Mardi | 2016-11-22
16h00-17h20 en salle des thèses

Simone RIGHI – Karoly TAKACS

Previous studies show that direct and indirect reciprocity are good candidates to explain the fundamental problem of evolution of cooperation between unrelated individuals. The importance of different forms of reciprocity, however, might be differentiated and one form could potentially drive out the feasibility of others. In the Prisoner’s Dilemma played in random networks we show that when present, direct reciprocity gains dominance over network-based indirect reciprocity as well as over network-independent generalized reciprocity. When direct reciprocity is absent, indirect reciprocity strategies are able to become dominant in the population and to supply a higher cooperation rate than that is attainable by direct reciprocity. Forgiveness is a characteristic that improves the performance of all reciprocal strategies in the long run.Limited memory quickly disrupts the reign of direct reciprocity and leaves place for the spread of network-based indirect reciprocity. Population size and network density seems to have an inversely U-shaped effect on cooperation in the absence of direct reciprocal strategies.