Abstract
The main purpose of this study is to apply the “selective play paradigm” to explain how cooperation emerges in one-shot prisoner's dilemmas. A unique feature of the selective play paradigm is the option for not playing a PD game. For this purpose, a computer simulation of 100-actor groups was conducted. At the beginning of each replication, each simulated actor was randomly assigned to one of eleven levels of trust, which indicates the actor's estimate of the overall cooperation rate in the group. Each simulated actor, then, decided whether or not to interact with the previous partner based on the calculated expected gains from interacting with one of the other partners. Results of the simulation show that: (1) when substantial opportunity costs exist, having a high level of trust benefits the actor; (2) the above effect of trust depends on the actor's cooperativeness in PD games; (3) but does not depend on the overall cooperation rate in the group.
