Game Theory Fundamentals

Read runtime and advisory rules before applying this skill. Other Foreman layers and the catalog are in ../../content/, relative to this SKILL.md.

Game Theory Fundamentals

BEAUTIFUL MIND. John Nash's Nobel Prize-winning insight fits in a single sentence: the best strategy for any player depends on what every other player is doing. This observation -- that optimal choices are interdependent, not independent -- is the foundation of game theory, and every entrepreneur already applies it without knowing the name. When a founder prices a product, negotiates a term sheet, or decides whether to enter a market, they are playing a game in the formal sense: multiple actors making strategic choices where each outcome depends on the choices of all. The difference between intuition and game theory is precision. Intuition says "consider the competition." Game theory says exactly how.

The Framework

The Three Essential Elements

Every strategic interaction -- from a pricing war between two SaaS competitors to a cofounder equity negotiation -- reduces to three elements. Identifying them is the first step toward rigorous analysis.

Element 1: Players. The decision-makers in the game. A minimum of two. In business, players are not always obvious. A startup entering a market has visible players (direct competitors) and hidden ones (regulators, platform owners, complementary product makers, the customer choosing between alternatives). Misidentifying who the players are is the most common source of strategic error.

Element 2: Strategies. The set of available actions each player can take. A strategy is a complete plan of action -- not just "lower prices" but "lower prices by 15% in Q3 and match any further competitor reduction within 48 hours." The strategies available to each player define the boundaries of the game. Expanding or constraining available strategies changes the game fundamentally.

Element 3: Payoffs. The outcome each player receives for each combination of strategies. Payoffs are not always financial. They can include market share, reputation, strategic positioning, time saved, or optionality preserved. The critical insight: a player's payoff depends not only on their own strategy but on the combination of all players' strategies.

A game is fully defined by specifying these three elements. Everything else -- equilibria, optimal strategies, competitive dynamics -- follows from them.

The Payoff Matrix: Mapping Strategic Interactions

The payoff matrix is the fundamental tool for visualizing a two-player game. Each row represents a strategy available to Player A; each column represents a strategy available to Player B. Each cell contains the payoffs for both players under that combination of strategies.

Example: Two Competing Coffee Shops

Reading this matrix reveals the strategic tension immediately. Both shops charging premium prices produces good results for both (80, 80). Both discounting produces mediocre results for both (50, 50). But each shop is individually tempted to discount while the other charges premium prices, capturing 100 while the competitor earns only 30.

This single table encodes the entire competitive dynamic. No narrative strategy document communicates as much information as concisely.

Optimal, Worst, and Relative Optimal Strategies

Three types of strategy merit identification in any game:

Optimal Strategy (Dominant Strategy). A strategy that produces a higher payoff than all alternatives, regardless of what other players do. When a dominant strategy exists, the decision is straightforward -- play it. In the Prisoner's Dilemma, confessing is the dominant strategy for both suspects. In business, a dominant strategy is rare and valuable: it means the right move does not depend on competitor intelligence.

Worst Strategy (Dominated Strategy). A strategy that always produces a lower payoff than some alternative, regardless of opponent behavior. Dominated strategies should be eliminated immediately from consideration. Their removal often simplifies the game considerably, sometimes revealing dominant strategies that were previously obscured.

Relative Optimal Strategy (Best Response). The best strategy given a specific choice by the other player. Most real business situations lack dominant strategies. Instead, the best action depends on the competitor's action. "If they price high, we should price high. If they price low, we should price low." Mapping the relative optimal strategy for each possible competitor action is the foundation for finding Nash equilibrium.

Nash Equilibrium: The Stable Outcome

Nash equilibrium is the combination of strategies where no player can improve their payoff by unilaterally changing their own strategy. Once reached, no one has an incentive to deviate.

The method for finding it in a simple matrix:

1. For each of Player B's strategies, identify Player A's best response (underline A's highest payoff in each column).
2. For each of Player A's strategies, identify Player B's best response (underline B's highest payoff in each row).
3. Any cell where both payoffs are underlined is a Nash equilibrium.

Applied to the Coffee Shop Example:

• If B prices premium, A's best response is discount (100 > 80). If B discounts, A's best response is discount (50 > 30).
• If A prices premium, B's best response is discount (100 > 80). If A discounts, B's best response is discount (50 > 30).
• Nash equilibrium: Both discount (50, 50).

The equilibrium is stable but suboptimal -- both would prefer (80, 80). This tension between individual rationality and collective benefit is the central paradox of non-cooperative games, and it recurs across virtually every competitive business situation.

Multiple Equilibria: Risk Control vs. Pareto Optimum

Many games produce multiple Nash equilibria. Two principles govern which one is likely to prevail:

Risk Control Principle. Players choose the equilibrium that minimizes their worst-case exposure. Consider two tech companies deciding whether to adopt a shared standard:

Both (Adopt, Adopt) and (Go Alone, Go Alone) are Nash equilibria. The risk-control principle predicts (Go Alone, Go Alone): if Firm A adopts the standard but B does not, A earns only 10. Playing it safe yields a guaranteed 60. Without trust or communication, risk aversion drives players to the inferior equilibrium.

Pareto Optimum Principle. When players can communicate or have established trust, they gravitate toward the equilibrium that maximizes collective payoffs. Here, (Adopt, Adopt) is Pareto optimal -- both earn 100 instead of 60.

The practical implication for entrepreneurs: the ability to communicate, build trust, and make credible commitments determines whether markets settle into productive equilibria or defensive, low-value ones.

Static vs. Dynamic Games

Static games are played simultaneously -- all players choose their strategies without knowing what others have chosen. Most pricing decisions, sealed-bid auctions, and market entry timing decisions approximate static games. The payoff matrix is the primary analytical tool.

Dynamic games are played sequentially -- one player moves first, the next player observes that move and responds, and so on. Product launches, negotiation sequences, and competitive responses are dynamic games. The primary analytical tool is the game tree, analyzed through backward induction: start at the final decision point, determine the rational choice, then work backwards to determine optimal play at each preceding stage.

The distinction matters because first-mover advantage and second-mover advantage are properties of dynamic games. A simultaneous game has no "first mover." Identifying whether a business situation is closer to a static or dynamic game determines which analytical tools apply.

Prompts

Prompt 1 -- Map Your Competitive Game:

I operate in [describe your market/industry]. Help me map the game theory structure of my competitive situation. Identify the key players (direct competitors, substitutes, adjacent entrants, platform owners, regulators), the strategies available to each, and the approximate payoffs. Present this as a payoff matrix for the two most important competitive interactions I face. Then identify the Nash equilibrium and explain whether it represents a good or bad outcome for my business.

Prompt 2 -- Find the Equilibrium in a Specific Decision:

I am facing this strategic decision: [describe decision, e.g., pricing change, market entry, partnership negotiation]. The other key player is [describe]. My available strategies are [list]. Their likely strategies are [list]. Help me construct the payoff matrix, find the Nash equilibrium, and assess whether I should play the equilibrium strategy or attempt to shift the game structure. Consider whether this is a static or dynamic game.

Prompt 3 -- Identify Dominated Strategies:

Here are the strategic options I am considering: [list all options]. For each, help me determine if any are dominated -- meaning there is always a better alternative regardless of what competitors do. Eliminate the dominated strategies and show me the simplified decision. Then identify my best response to each likely competitor action.

Prompt 4 -- Assess Multiple Equilibria:

My competitive situation appears to have multiple stable outcomes: [describe the possible equilibria]. Help me assess which equilibrium is most likely to prevail by applying the risk-control principle and the Pareto optimum principle. What would it take to shift the market from the risk-control equilibrium to the Pareto-optimal one? What commitments, signals, or structures would be required?

Use Cases

Validation-Stage SaaS Founder Choosing a Pricing Model

A B2B SaaS founder must choose between per-seat pricing and usage-based pricing. Two direct competitors face the same decision. If all three adopt per-seat pricing, customers compare on features and margins remain healthy. If one switches to usage-based while others charge per-seat, the switcher captures price-sensitive customers but compresses margins industry-wide. If all switch to usage-based, margins compress for everyone. Mapping this as a three-player game reveals that per-seat pricing is a Nash equilibrium only if no player has an incentive to defect. Analysis shows that the smallest competitor (the startup) has the strongest incentive to defect -- it gains market share by undercutting on price structure. The founder uses this insight to time the pricing model switch strategically, capturing share before competitors respond.

Growth-Stage Marketplace Evaluating Geographic Expansion

A two-sided marketplace operating in three cities must decide which of two new metros to enter next. A well-funded competitor faces the same decision. If both enter the same city, they split the market and burn cash competing for supply. If they enter different cities, each captures a local monopoly position. The game matrix reveals two Nash equilibria -- one where the startup enters City A and the competitor enters City B, and vice versa. The startup uses signaling (public announcements, early partnerships, office lease signings) to commit credibly to City A, steering the competitor toward City B and achieving the Pareto-optimal split.

Early-Traction Founder Negotiating with a Strategic Partner

A hardware startup needs a distribution partnership with one of two major retailers. Each retailer values exclusivity (higher payoff from being the sole distributor) but will accept non-exclusive terms rather than lose the deal entirely. The founder maps the negotiation as a dynamic game: she moves first by approaching both retailers simultaneously. Backward induction shows that each retailer's best response to a simultaneous approach is to offer competitive terms, because the alternative -- refusing and watching the competitor win exclusive access -- is worse. The dynamic game structure gives the founder leverage that a sequential approach would not.

Anti-Patterns

1. Treating all interactions as games against a single competitor. Most markets involve multiple players with heterogeneous strategies, asymmetric information, and different payoff structures. Reducing a five-player market to a two-player matrix can miss critical dynamics -- particularly the role of complementors, regulators, or platform owners who shape the payoff structure without competing directly.

2. Assuming players are perfectly rational. Game theory's predictions assume rational actors. Real competitors make emotional decisions, suffer from cognitive biases, and sometimes act against their own interests. The model identifies what rational actors would do, which is useful as a baseline -- but deviations from rationality are the norm, not the exception. Use the framework to identify the rational benchmark, then layer behavioral analysis on top.

3. Ignoring the distinction between one-shot and repeated games. A strategy that is optimal in a single interaction may be destructive in a repeated game. Aggressive pricing that wins one quarter's market share may trigger retaliation that destroys profitability for years. Always determine whether the game repeats before selecting a strategy.

4. Confusing Nash equilibrium with the best outcome. Nash equilibrium is the stable outcome, not necessarily the desirable one. The Coffee Shop example demonstrates this clearly: the equilibrium (both discount) is worse for both players than the cooperative outcome (both premium). Recognizing a bad equilibrium is the first step toward restructuring the game to escape it.

5. Mapping the game after the decision is made. Game theory is a planning tool, not a post-hoc rationalization device. Constructing the payoff matrix after committing to a strategy produces confirmation bias, not strategic insight. The matrix should be built before the decision, when it can actually influence the choice.

By Stage

Output Template

markdownCopy code
# Game Theory Analysis

**Market/Situation:** [Description]
**Company:** [Name]
**Date:** [Date]

## Player Map

| Player | Type | Key Strategies | Primary Objective |
|--------|------|---------------|-------------------|
| [Our Company] | Protagonist | [Strategy 1, Strategy 2] | [Objective] |
| [Competitor A] | Direct rival | [Strategy 1, Strategy 2] | [Objective] |
| [Player C] | [Platform/Regulator/etc.] | [Strategy 1, Strategy 2] | [Objective] |

## Payoff Matrix (Primary Competitive Interaction)

|  | Competitor Strategy 1 | Competitor Strategy 2 |
|--|----------------------|----------------------|
| **Our Strategy 1** | Us: [payoff], Them: [payoff] | Us: [payoff], Them: [payoff] |
| **Our Strategy 2** | Us: [payoff], Them: [payoff] | Us: [payoff], Them: [payoff] |

## Equilibrium Analysis

- **Nash Equilibrium:** [Strategy combination]
- **Is the equilibrium desirable?** [Yes/No -- explain]
- **Dominant strategies identified:** [Yes/No -- which player, which strategy]
- **Dominated strategies eliminated:** [List]

## Strategic Recommendation

Given the equilibrium analysis:
1. [Primary strategic action]
2. [Contingency if competitor deviates]
3. [Opportunity to restructure the game]

## Game Type Assessment
- **Static or Dynamic:** [Assessment and why]
- **One-shot or Repeated:** [Assessment and implications]
- **Information:** [Complete or incomplete]

Related Skills

• Prisoner's Dilemma and Cooperation -- The most famous application of these fundamentals, explaining why rational actors consistently produce collectively irrational outcomes and how to escape the trap.
• Information Asymmetry -- Extends the framework to games where players hold different information, fundamentally altering equilibrium outcomes.
• First-Mover Strategy -- Applies dynamic game analysis to the question of timing: when moving first wins, when waiting wins, and how to make commitments credible.
• Negotiation Game Theory -- Applies Nash equilibrium concepts specifically to bilateral bargaining, providing the mathematical foundation for splitting deal surplus.
• Porter's Five Forces -- Five Forces describes the structural dynamics of an industry; game theory explains the strategic logic that drives those dynamics and predicts how they evolve.
• Scenario Planning -- Combines with game theory to project how different competitive strategies might unfold under multiple future conditions.