An Analysis of the Generalized Tit-for-Tat Strategy Within the Framework of Memory-One Strategies

The Tit-for-tat strategy is a traditional strategy in game theory. In the Prisoner’s Dilemma, the TFT strategy has been proven to be strong. However, within a four-component Memory-One strategy framework, the experiment proves that when the TFT strategy faces a random strategy, the expected payoffs of both sides are equal in most cases, that is, the TFT strategy does not show an absolute advantage. In this paper, a Generalized TFT strategy is set up to study the impact of each variable on its strengths. The results show that: within the framework, in order to ensure a higher expected payoff, the user of the Generalized TFT strategy should appropriately increase the possibility of choosing a strategy opposite to the opponent’s strategy in the previous round currently. These works expound the nature of the TFT strategy from a new perspective, enrich its core, and solve the problem of defining the TFT strategy in evolutionary games.