Imitator strategies (Q1087498)

An imitator strategy in an evolutionary game is a pair (u,p) consisting of a conventional and generally-known strategy u (the personal strategy) and a probability p of using the opponent's personal strategy rather than one's own. The question of whether imitator strategies \((p>0)\) can be used to invade populations using conventional ESSs is explored, and is found to depend in part on the variability of strategy present in the initial population. Interior values of p \((0<p<1)\) are found never to be optimal. Examples are used repeatedly to illustrate the theory. The imitator strategy is also compared with the successful Tit-for-Tat strategy considered in the \textit{R. Axelrod} and \textit{W. D. Hamilton} study of the prisoner's dilemma [The evolution of cooperation. Science 211, 1390-1396 (1981)].