State-Dependent Risk Preferences in Evolutionary Games

Abstract. There is much empirical evidence that human decisionmaking under risk does not correspond the decision-theoretic notion of “rational” decision making, namely to make choices that maximize the expected value. An open question is how such behavior could have arisen evolutionarily. We believe that the answer to this question lies, at least in part, in the interplay between risk-taking and sequentiality of choice in evolutionary environments. We provide analytical and simulation results for evolutionary game environments where sequential decisions are made between risky and safe choices. Our results show there are evolutionary games in which agents with state-dependent risk preferences (i.e., agents that are sometimes riskprone and sometimes risk-averse depending on the outcomes of their previous decisions) can outperform agents that make decisions solely based on the local expected values of the outcomes.

• Type: Conference
• Year: 2010
• Where: SBP
• Authors: Patrick Roos, Dana S. Nau
• Keywords: Applied Computing, Evolutionary Game Environments, Evolutionary Games, Expected Values, SBP 2010