A direct answer
The Prisoner’s Dilemma does not prove that people are selfish or that cooperation is irrational. It shows that, under a specific ordering of payoffs and independent choice, individually dominant strategies can produce an outcome that is worse for every participant than mutual cooperation.
Questions hidden inside
- Can a choice be rational for each person while the resulting outcome is bad for everyone?
- When do trust, communication, repetition, or enforceable rules make cooperation sustainable?
- Does choosing to cooperate express morality, strategy, loyalty, or some combination of them?
- Which real situations genuinely share the dilemma’s payoff structure?
Philosophical lenses
Rational choice
What should each participant do given the other person’s possible choices?
The standard one-shot structure gives each participant an individually dominant strategy even though the combination produces a jointly inferior result.
Ethics
Should moral reasons change how the participants rank their options?
Promises, loyalty, fairness, and concern for others may supply reasons that the stripped-down payoff model does not include.
Institutions
How can rules make cooperation credible and resilient?
Monitoring, repeated interaction, enforceable commitments, and fair procedures can change incentives rather than relying on trust alone.
Begin with a choice
You and another person each receive a greater personal reward for defecting, whatever the other chooses, but mutual cooperation rewards both more than mutual defection. What do you choose in a one-time anonymous round?
The separated-choice scenario
Two suspects are questioned separately. Each can cooperate with the other by remaining silent or defect by giving evidence. If both remain silent, each receives a light penalty. If one defects while the other remains silent, the defector goes free and the silent person receives the harshest penalty. If both defect, each receives a substantial penalty.
Assume each cares only about receiving the least severe penalty and cannot make a binding agreement. For either suspect, defecting produces a better personal result whether the other cooperates or defects. Yet when both follow that reasoning, both do worse than they would have done through mutual cooperation.
| Your choice | Other person cooperates | Other person defects |
|---|---|---|
| Cooperate | Both receive the second-best outcome | You receive the worst outcome |
| Defect | You receive the best outcome | Both receive the second-worst outcome |
The structure creates the dilemma
The exact number of years is not essential. What matters is the ranking of outcomes: exploiting a cooperator is best, mutual cooperation is next, mutual defection is worse, and being the sole cooperator is worst. In the standard one-shot game, defection is a dominant strategy because it ranks above cooperation for each possible choice by the other person.
Mutual defection is also a Nash equilibrium: once there, neither participant can improve by changing alone. That does not make it the best joint outcome. The case separates stability from collective benefit and exposes a coordination problem that goodwill by one person cannot solve on its own.
The formal labels cooperate and defect describe moves in a model. They do not automatically identify which real-world action is morally right.
Repetition changes what strategy can achieve
When people expect to meet again, a present choice can affect later treatment. Conditional cooperation, reputation, retaliation, forgiveness, and the possibility of rebuilding trust can make cooperative strategies more successful than they are in an isolated encounter. The result depends on details such as how likely another round is, whether mistakes occur, and what each participant can observe.
Institutions can also alter the game. Contracts, monitoring, shared norms, transparent records, insurance, or a trusted mediator may change the payoffs or make commitments credible. Rather than asking people to become less rational, a community can redesign incentives so that individually sensible action and mutual benefit point in the same direction.
- A known final round can differ from an indefinitely continuing relationship.
- Communication may reveal intentions without creating an enforceable commitment.
- Punishment can discourage exploitation but may also sustain costly cycles after mistakes.
- Unequal power and unequal payoffs can make a supposedly symmetric example misleading.
Use the model carefully
The Prisoner’s Dilemma can illuminate contributions to public goods, arms competition, environmental restraint, price competition, teamwork, and the sharing of information. It is most useful when each participant is tempted to shift a cost to others while benefiting from their restraint.
Not every conflict is a Prisoner’s Dilemma. Some situations are coordination problems in which people mainly need to align, while others involve incompatible interests, uncertain information, more than two strategies, or moral duties that should not be represented as personal payoffs. A responsible application states who the players are, which choices they control, how they rank the outcomes, and whether those assumptions are justified.
Change one fact
Does the prospect of an ongoing relationship change your first move?
Your earlier choice: Choose in a single anonymous round with no communication and no future consequences.
Continue with cooperation, rationality, trust, collective action.
Continue with Ada
Professor Ada Rowan
Test an idea or objection
Compare perspectives, test an assumption, or develop an objection using this reviewed page as context.
Related paths
Sources
- Prisoner’s Dilemma — Steven Kuhn (2025)
Current scholarly treatment of the Prisoner’s Dilemma, its payoff structure, history, one-shot and repeated forms, evolutionary variants, and philosophical applications. It supports a careful distinction between a formal strategic model and any ordinary situation loosely described as a dilemma.
View source - The Evolution of Cooperation — Robert Axelrod (2025)
Axelrod’s study of repeated Prisoner’s Dilemma tournaments and the conditions under which cooperation can emerge among strategically interdependent agents. It supports discussion of iteration, reciprocity, retaliation, forgiveness, and the limits of transferring tournament results directly to moral life.
View source
Last modified 2026-07-31. Reviewed by Thinking Paths editorial team.