The book is an introductory literature to game theory. The book mainly speaks about Prisoner’s dilemma (PD) which is also a model for demonstrating real behaviour of structures in our society.
In the standard PD, there are two players which can decide whether they will betray the other player1 (C) or refuse to testify (R). The award function for PD is something between these lines:
| A\B | Stay silent | Betray |
| Stay silent | 1\1 | 0\5 |
| Betray | 5\0 | 3\3 |
Axelrod shows that one iteration of PD (without the prospect of two players cooperating again) leads to uncooperative behaviour. Instead of risking being the one betrayed, it is advantageous both both players to betray each other (since the temptation to betray the other player is high). When we switch to iterated PD, though, players tend to develop cooperative strategy since they are rewarded more points in the long run (as can be seen in the table above). So instead of “playing it safe”, players cooperate to maximise their score which is beneficial for both sides. The temptation to betray still exists, though.
For this reason, it turns out that the optimal strategy is the so called tit for tat strategy. This strategy is based on reciprocating both betrayal and cooperation. This way, player shows that he is both forgiving (in case of betrayal) and shows that he does not let himself be abused. As a result, the other side has no other option but to cooperate.
Strategies based on reciprocity are also present in the world around us. As a great illustration of reciprocal strategy, Axelrod mentioned live and let live.
1. In the original assignment of PD, the player is a person who is accused of commiting a crime; there are indisputable proofs of him being guilty but the police also suspects that this pair worked together on “something bigger”.