Title:Towards a Programmable Framework for Agent Game Playing

Abstract:The field of Game Theory provides a useful mechanism for modeling many decision-making scenarios. In participating in these scenarios individuals and groups adopt particular strategies, which generally perform with varying levels of success. However, most results have focussed on players that play the same game in an iterated fashion. This paper describes a framework which can be used to observe agents when they do not know in advance which game they are going to play. That is, the same group of agents could first play a few rounds of the Iterated Prisoner's Dilemma, and then a few rounds of the Linear Public Goods Game, and then a few rounds of Minority Game, or perhaps all games in a strictly alternating fashion or a randomized instantiation of games. This framework will allow for investigation of agents in more complex settings, when there is uncertainty about the future, and limited resources to store strategies.