Introduction to Game Theory
Total Page:16
File Type:pdf, Size:1020Kb
Load more
Recommended publications
-
Lecture 4 Rationalizability & Nash Equilibrium Road
Lecture 4 Rationalizability & Nash Equilibrium 14.12 Game Theory Muhamet Yildiz Road Map 1. Strategies – completed 2. Quiz 3. Dominance 4. Dominant-strategy equilibrium 5. Rationalizability 6. Nash Equilibrium 1 Strategy A strategy of a player is a complete contingent-plan, determining which action he will take at each information set he is to move (including the information sets that will not be reached according to this strategy). Matching pennies with perfect information 2’s Strategies: HH = Head if 1 plays Head, 1 Head if 1 plays Tail; HT = Head if 1 plays Head, Head Tail Tail if 1 plays Tail; 2 TH = Tail if 1 plays Head, 2 Head if 1 plays Tail; head tail head tail TT = Tail if 1 plays Head, Tail if 1 plays Tail. (-1,1) (1,-1) (1,-1) (-1,1) 2 Matching pennies with perfect information 2 1 HH HT TH TT Head Tail Matching pennies with Imperfect information 1 2 1 Head Tail Head Tail 2 Head (-1,1) (1,-1) head tail head tail Tail (1,-1) (-1,1) (-1,1) (1,-1) (1,-1) (-1,1) 3 A game with nature Left (5, 0) 1 Head 1/2 Right (2, 2) Nature (3, 3) 1/2 Left Tail 2 Right (0, -5) Mixed Strategy Definition: A mixed strategy of a player is a probability distribution over the set of his strategies. Pure strategies: Si = {si1,si2,…,sik} σ → A mixed strategy: i: S [0,1] s.t. σ σ σ i(si1) + i(si2) + … + i(sik) = 1. If the other players play s-i =(s1,…, si-1,si+1,…,sn), then σ the expected utility of playing i is σ σ σ i(si1)ui(si1,s-i) + i(si2)ui(si2,s-i) + … + i(sik)ui(sik,s-i). -
Learning and Equilibrium
Learning and Equilibrium Drew Fudenberg1 and David K. Levine2 1Department of Economics, Harvard University, Cambridge, Massachusetts; email: [email protected] 2Department of Economics, Washington University of St. Louis, St. Louis, Missouri; email: [email protected] Annu. Rev. Econ. 2009. 1:385–419 Key Words First published online as a Review in Advance on nonequilibrium dynamics, bounded rationality, Nash equilibrium, June 11, 2009 self-confirming equilibrium The Annual Review of Economics is online at by 140.247.212.190 on 09/04/09. For personal use only. econ.annualreviews.org Abstract This article’s doi: The theory of learning in games explores how, which, and what 10.1146/annurev.economics.050708.142930 kind of equilibria might arise as a consequence of a long-run non- Annu. Rev. Econ. 2009.1:385-420. Downloaded from arjournals.annualreviews.org Copyright © 2009 by Annual Reviews. equilibrium process of learning, adaptation, and/or imitation. If All rights reserved agents’ strategies are completely observed at the end of each round 1941-1383/09/0904-0385$20.00 (and agents are randomly matched with a series of anonymous opponents), fairly simple rules perform well in terms of the agent’s worst-case payoffs, and also guarantee that any steady state of the system must correspond to an equilibrium. If players do not ob- serve the strategies chosen by their opponents (as in extensive-form games), then learning is consistent with steady states that are not Nash equilibria because players can maintain incorrect beliefs about off-path play. Beliefs can also be incorrect because of cogni- tive limitations and systematic inferential errors. -
Evolutionary Game Theory: ESS, Convergence Stability, and NIS
Evolutionary Ecology Research, 2009, 11: 489–515 Evolutionary game theory: ESS, convergence stability, and NIS Joseph Apaloo1, Joel S. Brown2 and Thomas L. Vincent3 1Department of Mathematics, Statistics and Computer Science, St. Francis Xavier University, Antigonish, Nova Scotia, Canada, 2Department of Biological Sciences, University of Illinois, Chicago, Illinois, USA and 3Department of Aerospace and Mechanical Engineering, University of Arizona, Tucson, Arizona, USA ABSTRACT Question: How are the three main stability concepts from evolutionary game theory – evolutionarily stable strategy (ESS), convergence stability, and neighbourhood invader strategy (NIS) – related to each other? Do they form a basis for the many other definitions proposed in the literature? Mathematical methods: Ecological and evolutionary dynamics of population sizes and heritable strategies respectively, and adaptive and NIS landscapes. Results: Only six of the eight combinations of ESS, convergence stability, and NIS are possible. An ESS that is NIS must also be convergence stable; and a non-ESS, non-NIS cannot be convergence stable. A simple example shows how a single model can easily generate solutions with all six combinations of stability properties and explains in part the proliferation of jargon, terminology, and apparent complexity that has appeared in the literature. A tabulation of most of the evolutionary stability acronyms, definitions, and terminologies is provided for comparison. Key conclusions: The tabulated list of definitions related to evolutionary stability are variants or combinations of the three main stability concepts. Keywords: adaptive landscape, convergence stability, Darwinian dynamics, evolutionary game stabilities, evolutionarily stable strategy, neighbourhood invader strategy, strategy dynamics. INTRODUCTION Evolutionary game theory has and continues to make great strides. -
Lecture Notes
GRADUATE GAME THEORY LECTURE NOTES BY OMER TAMUZ California Institute of Technology 2018 Acknowledgments These lecture notes are partially adapted from Osborne and Rubinstein [29], Maschler, Solan and Zamir [23], lecture notes by Federico Echenique, and slides by Daron Acemoglu and Asu Ozdaglar. I am indebted to Seo Young (Silvia) Kim and Zhuofang Li for their help in finding and correcting many errors. Any comments or suggestions are welcome. 2 Contents 1 Extensive form games with perfect information 7 1.1 Tic-Tac-Toe ........................................ 7 1.2 The Sweet Fifteen Game ................................ 7 1.3 Chess ............................................ 7 1.4 Definition of extensive form games with perfect information ........... 10 1.5 The ultimatum game .................................. 10 1.6 Equilibria ......................................... 11 1.7 The centipede game ................................... 11 1.8 Subgames and subgame perfect equilibria ...................... 13 1.9 The dollar auction .................................... 14 1.10 Backward induction, Kuhn’s Theorem and a proof of Zermelo’s Theorem ... 15 2 Strategic form games 17 2.1 Definition ......................................... 17 2.2 Nash equilibria ...................................... 17 2.3 Classical examples .................................... 17 2.4 Dominated strategies .................................. 22 2.5 Repeated elimination of dominated strategies ................... 22 2.6 Dominant strategies .................................. -
A New Approach to Stable Matching Problems
August1 989 Report No. STAN-CS-89- 1275 A New Approach to Stable Matching Problems bY Ashok Subramanian Department of Computer Science Stanford University Stanford, California 94305 3 DISTRIBUTION /AVAILABILITY OF REPORT Unrestricted: Distribution Unlimited GANIZATION 1 1 TITLE (Include Securrty Clamfrcat!on) A New Approach to Stable Matching Problems 12 PERSONAL AUTHOR(S) Ashok Subramanian 13a TYPE OF REPORT 13b TtME COVERED 14 DATE OF REPORT (Year, Month, Day) 15 PAGE COUNT FROM TO August 1989 34 16 SUPPLEMENTARY NOTATION 17 COSATI CODES 18 SUBJECT TERMS (Contrnue on reverse If necessary and jdentrfy by block number) FIELD GROUP SUB-GROUP 19 ABSTRACT (Continue on reverse if necessary and identrfy by block number) Abstract. We show that Stable Matching problems are the same as problems about stable config- urations of X-networks. Consequences include easy proofs of old theorems, a new simple algorithm for finding a stable matching, an understanding of the difference between Stable Marriage and Stable Roommates, NTcompleteness of Three-party Stable Marriage, CC-completeness of several Stable Matching problems, and a fast parallel reduction from the Stable Marriage problem to the ’ Assignment problem. 20 DISTRIBUTION /AVAILABILITY OF ABSTRACT 21 ABSTRACT SECURITY CLASSIFICATION q UNCLASSIFIED/UNLIMITED 0 SAME AS Rf’T 0 DTIC USERS 22a NAME OF RESPONSIBLE INDIVIDUAL 22b TELEPHONE (Include Area Code) 22c OFFICE SYMBOL Ernst Mavr DD Form 1473, JUN 86 Prevrous edrtions are obsolete SECURITY CLASSIFICATION OF TYS PAGt . 1 , S/N 0102-LF-014-6603 \ \ ““*5 - - A New Approach to Stable Matching Problems * Ashok Subramanian Department of Computer Science St anford University Stanford, CA 94305-2140 Abstract. -
8. Maxmin and Minmax Strategies
CMSC 474, Introduction to Game Theory 8. Maxmin and Minmax Strategies Mohammad T. Hajiaghayi University of Maryland Outline Chapter 2 discussed two solution concepts: Pareto optimality and Nash equilibrium Chapter 3 discusses several more: Maxmin and Minmax Dominant strategies Correlated equilibrium Trembling-hand perfect equilibrium e-Nash equilibrium Evolutionarily stable strategies Worst-Case Expected Utility For agent i, the worst-case expected utility of a strategy si is the minimum over all possible Husband Opera Football combinations of strategies for the other agents: Wife min u s ,s Opera 2, 1 0, 0 s-i i ( i -i ) Football 0, 0 1, 2 Example: Battle of the Sexes Wife’s strategy sw = {(p, Opera), (1 – p, Football)} Husband’s strategy sh = {(q, Opera), (1 – q, Football)} uw(p,q) = 2pq + (1 – p)(1 – q) = 3pq – p – q + 1 We can write uw(p,q) For any fixed p, uw(p,q) is linear in q instead of uw(sw , sh ) • e.g., if p = ½, then uw(½,q) = ½ q + ½ 0 ≤ q ≤ 1, so the min must be at q = 0 or q = 1 • e.g., minq (½ q + ½) is at q = 0 minq uw(p,q) = min (uw(p,0), uw(p,1)) = min (1 – p, 2p) Maxmin Strategies Also called maximin A maxmin strategy for agent i A strategy s1 that makes i’s worst-case expected utility as high as possible: argmaxmin ui (si,s-i ) si s-i This isn’t necessarily unique Often it is mixed Agent i’s maxmin value, or security level, is the maxmin strategy’s worst-case expected utility: maxmin ui (si,s-i ) si s-i For 2 players it simplifies to max min u1s1, s2 s1 s2 Example Wife’s and husband’s strategies -
Economics 201B Economic Theory (Spring 2021) Strategic Games
Economics 201B Economic Theory (Spring 2021) Strategic Games Topics: terminology and notations (OR 1.7), games and solutions (OR 1.1-1.3), rationality and bounded rationality (OR 1.4-1.6), formalities (OR 2.1), best-response (OR 2.2), Nash equilibrium (OR 2.2), 2 2 examples × (OR 2.3), existence of Nash equilibrium (OR 2.4), mixed strategy Nash equilibrium (OR 3.1, 3.2), strictly competitive games (OR 2.5), evolution- ary stability (OR 3.4), rationalizability (OR 4.1), dominance (OR 4.2, 4.3), trembling hand perfection (OR 12.5). Terminology and notations (OR 1.7) Sets For R, ∈ ≥ ⇐⇒ ≥ for all . and ⇐⇒ ≥ for all and some . ⇐⇒ for all . Preferences is a binary relation on some set of alternatives R. % ⊆ From % we derive two other relations on : — strict performance relation and not  ⇐⇒ % % — indifference relation and ∼ ⇐⇒ % % Utility representation % is said to be — complete if , or . ∀ ∈ % % — transitive if , and then . ∀ ∈ % % % % can be presented by a utility function only if it is complete and transitive (rational). A function : R is a utility function representing if → % ∀ ∈ () () % ⇐⇒ ≥ % is said to be — continuous (preferences cannot jump...) if for any sequence of pairs () with ,and and , . { }∞=1 % → → % — (strictly) quasi-concave if for any the upper counter set ∈ { ∈ : is (strictly) convex. % } These guarantee the existence of continuous well-behaved utility function representation. Profiles Let be a the set of players. — () or simply () is a profile - a collection of values of some variable,∈ one for each player. — () or simply is the list of elements of the profile = ∈ { } − () for all players except . ∈ — ( ) is a list and an element ,whichistheprofile () . -
Alpha-Beta Pruning
Alpha-Beta Pruning Carl Felstiner May 9, 2019 Abstract This paper serves as an introduction to the ways computers are built to play games. We implement the basic minimax algorithm and expand on it by finding ways to reduce the portion of the game tree that must be generated to find the best move. We tested our algorithms on ordinary Tic-Tac-Toe, Hex, and 3-D Tic-Tac-Toe. With our algorithms, we were able to find the best opening move in Tic-Tac-Toe by only generating 0.34% of the nodes in the game tree. We also explored some mathematical features of Hex and provided proofs of them. 1 Introduction Building computers to play board games has been a focus for mathematicians ever since computers were invented. The first computer to beat a human opponent in chess was built in 1956, and towards the late 1960s, computers were already beating chess players of low-medium skill.[1] Now, it is generally recognized that computers can beat even the most accomplished grandmaster. Computers build a tree of different possibilities for the game and then work backwards to find the move that will give the computer the best outcome. Al- though computers can evaluate board positions very quickly, in a game like chess where there are over 10120 possible board configurations it is impossible for a computer to search through the entire tree. The challenge for the computer is then to find ways to avoid searching in certain parts of the tree. Humans also look ahead a certain number of moves when playing a game, but experienced players already know of certain theories and strategies that tell them which parts of the tree to look. -
14.12 Game Theory Lecture Notes∗ Lectures 7-9
14.12 Game Theory Lecture Notes∗ Lectures 7-9 Muhamet Yildiz Intheselecturesweanalyzedynamicgames(withcompleteinformation).Wefirst analyze the perfect information games, where each information set is singleton, and develop the notion of backwards induction. Then, considering more general dynamic games, we will introduce the concept of the subgame perfection. We explain these concepts on economic problems, most of which can be found in Gibbons. 1 Backwards induction The concept of backwards induction corresponds to the assumption that it is common knowledge that each player will act rationally at each node where he moves — even if his rationality would imply that such a node will not be reached.1 Mechanically, it is computed as follows. Consider a finite horizon perfect information game. Consider any node that comes just before terminal nodes, that is, after each move stemming from this node, the game ends. If the player who moves at this node acts rationally, he will choose the best move for himself. Hence, we select one of the moves that give this player the highest payoff. Assigning the payoff vector associated with this move to the node at hand, we delete all the moves stemming from this node so that we have a shorter game, where our node is a terminal node. Repeat this procedure until we reach the origin. ∗These notes do not include all the topics that will be covered in the class. See the slides for a more complete picture. 1 More precisely: at each node i the player is certain that all the players will act rationally at all nodes j that follow node i; and at each node i the player is certain that at each node j that follows node i the player who moves at j will be certain that all the players will act rationally at all nodes k that follow node j,...ad infinitum. -
How to Win at Tic-Tac-Toe
More Than Child’s Play How to Get N in a Row Games with Animals Hypercube Tic-Tac-Toe How to Win at Tic-Tac-Toe Norm Do Undoubtably, one of the most popular pencil and paper games in the world is tic-tac-toe, also commonly known as noughts and crosses. In this talk, you will learn how to beat your friends (at tic-tac-toe), discover why snaky is so shaky, and see the amazing tic-tac-toe playing chicken! March 2007 Norm Do How to Win at Tic-Tac-Toe Tic-Tac-Toe is popular: You’ve all played it while sitting at the back of a boring class. In fact, some of you are probably playing it right now! Tic-Tac-Toe is boring: People who are mildly clever should never lose. More Than Child’s Play How to Get N in a Row Some Facts About Tic-Tac-Toe Games with Animals Games to Beat your Friends With Hypercube Tic-Tac-Toe Some facts about tic-tac-toe Tic-Tac-Toe is old: It may have been played under the name of “terni lapilli” in Ancient Rome. Norm Do How to Win at Tic-Tac-Toe Tic-Tac-Toe is boring: People who are mildly clever should never lose. More Than Child’s Play How to Get N in a Row Some Facts About Tic-Tac-Toe Games with Animals Games to Beat your Friends With Hypercube Tic-Tac-Toe Some facts about tic-tac-toe Tic-Tac-Toe is old: It may have been played under the name of “terni lapilli” in Ancient Rome. -
Daniel Kahneman Curriculum Vitae August 2016
Daniel Kahneman Curriculum Vitae August 2016 Born: 1934, Tel Aviv, Israel Citizenship: US, Israel Education Ph.D. University of California, Berkeley, 1961 (Psychology) B.A. The Hebrew University, Jerusalem, 1954, (Psychology and Mathematics) Professional Positions Held 2007- Professor of Psychology and Public Affairs, Emeritus, Woodrow Wilson School, Princeton University 2007- Eugene Higgins Professor of Psychology, Emeritus, Princeton University 2000- Fellow, Center for Rationality, Hebrew University, Jerusalem 1993-2007 Eugene Higgins Professor of Psychology, Princeton University 1993-2007 Professor of Psychology and Public Affairs, Woodrow Wilson School, Princeton University 1991-1992 Visiting Scholar, Russell Sage Foundation 1986-1994 Professor of Psychology, University of California, Berkeley 1984-1986 Associate Fellow, Canadian Institute for Advanced Research 1978-1986 Professor of Psychology, The University of British Columbia 1977-1978 Fellow, Center for Advanced Studies in the Behavioral Sciences 1968-1969 Visiting Scientist (summers), Applied Psychological Research Unit, Cambridge, England 1966-1967 Fellow, Center for Cognitive Studies; Lecturer in Psychology, Harvard University 1965-1966 Visiting Scientist, Department of Psychology, University of Michigan 1961-1978 Lecturer to Professor, The Hebrew University, Jerusalem, Israel LINKS TO RECENT LECTURES Video and Audio Links “Interview with Leif Nelson,” The Society for Judgment and Decision Making, November 2015. “Thinking That We Know”, Sackler Colloquium to the National Academy -
Game Theory with Translucent Players
Game Theory with Translucent Players Joseph Y. Halpern and Rafael Pass∗ Cornell University Department of Computer Science Ithaca, NY, 14853, U.S.A. e-mail: fhalpern,[email protected] November 27, 2012 Abstract A traditional assumption in game theory is that players are opaque to one another—if a player changes strategies, then this change in strategies does not affect the choice of other players’ strategies. In many situations this is an unrealistic assumption. We develop a framework for reasoning about games where the players may be translucent to one another; in particular, a player may believe that if she were to change strategies, then the other player would also change strategies. Translucent players may achieve significantly more efficient outcomes than opaque ones. Our main result is a characterization of strategies consistent with appro- priate analogues of common belief of rationality. Common Counterfactual Belief of Rationality (CCBR) holds if (1) everyone is rational, (2) everyone counterfactually believes that everyone else is rational (i.e., all players i be- lieve that everyone else would still be rational even if i were to switch strate- gies), (3) everyone counterfactually believes that everyone else is rational, and counterfactually believes that everyone else is rational, and so on. CCBR characterizes the set of strategies surviving iterated removal of minimax dom- inated strategies: a strategy σi is minimax dominated for i if there exists a 0 0 0 strategy σ for i such that min 0 u (σ ; µ ) > max u (σ ; µ ). i µ−i i i −i µ−i i i −i ∗Halpern is supported in part by NSF grants IIS-0812045, IIS-0911036, and CCF-1214844, by AFOSR grant FA9550-08-1-0266, and by ARO grant W911NF-09-1-0281.