A Mathematical Model for Behavioral Changes by Pair Interactions and Its Relation to Game Theory

A mathematical model for behavioral changes by pair interactions (i.e. due to direct contact) of individuals is developed. Three kinds of pair interactions can be distinguished: Imitative processes, avoidance processes, and compromising processes. Representative solutions of the model for two different interacting subpopulations are illustrated by computational results. The equations of game theory are shown to result for a special case of imitative processes. Moreover, a stochastic version of game theory is formulated. It allows the derivation of equations for the most probable or the expected distribution of behavioral strategies and of (co)variance equations. The knowledge of the (co)variances is necessary for the calculation of the reliability of game theoretical descriptions. The use and application of the introduced equations is illustrated by concrete examples. Especially, computational results for the selforganization of social conventions by competition of two equivalent strategies are presented.