Game Theory and Incomplete Information

Author: Chen Jianguo
Publisher:
Publish Date: 2006-12-01
Features: In traditional game theory, "incomplete information" in games refers only to the fact that some participants do not know the types of their opponents. In fact, in social and economic games, the payoff matrices of game participants are often unknown to their opponents. This book discusses the behavior of game participants under asymmetric information—primarily the payoff matrices of participants. Clearly, if game participants decide their strategy choices based solely on their estimated information about their opponents' payoff matrices rather than the true payoff matrices of their opponents, the game typically cannot reach its Nash equilibrium. However, under long-term conditions, this book demonstrates that whether it is a fixed-opponent repeated game or a non-fixed-opponent repeated game, if the history of the game is common knowledge to all participants, or if the information about opponents can be infinitely close under long-term conditions, then the repeated game can converge to its Nash equilibrium point. This book also criticizes the von Neumann-Morgenstern criterion (expected utility maximization, or the compound game axiom) for individual decision-making under uncertainty, arguing that the goal of individual decision-making is not to maximize expected utility. Based on the proposal of the welfare stability axiom, the author extends and proves Allais's utility criterion. This book is suitable as a reference for senior undergraduate and graduate students in economics and social science departments, as well as for economists and social science researchers and readers interested in game theory.