Prisoner's Dilemma Payoff Matrix
|
🤝 Cooperate |
🗡️ Defect |
| 🤝 Cooperate |
3 / 3 |
0 / 5 |
| 🗡️ Defect |
5 / 0 |
1 / 1 |
Key Concepts
🎯 Nash Equilibrium – Both defect (1,1) is stable – neither can improve by changing alone
🤝 Pareto Optimal – Both cooperate (3,3) maximizes total welfare but is unstable
🔄 Shadow of the Future – Repeated play allows cooperation to emerge via reciprocity
🧬 Evolutionary Stability – Strategies compete, reproduce proportionally to fitness
Robert Axelrod's Tournament (1984)
In the landmark tournament, Tit-for-Tat – the simplest reciprocal strategy – won against 62 other programs.
Its four key properties: Nice (never defect first), Retaliatory (punish defection immediately),
Forgiving (resume cooperation after retaliation), and Clear (predictable to opponents).
Active Strategies
Start by cooperating, then copy opponent's last move. Nice, retaliatory, forgiving, and transparent – the tournament winner.
Unconditional cooperation regardless of opponent's history. Altruistic but exploitable by defectors.
Unconditional defection. Maximizes short-term gain but triggers retaliation from reciprocal strategies.
50/50 coin flip each round. Unpredictable and impossible to exploit, but equally impossible to cooperate with reliably.
If last round was rewarding (CC or DC), repeat; otherwise switch. Self-correcting and exploits unconditional cooperators.
Cooperate until opponent defects once – then defect forever. Unforgiving memory makes it devastating against serial defectors.