Open this lesson in your favourite AI. It'll walk you through the why, explain the demo, and quiz you on the try-it list.
Zero-sum games are the cinema of game theory — chess, poker, sports — and they are the exception, not the rule. Most strategic interactions you'll encounter are non-zero-sum: both sides can gain, both can lose, or one can gain while the other is indifferent. Mistaking a non-zero-sum interaction (a team, a merger, a negotiation) for a zero-sum one is one of the classic ways smart people sabotage themselves.
A game is zero-sum if, for every outcome, payoffs sum to zero (or any constant — 'constant-sum' generalises cleanly). In 2 players this reduces to .
Use these three in order. Each builds on the one before.
Explain in one paragraph what 'zero-sum' means and why it's a specific (not typical) kind of game.
Walk me through why two-player zero-sum games have especially clean solutions (minimax) while general-sum games don't. What breaks down?
Economists say international trade is 'positive-sum'. Using the game-theoretic definition, what exactly does that mean, and how would you test the claim against a real trade relationship?