Sympathy and Interaction Frequency in the Prisoner´s Dilemma

People deal with relationships at home, workplace, and local community. There are various levels of friendships. According to the level of social interaction, different degrees of sympathy and interaction frequency exist. This paper purports to examine how interaction frequency influences the development of sympathy factor, and thereby the emergence of social action patterns in an agent-based simulation model when learning takes place locally. In our artificial society, an agent frequently interacts with other agents who are closely located, and infrequently with those who are remotely located. By endogenizing the sympathy factor based on behavioral patterns of the agents, we examine how the interaction frequency affects patterns of emerging social actions. The results obtained are as follows: (1) the closer the distance and the relationship between agents, the more cooperative the behavior of agents are; (2) since cooperative behaviors promote the emotion of sympathy, frequent interactions induce more cooperative behaviors through strengthening sympathy than they do without this effect of sympathy. (1) and (2) together give rise to three patterns of social behaviors: cooperation with tolerance among players who are in a close relationship, reciprocal cooperation among players in a moderately close relationship, and no coorperation among strangers.