Founded by John von Neumann and Oskar Morgenstern (Theory of Games and Economic Behavior, 1944), with John F. Nash Jr.'s equilibrium (Equilibrium Points in N-Person Games, 1950) and Thomas C. Schelling's The Strategy of Conflict (1960).
Lays out a strategic situation as players, the moves open to each, and their payoffs, then finds the stable outcomes and asks whether the game itself could be changed.
An entry game solved from the last move back, and changed by a commitment
Each end shows both players' yearly profit. The incumbent picks its best reply first; the startup then picks knowing that reply. Raise the incumbent's profit from fighting to 6, as a binding price-match promise to its customers would, and watch the startup's best move change. Illustrative.
Payoffs read Startup / Incumbent.
Enter, then Fight on price
Enter, then Share the market
Stay out
An example
A startup selling scheduling software to dental practices was deciding whether to enter a region where one incumbent served 80 percent of practices. The incumbent's sales head had promised any newcomer a price war. The founders took the threat at face value until their advisor drew the game.
Two players, two moves each. The startup could enter or stay out; if it entered, the incumbent could fight on price or share the market. The advisor put rough yearly profits on each end. A price war would cost the startup $2 million and cut the incumbent's profit from $9 million to $3 million; sharing would leave the incumbent $5 million. Working back from the last move, the incumbent's best reply to entry was to share, so the threat was not credible and entry was worth about $3 million a year.
The advisor then asked what would make the threat real: a binding price-match promise to customers, which would make fighting automatic. The incumbent had made none. The startup entered with the smaller practices and watched the incumbent's contracts for that clause.
Why it works
The tool makes every party's choice depend on the others' in plain view. Most strategy talk plans one move and assumes the other side stands still, or takes its statements at face value. Laying out players, moves, and payoffs from each player's own seat replaces both habits with a question that has an answer: given what the others will do, what is each player's best move?
Working back from the last move is what exposes empty threats. A threat matters only if carrying it out would pay the threatener when the moment came. The second insight is that the game is not fixed. Payoffs, moves, timing, and information can be changed by contracts, commitments, and sequencing, and changing the game often matters more than playing it well.
The thinking behind it
John von Neumann John von Neumann and Oskar Morgenstern founded game theory as a field with Theory of Games and Economic Behavior (1944). Its subject is the game of strategy, in which each player's best choice depends on what the others choose, and the same book set out expected-utility theory for rational choice under risk. The entry game in the figure is a small game of strategy: the startup's best opening move depends on the incumbent's best reply, so the reply is worked out first.
John F. Nash Jr. John F. Nash Jr. defined, in Equilibrium Points in N-Person Games (1950), the equilibrium that bears his name: a set of strategies, one per player, in which no player can do better by changing strategy alone. He proved that every finite game has at least one such equilibrium, in mixed strategies if not in pure ones. It is the test of a stable outcome used in step five: an outcome is stable when nobody gains by moving alone.
Thomas C. Schelling Thomas C. Schelling, who shared the 2005 Nobel Memorial Prize in Economic Sciences for work on conflict and cooperation, introduced focal points and credible commitment in The Strategy of Conflict (1960). A player can gain by binding itself in advance, because the commitment changes what the others expect it to do. That is the move the figure lets the reader make: raise the incumbent's payoff from fighting, as a binding promise would, and the startup's best move changes with it.
How to run it
- Name the players. Specific firms or people, not the market or the regulator, and anyone who could join.
- List the moves each can really make. Include staying out, waiting, and exiting, the moves most often left off.
- Put payoffs on every combination. Rough numbers or a ranking will do, so long as each is written from that player's seat.
- Settle the shape of the game. Sequential or simultaneous, played once or repeated; repetition can sustain cooperation a single round cannot.
- Find the stable outcome. In a sequential game, work back from the last move; in a simultaneous one, find where nobody gains by switching alone.
- Test every threat and promise. If carrying it out would not pay the one who made it, set it aside.
- Ask what would change the game. Commitments, contracts, timing, information, and who takes part can each move the outcome.
Where it goes wrong
The game is the wrong game. A third party, a coalition, or the option to wait is missing, and the stable outcome is stable only on paper. Spot it when the real players do something the model has no branch for, and add the move.
The payoffs come from the modeler's own seat. Rivals are given the modeler's preferences, costs, or fears. Write each player's payoffs from what that player has done before, and ask someone close to them to check.
A repeated relationship gets modeled as one round. Pride, reputation, and long memory change what players do when they will meet again. Use the model to frame the choice, and check its prediction against the history between the players.
When to use something else
- Nash Equilibrium. When the players choose at the same time, and the question is which combinations nobody would leave on their own.
- Backward Induction. When the game runs through many moves in a fixed order and each needs solving from the end.
- Repeated Games. When the same players meet again and again, and cooperation depends on the future.
Where it sits in SOLVE eX
Go to the source
- Theory of Games and Economic Behavior, 1944
- Equilibrium Points in N-Person Games, Proceedings of the National Academy of Sciences, 1950
- The Strategy of Conflict, 1960