Game Theory: A Simple Strategy That Will Change Your Life Forever¶
Summary¶
Pursuit of Wonder explains game theory through the iterated prisoner's dilemma and Robert Axelrod's 1980 tournaments, demonstrating that the simplest strategy — Tit for Tat — consistently outperforms complex and cunning strategies.
Key Takeaways¶
The Roommate Dish Problem¶
A version of the prisoner's dilemma: you and a roommate agree to split dish duty. They keep defecting (not doing dishes). You face the question: do the dishes (cooperate) or let them pile up (defect)? What precedent do you set?
Game Theory Basics¶
Game = any interaction between multiple decision-makers where outcome depends on others' choices.
| Type | Description | Examples |
|---|---|---|
| Cooperative | Shared goals, free information exchange | Sports teams, business partnerships, trade agreements |
| Non-cooperative | Independent self-interest, winners and losers | Chess, poker, Golden Balls game show |
One-Off vs. Repeated Games¶
One-off games (Golden Balls split-or-steal): The rational choice is always to steal (weakly dominant strategy).
Repeated games (real life): Interactions are never one-offs — there are lingering effects, retaliation, resource changes, relationship strain.
Axelrod's Tournament (1980)¶
Setup: 14 programs competed in iterated prisoner's dilemma (200 rounds per game, each against every other + self).
Scoring: Cooperate-Cooperate = 3 points each; Defect-Cooperate = 5/0; Defect-Defect = 1 point each.
Winner: Tit for Tat — the simplest strategy submitted.
Tit for Tat Strategy¶
- Start with cooperation — always
- Copy opponent's last move — cooperate unless they defected
- Immediately retaliate — defect back when defected against
- Immediately forgive — return to cooperating when opponent cooperates again
Results: - Never won any individual game (can only lose or draw one-on-one) - Won the tournament by cooperating with enough players to get highest total score - Won again in second tournament (62 programs, unknown round count)
Why Tit for Tat Wins¶
"Its niceness prevents it from getting into unnecessary trouble. Its retaliation discourages the other side from persisting whenever defection is tried. Its forgiveness helps restore mutual cooperation. And its clarity makes it intelligible to the other player, thereby eliciting long-term cooperation." — Axelrod
Generous Tit for Tat¶
In even more realistic, chaotic conditions, a more generous version — occasionally forgiving defections instead of reciprocating them — proved even more effective.
Nasty players found themselves in defecting wars leading to mutual destruction.
The Core Insight¶
"What makes it possible for cooperation to emerge is the fact that the players might meet again."
Life Lessons¶
- Being nice is a strength, not a weakness
- Holding grudges is a weakness; forgiveness is a strength
- Weakness is also a weakness — letting someone do wrong without consequence leads to exploitation
- Consequence must be proportional, consistent, and clear — not opaque or manipulative
- A strategy always focused on winning can be the least effective at winning overall
- "Let's be sure we do the dishes"
Limitations¶
- Simulations can never fully reproduce real-world complexity
- Real interactions involve many people, shifting goals, asymmetric resources, emotional and irrational human nature
- But the core principles hold: lead with cooperation, retaliate proportionally, forgive quickly, be clear