Abstract
Recent articles in this journal demonstrated that cooperation until provocation is the best approach to iterated Prisoner's Dilemma when the goal is the maximization of payoffs. However, analysis of the tournament results discussed in these previous works shows that this same strategy is not necessarily beneficial when the desired end is outscormg the opponent. In the environment considered here, when either consistent victory or maximum average margin of victory is the predetermined policy objective, strategies which usually cooperate until provoked, but occasionally take advantage of cooperative opponents are clearly most successful. Even when playing a game as straightforward as Prisoner's Dilemma, strategies must be specifically tailored to fit objectives.

This publication has 5 references indexed in Scilit: