The prisoner’s dilemma
Sometimes the safest choice for each person creates a worse result for everyone. The dilemma shows why—and how better rules can make cooperation practical.
You and your flatmate each decide whether to clean tonight. If both clean, the job is quick and the kitchen is pleasant. If only one cleans, the other enjoys a clean room without lifting a finger. If neither cleans, both avoid the work—but both wake up to a mess.
The temptation is simple: leaving the job gives you the easier evening whether your flatmate cleans or not. Now test that logic yourself.
Choose before you know what they will do.
Points are a simplified blend of free time and having a usable kitchen. More points means a better evening for that person.
Your turn tonight
Your flatmate is behind a closed door. You cannot coordinate first.
What do you choose?
Every possible ending
your points · their pointsThem →
clean
leave it
Choose your action on the left, then inspect either response from your flatmate.
“Leave it” wins each comparison—yet loses the bigger one.
Compare your own points while holding the other person’s choice still.
If your flatmate cleans…
You get more by leaving the work to them.
If your flatmate leaves it…
You still get more by avoiding the work.
Both people follow the same self-protective logic and end at 1 · 1, even though 3 · 3 was available.
Use it to spot situations where trust, not goodwill, is the bottleneck.
The names change, but the structure repeats: cooperation costs you something now; taking the easy option is tempting; and if everyone takes it, everyone is worse off.
A group assignment
Each member can prepare properly or hope the others carry the presentation. Slacking saves time—unless everyone has the same idea.
What knowing the dilemma helps you do
Do not merely ask people to “be nicer.” Change the conditions that make cooperation feel risky.
Use named sections and short progress check-ins, so effort is observable before the deadline.
Share a rough outline early. People can reciprocate before anyone has risked the whole grade.
A clear backup plan makes accountability predictable rather than personal.
Cooperation grows when people can safely expect it back.
Turn on the conditions below. They do not guarantee trust; they reduce the risk of being the only person who cooperates.
Good intentions must carry all the weight.
With no feedback, visibility, or commitment, taking the cooperative risk is hardest to justify.
When cooperation fails, look beyond the people. The structure of the choice may be doing the damage.
Knowing the prisoner’s dilemma is practical because it changes your next question from “Why are they selfish?” to “What would make cooperation safe, visible, and worth repeating?”