Abstract
This text offers a review of the literature on the evolution of cooperation in the Prisoner's Dilemma. The scope of the article is limited to the two-person repeated game and with respect to the emergence of cooperation it is focused upon the mechanism of direct reciprocity.
The results of mathematical analysis and those of the related agent-based models are presented side by side. The various ways of examining the success of strategies are the dominant topic in the first part of the article.
The second part describes multi-agent simulations and their outcomes and argues that the three most important factors (at least from the international relations point of view) influencing prospects of cooperation are noise, the structure of interactions, and the payoff matrix.