Minimax theorem

The minimax theorem says that in a zero-sum game, each player can choose a strategy that minimizes their maximum possible loss. In Intermediate Microeconomic Theory, it explains how rational players make decisions when outcomes depend on an opponent's move.

What is the minimax theorem?

The minimax theorem is a game theory result for zero-sum games, where one player's gain is exactly the other player's loss. In that setting, it says there is a stable strategy profile where each player is choosing the best response to the worst case they might face.

In practical terms, minimax means you do not pick the option with the highest possible payoff. You pick the option that gives you the best outcome among your worst possible outcomes. That is why it is called a "maximize the minimum" strategy, or equivalently, a "minimize the maximum loss" strategy.

This shows up most clearly when players do not have a pure strategy that works well every time. If your opponent can observe or anticipate your move, then a single fixed action may be easy to exploit. The theorem tells you that randomizing across actions with a mixed strategy can create an equilibrium where neither side can do better by changing alone.

In Intermediate Microeconomic Theory, this idea sits inside the static games unit. You may see it in payoff matrices, where each cell shows outcomes for both players, or in problems where you are asked to find a saddle point, a mixed strategy equilibrium, or the value of the game. The theorem is what justifies treating those solutions as economically meaningful rather than just algebraic tricks.

A simple way to think about it is this: if you are choosing under strategic uncertainty, minimax asks, "What is the safest strategy if my opponent behaves optimally against me?" That does not mean you are timid. It means you are rational in a competitive setting where the other side is trying to hurt your outcome as much as possible.

The theorem is closely tied to zero-sum logic, but it does not mean every game has a pure strategy solution. Many games only have a minimax solution in mixed strategies, where probabilities over actions do the work of making you unpredictable.

Why the minimax theorem matters in Intermediate Microeconomic Theory

The minimax theorem matters because it gives you a clean way to analyze competition when one person's best move depends on another person's response. In Intermediate Microeconomic Theory, that is the basic structure of static games, especially when incentives are directly opposed.

It also connects the math of game theory to economic behavior. Instead of just saying, "each player chooses something," you can explain why a player might randomize, accept a lower expected payoff on some moves, and still be acting optimally overall. That is a big shift from simple utility maximization, where there is no opponent actively trying to exploit your choice.

You will also see it as a bridge to Nash equilibrium. The minimax result helps show why mixed strategy equilibria can exist in zero-sum games and why the equilibrium value is stable. If you can identify the worst-case payoff in a payoff matrix, you are already doing the kind of reasoning the theorem formalizes.

It shows up in class when you analyze payoffs, compare pure and mixed strategies, or explain why a player would avoid being predictable. It also gives you a vocabulary for describing strategic uncertainty in markets, bargaining, and competition, even when the exact model is stylized.

How the minimax theorem connects across the course

Zero-sum game

The minimax theorem is built for zero-sum games, where one player’s gain is the other player’s loss. That structure is what makes the "best worst-case" logic work. If the game is not zero-sum, you usually need different equilibrium tools because both players can sometimes benefit or lose together.

Mixed strategy

Minimax often leads you to mixed strategies instead of a single fixed action. By randomizing, a player can make their behavior harder to exploit and improve the worst-case outcome. In many textbook examples, the minimax solution is only available once players choose probabilities over actions.

Nash equilibrium

A minimax solution in a zero-sum game lines up with a Nash equilibrium idea, because neither player can improve by changing alone. The difference is that minimax highlights the worst-case logic, while Nash equilibrium is the broader equilibrium concept used across many game types. In zero-sum games, they fit together neatly.

Payoff Matrix

Payoff matrices are the most common place you will calculate minimax in an intermediate micro class. You compare the payoffs across rows or columns, look for the best worst-case choice, and check whether a pure strategy saddle point exists. If not, the matrix often pushes you toward mixed strategies.

Is the minimax theorem on the Intermediate Microeconomic Theory exam?

A quiz question or problem set will usually give you a payoff matrix and ask you to identify the minimax strategy, the maximin strategy, or the value of the game. You may need to compare row minima and column maxima, check for a saddle point, or solve for mixed strategy probabilities.

If the question is conceptual, you should explain that minimax is the strategy that protects a player against the worst possible opponent response in a zero-sum setting. If the game has no pure strategy equilibrium, the right move is often to show why mixing is necessary rather than guessing a single action.

For written responses, use the payoff table directly. Point to the worst-case payoff for each choice, then explain why the rational player picks the option with the least bad outcome. That kind of step-by-step reasoning is exactly what instructors look for in static game analysis.

The minimax theorem vs Nash equilibrium

These overlap in zero-sum games, but they are not the same idea. Minimax is a rule for choosing the strategy that protects you against the worst case, while Nash equilibrium is the broader condition that no player wants to change alone. In a zero-sum game, a minimax solution usually gives a Nash equilibrium, but Nash equilibrium also applies to many non-zero-sum games where minimax does not.

Key things to remember about the minimax theorem

• The minimax theorem applies to zero-sum games, where each player's loss is the other player's gain.

• It tells you to choose the strategy with the best worst-case payoff, not the highest possible payoff.

• In many games, the minimax solution appears only when players use mixed strategies.

• The theorem connects directly to Nash equilibrium in zero-sum settings.

• In Intermediate Microeconomic Theory, you usually use it to analyze payoff matrices and strategic uncertainty.

Frequently asked questions about the minimax theorem

What is the minimax theorem in Intermediate Microeconomic Theory?

It is the result that in a zero-sum game, each player can choose a strategy that minimizes their maximum possible loss. In microeconomics, this is used to analyze strategic interactions where your payoff depends on an opponent's move. It is especially common in static game problems with payoff matrices.

How is minimax different from maximin?

Minimax focuses on minimizing the maximum loss, while maximin focuses on maximizing the minimum gain. In many zero-sum game setups, the two ideas are mirror images from each player's point of view. If you are reading a payoff matrix, maximin is usually the row player's perspective and minimax is the column player's perspective.

Does the minimax theorem always give a pure strategy?

No. A lot of games do not have a pure strategy solution that satisfies minimax. In those cases, the equilibrium comes from a mixed strategy, where you choose probabilities over actions to make yourself harder to exploit.

How do you find minimax in a payoff matrix?

You usually look at each possible strategy and find its worst-case payoff, then choose the strategy with the best of those worst-case outcomes. If the game has a saddle point, that can be the minimax solution in pure strategies. If not, you often need mixed strategy methods or algebra to solve it.