Tit for Tat

principle

In repeated trust games, begin with cooperation, then copy the other party’s previous move. This makes cooperation rewarding and defection costly without turning one betrayal into permanent hostility.

The winning algorithm needs only one move of memory: if the other player cooperated last time, cooperate now; if they defected, defect. It succeeds without predicting motives or being unconditionally generous.

E1

Reciprocity with a reset button

Tit for Tat links today’s treatment directly to yesterday’s behavior. Cooperation is answered with cooperation, preserving the gains from mutual trust; defection triggers an immediate cost, so exploitation is not rewarded. Crucially, punishment lasts only as long as defection does. The moment the other party cooperates again, the strategy reciprocates, reopening the relationship instead of carrying a permanent grudge. It is Trust but Verify expressed as a decision rule: offer trust first, retain memory, and let observed behavior determine the next move.

E1

It needs another round

The strategy depends on repeated interaction. In a one-shot exchange, there is no next move through which cooperation can be rewarded or defection punished. It also offers no richer diagnosis of why someone defected; it simply mirrors the recorded move.

Make the next move conditional

In one ongoing relationship, define one observable cooperative behavior and your matching response. Offer cooperation now; on the next round, reciprocate what the other party actually did—and restore cooperation immediately if they do.

Episodes that teach this