 1 3. 13

Pr(reaching information set)=pr(not out )=1-1 4 = 3 4 Pr(L/not out )= P r(l).p r(not out L) P r(l).p r(not out L)+P r(r).p r(not out R) = (1/2).1 (1/2).1+((1/4).1 = 2 3 Pr(L/not out )=1 2 3 = 1 3 14

When players are not playing completely mixed strategies, some information sets may not be reached with positive probability. We then cannot use Baye s rule to compute conditional probabilities. In such cases, the weak PBE concept allows us to assign any beliefs at these information sets (hence the term weak ). 15

Definition: A profile of strategies and system of beliefs (σ, µ) is a weak PBE in extensive form game Γ E if it has the following properties: (i) The strategy profile σ is sequentially rational given belief system µ; (ii) The system of beliefs µ is derived through Bayes rule whenever possible. 16

Player 1 Out 2 2 In Player 1 U D p Player 2 (1-p) L R L R 3 0 0 1 1 0 0 3 Player 2's optimal strategy is to play L for p>(2/3) But only p=1 is sequentially rational 17

One shortcoming of the weak PBE concept is that it places no restrictions on beliefs off the equilibrium path. Because of this, a weak PBE need not be subgame perfect. Lets go back to the 1st game we looked at. 18

Firm E Out In U(E) 0 U(I) 2 Firm E Fight Accommodate Firm I Fight Accomm Fight Accomm -3 1-2 3-1 -2-1 1 Firm E Firm I Accommodat Fight Accommodat 3,1-2,-1 Fight 1,-2-3,-1 I Accomm if E Fight if E plays In plays In Out, Accomm if In 0,2 0,2 E Out, Fight if In 0,2 0,2 In, Accomm if In 3,1-2,-1 In, Fight if In 1,-2-3,-1 19

A weak PBE of this game is (σ E, σ I )=((out, accommodate if in),(fight if firm E plays in )) with beliefs for firm I that assign probability 1 to firm E having played fight. But these strategies are not subgame perfect; they are not a NE in the post-entry subgame. Problem arises because firm I s post-entry belief about firm E s post-entry play is unrestricted by the weak PBE concept - firm I s information set is off the equilibrium path. 20