This strategy is for the Iterated Prisoner's Dilemma. It analyzes the opponent’s behavior over time and tries to classify their strategy (e.g. Tit-for-Tat, Grim Trigger, Pavlov, Always Defect, Random, etc.).
- Round 1: Always cooperate.
- Last round: Defect, if total rounds are known.
- Pattern detection: Matches over 25 known strategies using heuristics.
- Response: Reacts based on detected behavior (e.g., mirrors Tit-for-Tat, punishes defectors, cooperates with cooperators).
- Fallback: Defaults to defect if no clear pattern is found and opponent defected last.