试析博弈困境的两种解决方案——以旅行者困境为例  

An Analysis on Two Solutions of Game Dilemmas

作　　者：郭美云[1] 党学哲 

机构地区：[1]西南大学逻辑与智能研究中心,重庆市400715

出　　处：《西南大学学报（社会科学版）》2015年第2期18-25,189,共8页Journal of Southwest University(Social Sciences Edition)

基　　金：重庆市人文社会科学重点研究基地重点项目"博弈的逻辑与认知基础研究"(14SKB047);项目负责人:郭美云;国家社会科学基金重点项目"现代逻辑视野的认知研究"(11AZD57);项目负责人:何向东;中央高校基本科研业务费专项资金资助项目"博弈中互动认知的逻辑研究"(SWU1309380);项目负责人:郭美云

摘　　要：在博弈困境中,纳什均衡所预测的博弈解与人们的直观相矛盾,而且越来越多的博弈实验表明,这与现实博弈的结果也存在巨大反差。博弈困境的出现引起了人们对经典博弈论理性预设的质疑,从而提出不同的逻辑模型和解决方案。本文以旅行者困境为例,对Halpern&Pass提出的重复后悔度最小化模型和Capraro提出的基于联盟与合作的概率推理模型两种方案进行分析比较,考察它们是如何成功地解释和预测旅行者困境中选手实际博弈行为,并对这两种方案的普适性问题和应用前景进行了展望,同时对博弈论的基础进行了哲学反思。In game dilemmas,the game solutions that Nash Equilibrium predict are contradictory with human being's intuitions.The game dilemmas give rise to some questionings on rational assumptions of the standard game theory.Hereby some logical models and solutions are proposed.In this paper,taking the Traveler's Dilemma as an example,we make a comparative study on the model of iterated regret minimization proposed by Halpern & Pass(2011)and the probabilistic reasoning model based on coalition and cooperation proposed by Capraro(2013).We show that how the two new solutions can successfully resolve the Traveler's Dilemma,and point out some problems when they are generalized and applied in other game dilemmas.Finally,we give some philosophical reflections on the foundations of game theory.

关 键 词：博弈困境 旅行者困境 解决方案 重复后悔最小化模型 概率推理模型 

分 类 号：B815[哲学宗教—逻辑学]