Strategy revision phase with payoff threshold in the public goods game

Abstract

Abstract Commonly, the strategy revision phase in evolutionary games relies on payoff comparison. Namely, agents compare their payoff with the opponent, assessing whether changing strategy can be potentially convenient. Even tiny payoff differences can be crucial in this decision process. In this work, we study the dynamics of cooperation in the public goods game, introducing a threshold ε in the strategy revision phase. In doing so, payoff differences narrower than ε entail the decision process reduces to a coin flip. Interestingly, with ordinary agents, results show that payoff thresholds curb the emergence of cooperation. Yet, the latter can be sustained by these thresholds if the population is composed of conformist agents, which replace the random-based revision with selecting the strategy of the majority. To conclude, agents sensible only to consistent payoff differences may represent ‘real-world’ individuals unable to properly appreciate advantages or disadvantages when facing a dilemma. These agents may be detrimental to the emergence of cooperation or, on the contrary, supportive when endowed with a conformist attitude.