Introduction to Game Theory
← Back
Foundations — Entry-Level Lesson
Game Theory
The mathematical study of strategic decision-making — how rational players interact when their choices affect each other, and why understanding the structure of the "game" matters more than understanding the players.
"Game theory is the study of mathematical models of strategic interactions. It has applications in economics, political science, psychology, and biology — anywhere rational agents make decisions that depend on each other."— Standard definition
What is game theory
Game theory is not about games. It is about any situation where the outcome of your decision depends on what other people decide — and their outcome depends on what you decide. A "game" in game theory is any interaction between two or more decision-makers ("players") where each player's result is affected by the choices of all players.
The key distinction from ordinary decision-making: in a game, your best move depends on what the other players do. A lumberjack deciding how to chop wood doesn't need game theory — the wood doesn't fight back. A general deciding how to attack an enemy does — the enemy is making strategic decisions in response. Whenever you face an intelligent, purposeful opponent who is also thinking about what you'll do, you are in a game.
Game theory provides a framework for thinking about these interactions systematically — identifying the players, the strategies available to each, the payoffs for each combination of strategies, and the equilibrium outcomes where no player can improve by unilaterally changing their approach.
Core vocabulary — the language of game theory
Players
The decision-makers
The individuals, companies, countries, or entities making strategic choices. Each player has their own objectives and information. Game theory assumes players are rational — they try to maximize their own payoff given what they know.
Example: Two companies deciding whether to lower prices. Each company is a "player."
Strategies
The available choices
A complete plan of action for every possible situation a player might face. A strategy isn't a single move — it's a rule for deciding what to do in every circumstance. "Always cooperate" is a strategy. "Cooperate first, then copy what the opponent did" is a different strategy.
Example: In pricing, a firm's strategy might be "match whatever price the competitor sets" or "always price 10% below."
Payoffs
The outcomes — what each player gains or loses
The reward (or punishment) each player receives for a given combination of strategies. Payoffs can be money, utility, satisfaction, survival — anything the players value. The payoff matrix shows what happens for every possible combination of all players' choices.
Example: If both firms keep prices high, both earn $10M. If one cuts prices, it earns $15M while the other earns $3M. If both cut, both earn $5M.
Dominant strategy
The move that's best no matter what others do
A strategy that gives a player a higher payoff regardless of what any other player does. If you have a dominant strategy, your decision is easy — play it. The problem: your opponent may also have a dominant strategy, and the resulting outcome may be bad for both of you (this is the Prisoner's Dilemma).
Example: In the Prisoner's Dilemma, confessing is dominant for both prisoners — but both confessing produces a worse outcome than both staying silent.
Nash Equilibrium
The stable outcome — named after John Nash (1950)
A set of strategies — one for each player — where no player can improve their payoff by unilaterally changing their strategy, given what the others are doing. It's the point where everyone is playing their best response to everyone else's best response. A Nash Equilibrium is stable but not necessarily optimal — it can be the outcome where everyone is stuck doing badly because changing alone would make things worse.
Example: Both prisoners confessing is a Nash Equilibrium — neither can improve by changing alone. But both staying silent would have been better for both.
Pareto Optimality
The outcome where no one can be made better off without making someone worse off
An outcome is Pareto optimal if there is no way to improve one player's payoff without reducing another's. The Nash Equilibrium of a game is often not Pareto optimal — which is the central tragedy of many strategic interactions. Players acting rationally in their own interest produce collective outcomes that rational collective action would avoid.
Example: Both prisoners staying silent is Pareto optimal (both get 1 year). But the Nash Equilibrium (both confess, both get 5 years) is not Pareto optimal.
Mixed Strategy
Randomizing your choices to stay unpredictable
Instead of always playing the same move ("pure strategy"), a player randomly selects from available strategies according to specific probabilities. Some games have no stable pure-strategy equilibrium — like rock-paper-scissors, where any predictable pattern is exploitable. In these cases, randomizing is the rational approach. Nash proved that every finite game has at least one equilibrium when mixed strategies are allowed.
Example: A soccer penalty kicker who always shoots left will be saved. Randomizing 60% left, 40% right (matching the goalkeeper's weaknesses) is the mixed strategy equilibrium.
Types of games — the six fundamental distinctions
Zero-sum vs. Non-zero-sum
Zero-sum: one player's gain is exactly another's loss. The total payoff is constant — the pie doesn't grow. Non-zero-sum: both players can gain (or both can lose). The pie can grow or shrink depending on choices. Most real-world interactions are non-zero-sum, which is why cooperation is possible.
Zero-sum: Chess, poker — Non-zero-sum: Trade, partnerships, most of life
Cooperative vs. Non-cooperative
Cooperative: players can make binding agreements and form coalitions. The question is how to divide the gains. Non-cooperative: players act independently — no enforceable agreements. Each player chooses their own strategy. Most game theory focuses on non-cooperative games.
Cooperative: Joint ventures, mergers — Non-cooperative: Price wars, elections
Simultaneous vs. Sequential
Simultaneous: players choose their strategies at the same time, without knowing the others' choices. Analyzed with payoff matrices. Sequential: players take turns — later players see what earlier players did. First-mover advantage (or disadvantage) is a key concept in sequential games.
Simultaneous: Sealed-bid auctions — Sequential: Chess, negotiations
Perfect vs. Imperfect Information
Perfect information: every player sees all previous moves. No hidden information. Imperfect information: players have private information others can't see. Introduces bluffing, signaling, and the strategic management of what others know about you.
Perfect: Chess, Go — Imperfect: Poker, Stratego, most business
One-shot vs. Repeated
One-shot: players interact once — no future consequences for today's behavior. Cheating is often rational. Repeated (iterated): players interact many times. Reputation, retaliation, and trust become viable strategies. This distinction transforms the Prisoner's Dilemma from a tragedy into a solvable problem.
One-shot: Buying from a stranger — Repeated: Long-term supplier relationship
Symmetric vs. Asymmetric
Symmetric: all players have the same strategies and payoffs — the game looks identical from every seat. Asymmetric: players have different roles, resources, or options. Most real-world games are asymmetric — you and your competitor don't have the same hand.
Symmetric: Prisoner's Dilemma — Asymmetric: Incumbent vs. startup
The classic games — five problems every strategist should know
1
The Prisoner's Dilemma
The most famous game in all of game theory
Two suspects are interrogated separately. Each can cooperate (stay silent) or defect (betray the other). If both cooperate: light sentences (1 year each). If one defects while the other cooperates: the defector goes free, the cooperator gets 10 years. If both defect: moderate sentences (5 years each). The dilemma: defecting is the dominant strategy for both — but mutual defection produces a worse outcome than mutual cooperation. Individual rationality leads to collective irrationality.
The lesson: In one-shot interactions, rational self-interest often produces suboptimal outcomes for everyone. This explains price wars, arms races, environmental destruction, and any situation where "if everyone acts selfishly, everyone suffers." The solution: change the game to a repeated interaction, where cooperation can evolve.
2
The Stag Hunt
The coordination problem — trust as a strategic variable
Two hunters can each hunt a stag (high value, but requires both to cooperate) or a hare (low value, but can be caught alone). If both hunt stag: big payoff for both. If one hunts stag while the other hunts hare: the stag hunter gets nothing. Both hunting hare: small but certain payoff for both. Unlike the Prisoner's Dilemma, there are two Nash Equilibria: both hunt stag (best outcome, but risky) or both hunt hare (safe, but mediocre).
The lesson: Sometimes the problem isn't conflicting interests — it's coordinating toward the better outcome when both sides benefit from cooperation but neither wants to risk cooperating alone. This is why communication, commitment devices, and trust matter: they help players coordinate on the superior equilibrium.
3
Chicken (Hawk-Dove)
The brinkmanship problem — who blinks first
Two drivers race toward each other. Each can swerve (safe but humiliating) or drive straight (brave but potentially fatal). If both swerve: tie — neither gains status. If one swerves and the other doesn't: the one who drove straight wins prestige. If both drive straight: catastrophic collision. There are two Nash Equilibria (one swerves, the other doesn't), and the game rewards the player who can credibly commit to not swerving.
The lesson: In brinkmanship situations (nuclear deterrence, labor strikes, business ultimatums), the player who can credibly eliminate their own option to back down gains the advantage. The paradox: rational flexibility is a weakness; visible irrationality is a strength. This is why leaders sometimes burn bridges, make public commitments, or tie their own hands — to make their threat credible.
4
Battle of the Sexes
The coordination problem with conflicting preferences
A couple wants to spend the evening together but disagrees on where: one prefers the opera, the other prefers a football match. Both prefer being together anywhere over being apart at their preferred venue. There are two Nash Equilibria (both go to opera, or both go to football) — but each player prefers a different one. The challenge is not cooperation itself but which cooperative outcome to coordinate on.
The lesson: Many business and personal conflicts aren't about whether to cooperate but about the terms of cooperation. Partnership negotiations, standard-setting in technology, merger terms — the parties agree that cooperation is better than competition, but each prefers a different version of the deal. The solution often involves negotiation, compromise, or accepting short-term turns.
5
The Tragedy of the Commons
The multi-player problem — collective overexploitation
Multiple farmers share a common grazing field. Each farmer benefits by adding one more cow — the grazing benefit goes entirely to them, while the cost (degradation of the field) is shared among all farmers. Individually rational: add more cows. Collectively catastrophic: the field is destroyed and everyone loses. This is a multi-player Prisoner's Dilemma — each player's dominant strategy (exploit the commons) produces collective ruin.
The lesson: Shared resources without governance are inevitably overexploited. This explains overfishing, pollution, climate change, and the depletion of any shared resource. Solutions: property rights (privatize the commons), regulation (enforce limits), or community norms (social pressure to restrain individual exploitation). Elinor Ostrom won the 2009 Nobel Prize for showing that communities can self-govern commons without privatization or top-down regulation.
Key strategies — the moves game theory has identified
01
Tit for Tat (reciprocity)
Start by cooperating. Then copy whatever the other player did last round. Nice (starts with cooperation), retaliatory (punishes cheating), forgiving (returns to cooperation), clear (easy to understand). Won Robert Axelrod's famous 1980 tournament against 62 other strategies. The mathematical proof that the Golden Rule works.
02
Minimax (minimize the maximum loss)
Choose the strategy that minimizes the worst possible outcome. Conservative and defensive — you may not win big, but you guarantee you won't lose big. The foundational strategy for zero-sum games, proven optimal by von Neumann in 1928.
03
Commitment (burning bridges)
Eliminate your own options to make your threat credible. Cortés burning his ships, a negotiator announcing "take it or leave it" publicly — removing the ability to back down makes your position stronger because the other player knows you can't retreat. Only works if the commitment is visible and irreversible.
04
Signaling (communicating through actions)
When you can't communicate directly, your actions send messages about your type, your intentions, or your capabilities. A company investing in expensive advertising signals product quality. A job candidate earning a degree signals capability. The signal must be costly enough that a faker can't afford to imitate it — this is what makes it credible.
05
Backward induction (think from the end)
In sequential games, start from the final move and work backward. At each stage, predict what the rational player would do — then use that knowledge to determine what happens at the previous stage. This unravels the entire game from the end to the beginning, revealing the optimal strategy at every point.
06
Mixed strategy (strategic randomization)
Randomize your choices according to calculated probabilities to prevent the opponent from predicting your move. Essential in any competitive situation where predictability is exploitable — penalty kicks, auditing schedules, military patrols, pricing in competitive markets.
In business
Game theory is the operating system of competitive strategy. Every pricing decision, every negotiation, every market entry is a game. The core business applications: Pricing: the Prisoner's Dilemma explains why competitors cut prices even when both would profit from keeping them high (and why cartels are unstable). Market entry: sequential game analysis reveals first-mover advantages and the credibility of incumbent threats. Negotiations: understanding Nash Equilibrium helps you identify the stable outcomes and avoid deals that will unravel. Auctions: game theory provides the mathematical foundation for optimal bidding. The deepest business lesson: don't just play the game — design the game. The player who can change the rules, the payoffs, or the structure of the interaction has more power than the player who merely optimizes within it.
In life
You play games every day — whether you know it or not. Every negotiation with a landlord, every interaction with a colleague, every decision about whether to trust someone is a game with players, strategies, and payoffs. Game theory's life lessons: Identify the game you're in. Is it zero-sum or non-zero-sum? One-shot or repeated? Cooperative or competitive? The answer determines the optimal strategy. Understand the other player's payoffs. You can't predict behavior without understanding incentives. Repeated interactions transform every game. The person you'll see again tomorrow is a fundamentally different strategic situation from the stranger you'll never encounter again — which is why trust, reputation, and relationships are the most valuable strategic assets in life. The structure matters more than the players. Bad outcomes are usually produced by bad game structures, not bad people. If you want different behavior — change the game.
Bottom line
Game theory is not abstract mathematics — it is the grammar of strategic interaction. It reveals why rational people produce irrational outcomes (the Prisoner's Dilemma), why cooperation requires structure (repeated games), why commitment beats flexibility (Chicken), and why the design of the game determines the behavior of the players more than the character of the players themselves. Learn game theory not to manipulate — but to see the invisible architecture of every interaction, and to understand that the most powerful strategic move is often not a better play within the game, but a change to the game itself.
