gametheory-framework-nash-and-classic-games
Use when setting up or reading a game: what defines a game and what 'rational' actually assumes, normal and extensive form representations, dominance and iterated elimination, Nash equilibrium including existence, mixed strategies and — importantly — what Nash equilibrium is not, and the classic games from the Prisoner's Dilemma through the other four worth knowing. Includes the router for the whole game-theory reference.
- License 许可
- MIT
- 最近更新
- 2026-11-15
- Stars
- ★ 5
- Forks
- 3
01 前言 Front Matter
- name
- gametheory-framework-nash-and-classic-games
- description
- Use when setting up or reading a game: what defines a game and what 'rational' actually assumes, normal and extensive form representations, dominance and iterated elimination, Nash equilibrium including existence, mixed strategies and — importantly — what Nash equilibrium is not, and the classic games from the Prisoner's Dilemma through the other four worth knowing. Includes the router for the whole game-theory reference.
03 SKILL.md 正文
Game Theory: The Framework, Representations, Dominance, Nash Equilibrium, and the Classic Games
Part 1 of 5 of the Game Theory reference (plugin
game-theory), covering §0–§5. Sibling skills:gametheory-zero-sum-sequential-repeated-and-information(§6–§9),gametheory-bargaining-cooperative-mechanism-design-and-matching(§10–§13),gametheory-evolutionary-empirical-limits-and-computation(§14–§16),gametheory-reference(§17–§21). Section numbers are shared across the set; a reference written as §N →skillpoints into that sibling skill.Currency: Settled mathematics — von Neumann and Morgenstern 1944, Nash 1950, Gale-Shapley 1962, Selten 1965, Harsanyi 1967, Maynard Smith 1973, Myerson-Satterthwaite 1983.
Scope. The mathematics of strategic interaction — situations where your best action depends on what others do, and theirs on what you do. ⚠️ That circularity is the whole subject; everything else is machinery for resolving it.
⚠️ GOTCHA boxes mark misconceptions, and game theory attracts a lot of them — §18 →
gametheory-referenceconsolidates.The three ideas that matter most:
- ⚠️ "Rational" means consistent, not selfish. A player maximizes their own utility function — and that function can include other people's welfare, fairness, or spite. Most criticism of game theory as "assuming people are selfish" attacks something the theory doesn't claim (§1.2).
- ⚠️ Equilibrium is a consistency condition, not a prediction and not a recommendation. Nash equilibrium says no one can gain by unilateral deviation. It does not say the outcome is good, or that players will find it (§4.3).
- ⚠️ The interesting results are mostly impossibility theorems. Arrow, Gibbard-Satterthwaite, Myerson-Satterthwaite, the folk theorem's anything-goes conclusion. Game theory's deepest contribution is telling you what cannot be achieved (§12 →
gametheory-bargaining-cooperative-mechanism-design-and-matching).
§0. Routing
| You want... | Go to |
|---|---|
| The framework and its assumptions | §1 |
| Representations | §2 |
| Dominance | §3 |
| Nash equilibrium | §4 |
| The classic games | §5 |
| Zero-sum and minimax | §6 → gametheory-zero-sum-sequential-repeated-and-information |
| Sequential games | §7 → gametheory-zero-sum-sequential-repeated-and-information |
| Repeated games | §8 → gametheory-zero-sum-sequential-repeated-and-information |
| Incomplete information and signalling | §9 → gametheory-zero-sum-sequential-repeated-and-information |
| Bargaining | §10 → gametheory-bargaining-cooperative-mechanism-design-and-matching |
| Cooperative game theory | §11 → gametheory-bargaining-cooperative-mechanism-design-and-matching |
| Mechanism design and auctions | §12 → gametheory-bargaining-cooperative-mechanism-design-and-matching |
| Matching markets | §13 → gametheory-bargaining-cooperative-mechanism-design-and-matching |
| Evolutionary game theory | §14 → gametheory-evolutionary-empirical-limits-and-computation |
| Where the theory fails empirically | §15 → gametheory-evolutionary-empirical-limits-and-computation |
| Computational game theory | §16 → gametheory-evolutionary-empirical-limits-and-computation |
| Misconceptions | §17 → gametheory-reference |
| Formulas | §18 → gametheory-reference |
| Books | §19 → gametheory-reference |
| Quick reference | §20 → gametheory-reference |
§1. The Framework
1.1 What defines a game
Players who decides
Strategies ⚠️ a COMPLETE contingent plan — what to do at every point you might act
Payoffs utility over outcomes
Information who knows what, when
Timing simultaneous or sequential
⚠️ "Strategy" is a technical term and it trips people up: it is not a single move. In chess, a strategy specifies your response to every possible position — an astronomically large object. This matters because equilibrium is defined over strategies, not moves.
Common knowledge — ⚠️ everyone knows X, everyone knows everyone knows X, ad
infinitum. This is far stronger than "everyone knows" and it does real work in the
theory (§18 → gametheory-reference's electronic mail game).
1.2 ⚠️ What "rational" actually assumes
A rational player has a complete, transitive preference ordering and maximizes expected utility with respect to it (von Neumann-Morgenstern, 1944, from four axioms: completeness, transitivity, continuity, independence).
⚠️ GOTCHA — this is the most misunderstood point in the subject. ⚠️ Rationality does NOT mean selfish, money-maximizing, cold, or unemotional. A player whose utility function values their child's welfare above their own is rational. A player who derives utility from punishing unfairness is rational. The assumption is consistency, not content.
⚠️ The real limitation is different and worth stating precisely: the theory assumes unlimited computational ability, correct beliefs about others' payoffs, and common knowledge of rationality. Those are the assumptions that actually fail (§15 →
gametheory-evolutionary-empirical-limits-and-computation, §16 →gametheory-evolutionary-empirical-limits-and-computation).
⚠️ Utility is ordinal in preference and cardinal only up to positive affine
transformation. Interpersonal comparison of utility is not licensed by the
framework — which is why §11 → gametheory-bargaining-cooperative-mechanism-design-and-matching and §12 → gametheory-bargaining-cooperative-mechanism-design-and-matching have to work hard to say anything about fairness.
§2. Representations
Normal (strategic) form — a payoff matrix; assumes simultaneous choice. Extensive form — a game tree with nodes, branches, and information sets (⚠️ a set of nodes a player cannot distinguish between — this is how you represent imperfect information, and a simultaneous game is just a tree with a big information set).
⚠️ Perfect vs complete information — routinely confused:
- Perfect information: every player knows the full history when they move. ⚠️ Chess has it; poker does not.
- Complete information: everyone knows everyone's payoff functions. ⚠️ An auction
with private valuations lacks it.
They are independent. §7 →
gametheory-zero-sum-sequential-repeated-and-informationhandles imperfect; §9 →gametheory-zero-sum-sequential-repeated-and-informationhandles incomplete.
§3. Dominance
Strictly dominated — worse than another strategy against every opponent profile. ⚠️ A rational player never plays one, so you can delete it — and iterated elimination of strictly dominated strategies is order-independent. Weakly dominated — never better, sometimes worse. ⚠️ Iterated elimination of weakly dominated strategies IS order-dependent, and can eliminate legitimate equilibria. Handle with care.
Dominant strategy equilibrium — everyone has a single best strategy regardless of
others. ⚠️ Rare and extremely strong when it exists — it needs no beliefs about others
at all, which is exactly why §12 → gametheory-bargaining-cooperative-mechanism-design-and-matching prizes dominant-strategy mechanisms.
Rationalizability — ⚠️ the weaker solution concept: strategies surviving iterated elimination of never-best-responses. Every Nash equilibrium is rationalizable; the converse fails.
§4. Nash Equilibrium
4.1 Definition and existence
⚠️ A strategy profile where no player can improve by unilaterally deviating. Each player's strategy is a best response to the others'.
Nash (1950): ⚠️ every finite game has at least one equilibrium in mixed strategies.
The proof is a fixed-point argument (Kakutani/Brouwer) — ⚠️ which is why it's
non-constructive, and why §16 → gametheory-evolutionary-empirical-limits-and-computation's complexity results are not a contradiction: existence is
guaranteed, finding it is hard.
4.2 Mixed strategies
Randomizing over pure strategies. ⚠️ The defining property people miss: in a mixed equilibrium, you are indifferent among all strategies you play with positive probability.
⚠️ GOTCHA — this produces the most counterintuitive result in basic game theory. ⚠️ Your equilibrium mixing probabilities are determined by making your OPPONENT indifferent, not by optimizing your own payoff. Consequence: if a player's payoffs change, it is the other player's equilibrium mix that shifts, not their own. This routinely surprises people and it's a good test of whether you've understood mixed equilibrium.
Interpretations: deliberate randomization (⚠️ genuinely correct in poker and
penalty kicks), population frequencies (§14 → gametheory-evolutionary-empirical-limits-and-computation), or beliefs about an opponent's
type (Harsanyi purification).
4.3 ⚠️ What Nash equilibrium is not
- ⚠️ Not necessarily efficient. The Prisoner's Dilemma's unique equilibrium is Pareto- dominated (§5.1). Equilibrium ≠ good outcome.
- ⚠️ Not unique. Most games have many; selecting among them is an unsolved problem
(refinements: subgame perfection §7 →
gametheory-zero-sum-sequential-repeated-and-information, trembling-hand, risk vs payoff dominance). - ⚠️ Not a prediction of behaviour without additional assumptions about how players reach it. The theory says where you'd stop, not how you get there.
- ⚠️ Not a recommendation. "Play your Nash strategy" is only advice if you believe others will too.
Correlated equilibrium (Aumann) — ⚠️ players observe a shared signal and condition
on it. Weaker than Nash, can achieve better payoffs, is computationally easy (§16 → gametheory-evolutionary-empirical-limits-and-computation),
and is arguably the more natural concept — a traffic light is a correlating device.
§5. The Classic Games
5.1 Prisoner's Dilemma
Cooperate Defect
Cooperate 3, 3 0, 5
Defect 5, 0 1, 1 ⚠️ unique equilibrium: (Defect, Defect)
⚠️ Defection strictly dominates, so the equilibrium is unique and Pareto-dominated. Individual rationality produces a collectively worse outcome — the canonical statement that these can conflict.
⚠️ GOTCHA — the Prisoner's Dilemma is enormously over-applied, and the specific error is diagnosable. ⚠️ It requires
T > R > P > SAND2R > T + S. Most real situations labelled "a prisoner's dilemma" are actually Stag Hunt (§5.2 — a coordination problem with a good equilibrium available) or Chicken. The distinction is not pedantic: a coordination failure needs communication and trust; a true PD needs enforcement or repetition (§8 →gametheory-zero-sum-sequential-repeated-and-information). ⚠️ Prescribing the wrong remedy follows directly from misdiagnosing the game.
5.2 The other four you need
STAG HUNT (assurance) Two equilibria: (Stag,Stag) payoff-dominant,
(Hare,Hare) risk-dominant. ⚠️ A TRUST problem, not a
greed problem — the good outcome IS an equilibrium
CHICKEN (hawk-dove) Two asymmetric pure equilibria + a mixed one.
⚠️ Commitment wins: visibly removing your own options
(throwing away the steering wheel) is a strategic ADVANTAGE
BATTLE OF THE SEXES Coordination with conflicting preferences.
⚠️ Both want to coordinate; they disagree on where
MATCHING PENNIES ⚠️ Zero-sum, NO pure equilibrium, unique mixed equilibrium
at 50/50. The archetype for §6
⚠️ Commitment as advantage (Schelling) is the deep idea in Chicken: in strategic settings, reducing your own options can improve your outcome — burning bridges, binding contracts, publicly irreversible positions. ⚠️ This inverts the decision-theoretic intuition that more options are always weakly better, and the inversion is real.
04 同仓库更多 Skills
white-paper-writing
Use when writing white papers, technical reports, policy documents, thought leadership pieces, or any long-form authoritative document. Triggers on requests to write, structure, outline, or review white papers, technical reports, position papers, policy briefs, or evidence-driven arguments. Also use for B2B content marketing strategy, technical documentation requiring authority, academic-style reports, or any document that must educate while persuading. Use when someone asks "how should I structure this" for a long-form document.
prose-technical-mechanics
Use when editing or writing prose at a technical/mechanical level: choosing between Anglo-Saxon and Latinate diction, placing information in topic vs stress positions, writing periodic vs cumulative sentences, handling sentence rhythm, punctuation as logic, transitions, paragraph structure, or phonesthetics. Use when someone wants to understand WHY prose works or doesn't, not just surface-level style advice. Triggers on word choice, sentence structure, emphasis, rhythm, diction, or any deep writing mechanics question.
prose-mechanics
Use when writing, editing, or analyzing prose style; when trying to avoid generic AI-sounding writing; when applying specific prose styles such as journalistic, academic, narrative, KJV biblical, Hemingway plain, Latinate/Jamesian, Orwellian, Augustan, Victorian, Early Modern (Bacon, Browne, Milton), or free indirect discourse; or when coaching someone on writing voice and mechanics. Triggers on any request to improve, analyze, or write in a specific prose style, to fix "AI-sounding" writing, to make prose more vivid or concrete, to vary sentence rhythm, or to understand how a particular writer achieves their effects.
narrative-structure
Use when structuring stories, pitches, presentations, white papers, or any persuasive narrative; analyzing the hero's journey in fiction or non-fiction; applying StoryBrand for business communication; designing narrative arcs for thought leadership; or understanding Vogler, Snyder, Propp, or Campbell frameworks. Triggers on any narrative structure, storytelling, pitch structure, StoryBrand, hero's journey, three-act structure, Freytag, Save the Cat, narrative arc, or "how should I structure this story/pitch/presentation" request.
legal-brief-writing
Use when writing legal briefs, motions, or any court filing; when analyzing judicial cognition and persuasion; when applying elite brief-writing techniques from Roberts, Clement, Kagan, Garner; or when structuring legal arguments for maximum persuasion. Also use for autistic/neurodivergent advocates who need to understand how their natural communication style maps to elite legal advocacy. Triggers on any legal writing, brief drafting, motion preparation, or legal argument structuring request.
journal-publishability-criteria
Use whenever a research paper, article or preprint is being written, revised, reviewed or published, and specifically before every publication of docs/paper.md. Covers what legitimate journals actually gate on, in order of gating — the editor's desk screen (scope, novelty, package problems), what reviewers judge (methodological soundness, claims the evidence can carry, statistical and qualitative reporting), EQUATOR reporting guidelines, the ICMJE authorship criteria, COPE's view of text recycling, the near-universal generative-AI policy (no AI author, disclosure, author responsibility), TOP 2025 transparency standards, how to verify a journal is legitimate, submitting without an institution, and a pre-submission checklist. Triggers on requests to evaluate, review, prepare or submit a paper, to judge publishability, to write a declarations or data-availability section, to disclose AI use, or to check a journal.