Axelrod ran computer tournaments in which submitted strategies played the iterated prisoner's dilemma against each other, and this book reports what won and why.
The winner in both tournaments was tit-for-tat, submitted by Anatol Rapoport, which cooperates on the first move and then copies the opponent's previous move. It was the shortest programme submitted. Axelrod identifies four properties that characterised successful strategies: being nice, meaning never defecting first; being retaliatory, meaning responding to defection immediately; being forgiving, meaning returning to cooperation as soon as the opponent does; and being clear, meaning simple enough that the opponent can recognise the pattern and adapt to it.
His theoretical result is that cooperation can emerge among self-interested agents without central authority or trust, provided the interaction is repeated and the shadow of the future is long enough, meaning the participants weight future encounters sufficiently relative to present gains.
The empirical applications include the live and let live system that developed in First World War trench warfare, where units facing each other across static lines developed tacit restraint that commands had to break up by rotating troops, and biological examples of reciprocal altruism.
The work has been enormously influential across biology, political science and economics. Later research has qualified it: tit-for-tat performs poorly with noise, since a single misperception produces endless mutual retaliation, and variants including generous tit-for-tat and win-stay-lose-shift do better in noisy environments.
Anyone designing repeated interactions, including negotiators, platform operators and partnership managers, who needs the conditions under which cooperation emerges.
Chosen by shared topics, the entries that hold them, and the people who recommended both.