The result depends on the other side’s response
In an ordinary scenario table, demand and price may look like external conditions. In a strategic problem, a competitor responds to the company’s decision and the company responds to the expected reaction. Game theory helps describe participants, available actions, information and payoffs. It does not create factual evidence about management intentions. If the payoffs are invented, equilibrium remains the result of an invented model even when calculated rigorously.
MIT teaching materials separately address strategic form, equilibrium, mixed strategies, repetition and information. That structure is useful for asking questions about a business. One cannot take the name of a famous game and declare that every industry must reproduce its outcome. First show that the actions, benefits and constraints correspond to the specific mechanism being analyzed.[1]
A matrix for two hypothetical producers
Imagine producers A and B. Each can keep the ordinary price or cut it. In a fictional one-shot problem, both earn 10 at ordinary prices. If A cuts while B holds, A earns 14 and B earns 4; with the reverse choice, the payoffs switch. If both cut, each earns 6. The figures are hypothetical and are not data from a real industry.
For A, cutting is better under either action by B: 14 exceeds 10 and 6 exceeds 4. The same is true for B. Thus, in this specific one-shot model, both choose a cut even though the joint result of 6 each is worse than the joint result of 10 each. Total profit is 12 in the first outcome and 20 in the second. This illustrates a conflict between individual incentives and the joint result; it does not recommend coordinating prices.
Theoretical analysis is not a legal assessment. Actual price coordination may breach applicable competition rules; this article does not describe how to engage in it. For a financial reader, the relevant point is different: a favorable aggregate outcome need not survive the independent actions of individual participants. A margin forecast should account for the competitor’s incentive to respond.
Change an assumption and recalculate
Suppose a price cut requires extra capacity and yields 9 rather than 14 when the competitor holds price. The original logic of a dominant action changes. This shows why the initial payoffs matter. They cannot be chosen to fit an appealing story and then presented as independent confirmation of that same story.
In real analysis, ask who can expand output, how quickly supply can change, whether financing limits exist and how costs differ. If actions are sequential rather than simultaneous, use a sequential structure: the first move changes the set of available responses. If a participant cannot observe the other’s costs, information and beliefs must be modeled separately. A complete-information matrix does not describe that problem automatically.
Repeated relationships are not one meeting
With repeated interaction, today’s choice may affect future responses. The duration of the relationship, observability of actions and value of future outcomes matter. Repetition does not guarantee cooperation and certainly does not determine whether a particular practice is lawful. An uncertain horizon, a new entrant or a technological change can alter incentives and destroy previously observed stability.
Do not confuse equilibrium with the best outcome for everyone. Equilibrium describes stability against permitted unilateral deviations under stated assumptions. It may not be unique and need not maximize the aggregate result. Therefore, “equilibrium” is not a synonym for fairness, business efficiency or an exact forecast of a future price.
Use the model when reading financial statements
List real decisions: price, capacity, quality and product-launch timing. For each, identify a possible competitor response and observable signs. Separate facts, estimates and assumptions about reactions. Then check how those assumptions enter [peer comparison](/guides/peer-comparison/) and cash-flow forecasts. A valuation point should have an explainable strategic premise, not merely extend the previous growth rate.
Questions about interaction complement the [analysis of complex financial systems](/guides/complex-systems-markets/). The useful result of game theory is not a secret way to guess a competitor but a transparent map of dependencies. It shows which change in information, costs or available actions could reverse the conclusion. If the model cannot be linked to testable observations, leave it as an educational analogy rather than presenting it as a confirmed company valuation.
Checking the stability of one outcome
For each participant in turn, hold the other’s action fixed and compare all of that participant’s alternatives. If any unilateral change improves the payoff, the candidate fails the selected equilibrium condition. This is a small computable check, not an assessment of character. The [Nash equilibrium](/glossary/ravnovesie-nesha/) definition helps distinguish such stability from maximum total profit.
Sources and review boundaries
[1] [Lecture Notes: Game Theory for Managers](https://ocw.mit.edu/courses/15-040-game-theory-for-managers-spring-2004/pages/lecture-notes/). Equilibrium, repeated games, information and incentives; not a description of a specific company.