Chapter 23: Game Theory
Core idea
Most economic decisions are made in isolation: a consumer optimizes against fixed prices, a small farmer plants without worrying about what the farm next door will do. Oligopoly is different. With only a handful of players, every major decision depends on what the others will choose — and they’re thinking the same thing about you. Game theory is the toolkit for reasoning about exactly this kind of interdependent choice. Its most famous result, the Prisoner’s Dilemma, shows why two parties who would each be better off cooperating will often each rationally choose to defect — and why repeating the game changes the answer.
Authors’ framing: Game theory studies the outcomes of decisions made when those decisions depend on the choices of others. It’s the natural language of oligopoly.
Why it matters
If you only ever learn one piece of game theory, learn the Prisoner’s Dilemma. It explains cartels, arms races, climate negotiations, ad spending, and why your gym membership autorenews.
Why “rational” doesn’t mean “optimal”
The Prisoner’s Dilemma’s punchline is unsettling: two perfectly rational, self-interested players can each follow their own best logic and arrive at a worse outcome than if they had just cooperated. This isn’t a quirk — it’s the structural problem at the heart of public goods, climate, defense spending, fisheries, and antitrust. Most “tragedy of the commons” stories are dressed-up Prisoner’s Dilemmas.
Why repeated games change everything
A one-shot Prisoner’s Dilemma rewards defection. A repeated one — where you’ll face the same opponent again tomorrow — rewards conditional cooperation. This is the entire reason oligopolists who legally can’t collude still end up with prices much higher than perfect competition would predict. They don’t sign agreements. They play tit-for-tat across thousands of pricing rounds and learn that punishing each other isn’t worth it.
A frame that travels
The same logic explains: arms races (both sides would rather disarm but neither dares go first), labor strikes (cooperation reduces strike costs but the temptation to free-ride is real), pricing wars, attack ads in politics, doping in sports, and overfishing in international waters. Once you have the frame, you’ll see it everywhere — and you’ll notice that the solution is almost always “make the game repeat enough times that defection stops paying.”
Key takeaways
Key takeaways
- Game theory analyzes decisions where each player's payoff depends on what the others choose — it's the math of interdependent strategy.
- A dominant strategy is one a player should choose regardless of what the opponent does. When both players have a dominant strategy, the outcome is locked in.
- The Prisoner's Dilemma: two parties acting rationally in their own interest reach a jointly worse outcome than mutual cooperation would have produced.
- One-shot games tend to lock in defection. Repeated games allow players to develop tit-for-tat strategies and reach a cooperative equilibrium.
- Tacit collusion = repeated-game cooperation without an explicit agreement. It explains why oligopolist prices stay high without anyone signing a contract.
- Successful tit-for-tat requires: ability to observe the other's moves, ability to retaliate quickly, and an indefinite (or unknown) end date.
- When the game has a known last round, cooperation unravels by backward induction — the last round is one-shot, so defect; therefore the second-to-last is effectively one-shot too; and so on.
Mental model — the Prisoner’s Dilemma payoff matrix
Read it as: Read each cell as “what happens if Adam picks the row and Karl picks the column.” The cooperative cell (green) gives each only 1 year — but neither can trust the other to choose it. Whatever Karl does, Adam is better off confessing (compare row 1 vs row 2 in each column). The same logic drives Karl. Both rationally pick the bottom-right (red) outcome — 3 years each, worse than the green cell they could have had.
Mental model — gas station tit-for-tat
Practical application
Diagnose any standoff as a Prisoner’s Dilemma
Build cooperation deliberately
-
Increase the shadow of the future. Long-term contracts, repeat business, public reputation systems — anything that makes the next interaction matter — raises the cost of defection.
-
Make moves visible. If players can see whether others are cooperating in near-real-time, retaliation becomes credible and quick.
-
Start by cooperating. Tit-for-tat’s secret is that it’s nice — it cooperates on round 1 — but punishes any defection immediately and forgives the moment the other side resumes cooperation.
-
Don’t be vindictive. Permanent retaliation locks in a bad equilibrium. Forgive once the other side returns to cooperation.
-
Change the payoffs if you can. Sometimes the structure is so adversarial that no level of repetition fixes it; you need to add an external cost to defection (a regulator, a contract penalty, a tax).
Example: gym memberships as a PD
You and your friend both want to get in shape. You’d both be best off if you each go three times a week (jointly: motivation, accountability, payoff). But on any given Tuesday, the rational move is to skip — the gym is far, you’re tired, missing one session won’t hurt.
If neither of you can see what the other did, you’re playing one-shot every Tuesday. Defection (skip) dominates. Within a month, neither of you is going.
Now add observability: a shared app that tells each of you when the other has checked in. Suddenly the game repeats with memory. If you skip, your friend skips tomorrow. The cost of defection now includes losing your training partner. Tit-for-tat kicks in. Attendance stays high — not because you’ve become more disciplined, but because the structure of the game changed. This is the entire premise behind workout buddies, accountability groups, and apps like Stickk. Game theory in your sneakers.
Caveats
Related lessons
Jump to…
Type to filter; press Enter to open