AGENT SKILL

gametheory-reference

Use when correcting a game-theory misconception, looking up a solution concept, formula or standard result, finding the canon, or needing a picker and a modelling checklist to run before claiming a strategic situation is a particular game. Companion to the other game-theory skills.

License 许可
MIT
最近更新
2026-11-15
Stars
★ 5
Forks
3

01 前言 Front Matter

name
gametheory-reference
description
Use when correcting a game-theory misconception, looking up a solution concept, formula or standard result, finding the canon, or needing a picker and a modelling checklist to run before claiming a strategic situation is a particular game. Companion to the other game-theory skills.

03 SKILL.md 正文

Game Theory: Misconceptions, Formulas, and Canon

Part 5 of 5 of the Game Theory reference (plugin game-theory), covering §17–§21. Sibling skills: gametheory-framework-nash-and-classic-games (§0–§5), 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). Section numbers are shared across the set; a reference written as §N → skill points 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 consolidates.

The three ideas that matter most:

  1. ⚠️ "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 → gametheory-framework-nash-and-classic-games).
  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 → gametheory-framework-nash-and-classic-games).
  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).

§17. Misconceptions

Misconception Correction
Rational means selfish ⚠️ Means consistent. Utility can include anyone's welfare (§1.2 → gametheory-framework-nash-and-classic-games)
Game theory assumes people are cold calculators ⚠️ It assumes consistency; the failing assumptions are computational (§1.2 → gametheory-framework-nash-and-classic-games, §15 → gametheory-evolutionary-empirical-limits-and-computation)
Nash equilibrium is the optimal outcome ⚠️ The PD's equilibrium is Pareto-dominated (§4.3 → gametheory-framework-nash-and-classic-games)
Nash equilibrium is unique Most games have many; selection is unsolved (§4.3 → gametheory-framework-nash-and-classic-games)
Nash equilibrium predicts behaviour ⚠️ It's a consistency condition, not a dynamic theory (§4.3 → gametheory-framework-nash-and-classic-games)
In a mixed equilibrium you optimize your own payoff ⚠️ Your mix makes your OPPONENT indifferent (§4.2 → gametheory-framework-nash-and-classic-games)
Most situations are prisoner's dilemmas ⚠️ Usually Stag Hunt or Chicken. Check T>R>P>S and 2R>T+S (§5.1 → gametheory-framework-nash-and-classic-games)
Most competition is zero-sum ⚠️ Rarely. Zero-sum thinking is a named bias (§6 → gametheory-zero-sum-sequential-repeated-and-information)
More options are always better ⚠️ Commitment — removing options — can win (§5.2 → gametheory-framework-nash-and-classic-games)
A threat in a Nash equilibrium is credible ⚠️ Subgame perfection exists to exclude these (§7 → gametheory-zero-sum-sequential-repeated-and-information)
Cooperation is impossible among self-interested agents ⚠️ Repetition, reputation, structure all sustain it (§8 → gametheory-zero-sum-sequential-repeated-and-information, §14 → gametheory-evolutionary-empirical-limits-and-computation)
The folk theorem shows cooperation emerges ⚠️ It shows ANYTHING can be an equilibrium — that's weak, not strong (§8 → gametheory-zero-sum-sequential-repeated-and-information)
Tit-for-tat is the best strategy ⚠️ Fragile to noise; no strategy is universally best (§8 → gametheory-zero-sum-sequential-repeated-and-information)
Second-price auctions are always truthful in practice ⚠️ Requires private values and a trusted auctioneer (§12 → gametheory-bargaining-cooperative-mechanism-design-and-matching)
A better voting system could be non-manipulable ⚠️ Gibbard-Satterthwaite says no (§12 → gametheory-bargaining-cooperative-mechanism-design-and-matching)
Efficient trade can always be arranged ⚠️ Myerson-Satterthwaite says no (§12 → gametheory-bargaining-cooperative-mechanism-design-and-matching)
Which side proposes in matching is a detail ⚠️ It determines who gets their best stable match (§13 → gametheory-bargaining-cooperative-mechanism-design-and-matching)
Adding capacity to a network helps ⚠️ Braess's paradox (§16 → gametheory-evolutionary-empirical-limits-and-computation)
Equilibria are what players will compute ⚠️ General Nash is PPAD-complete (§16 → gametheory-evolutionary-empirical-limits-and-computation)
Evolutionary game theory needs rational agents ⚠️ It needs none. Selection replaces reasoning (§14 → gametheory-evolutionary-empirical-limits-and-computation)

§18. Formulas and Facts

NASH EQUILIBRIUM
sᵢ* ∈ argmax uᵢ(sᵢ, s₋ᵢ*)  for all i
⚠️ Every finite game has a mixed equilibrium (Nash 1950, via fixed point)
⚠️ Mixed equilibrium: you are indifferent over your support

ZERO-SUM
⚠️ max min = min max (von Neumann 1928); solvable by LP

REPEATED GAMES
Cooperation sustainable when δ ≥ (T−R)/(T−P)   [grim trigger, PD]
⚠️ Folk theorem: any individually rational feasible payoff, for δ high enough

BARGAINING
Nash solution: max (u₁−d₁)(u₂−d₂)
⚠️ Rubinstein: more patient player gets more

COOPERATIVE
Shapley: φᵢ = Σ_S [|S|!(n−|S|−1)!/n!] · [v(S∪{i}) − v(S)]
⚠️ = average marginal contribution over arrival orders

AUCTIONS
⚠️ Second-price: bid your true value (dominant)
First-price: shade below value
⚠️ Revenue equivalence under private/independent values, risk neutrality, symmetry

COMPLEXITY ⚠️
Zero-sum Nash: P · General Nash: PPAD-complete
Correlated equilibrium: P · Shapley value: #P-hard
Price of Anarchy, selfish routing with linear latency: 4/3

EVOLUTIONARY
Replicator: ẋᵢ = xᵢ(fᵢ − f̄)
⚠️ ESS ⟹ Nash, but not conversely

EXPERIMENTAL ⚠️
Ultimatum: modal offers ~40–50%; offers <20% often rejected
p-beauty contest: equilibrium 0; observed 1–3 levels of reasoning

§19. Books

Author Work Why
Osborne & Rubinstein A Course in Game Theory ⚠️ The rigorous standard. Free from Osborne's site
Fudenberg & Tirole Game Theory The graduate reference
Osborne An Introduction to Game Theory ⚠️ The accessible entry point
Schelling The Strategy of Conflict ⚠️ Commitment, focal points, credibility. Almost no mathematics and one of the most influential books in the field
von Neumann & Morgenstern Theory of Games and Economic Behavior (1944) The founding text
Camerer Behavioral Game Theory ⚠️ §15 → gametheory-evolutionary-empirical-limits-and-computation, definitively
Roth Who Gets What — and Why ⚠️ §13 → gametheory-bargaining-cooperative-mechanism-design-and-matching from the person who built the markets; Nobel work, readable
Krishna Auction Theory §12 → gametheory-bargaining-cooperative-mechanism-design-and-matching
Nisan, Roughgarden, Tardos, Vazirani Algorithmic Game Theory ⚠️ §16 → gametheory-evolutionary-empirical-limits-and-computation. Free online
Maynard Smith Evolution and the Theory of Games §14 → gametheory-evolutionary-empirical-limits-and-computation, foundational
Nowak Evolutionary Dynamics §14 → gametheory-evolutionary-empirical-limits-and-computation, modern
Dixit & Nalebuff The Art of Strategy ⚠️ Popular, genuinely good on intuition
Axelrod The Evolution of Cooperation §8 → gametheory-zero-sum-sequential-repeated-and-information, and read the noise critiques alongside

§20. Quick Reference

20.1 Picker

Situation Tool
Simultaneous, complete information Nash equilibrium; check dominance first (§3 → gametheory-framework-nash-and-classic-games, §4 → gametheory-framework-nash-and-classic-games)
Strictly competitive ⚠️ Minimax — you get a genuine guarantee (§6 → gametheory-zero-sum-sequential-repeated-and-information)
Sequential, perfect information Backward induction → SPNE (§7 → gametheory-zero-sum-sequential-repeated-and-information)
Non-credible threats in the solution ⚠️ Apply subgame perfection (§7 → gametheory-zero-sum-sequential-repeated-and-information)
Ongoing relationship ⚠️ Repeated game; check the discount factor (§8 → gametheory-zero-sum-sequential-repeated-and-information)
Asymmetric information about types Bayesian game; signalling or screening (§9 → gametheory-zero-sum-sequential-repeated-and-information)
Splitting a surplus Nash bargaining or Rubinstein; ⚠️ improve your BATNA (§10 → gametheory-bargaining-cooperative-mechanism-design-and-matching)
Allocating joint costs or credit ⚠️ Shapley value (§11 → gametheory-bargaining-cooperative-mechanism-design-and-matching)
Designing rules for self-interested agents Mechanism design; ⚠️ check the impossibility theorems first (§12 → gametheory-bargaining-cooperative-mechanism-design-and-matching)
Selling a single item Second-price for truthfulness; ⚠️ check revenue-equivalence assumptions (§12 → gametheory-bargaining-cooperative-mechanism-design-and-matching)
Two-sided allocation without prices ⚠️ Deferred acceptance; decide who proposes (§13 → gametheory-bargaining-cooperative-mechanism-design-and-matching)
Populations without deliberation Replicator dynamics, ESS (§14 → gametheory-evolutionary-empirical-limits-and-computation)
Predicting real human play ⚠️ QRE, level-k — not plain Nash (§15 → gametheory-evolutionary-empirical-limits-and-computation)
Solving a large game computationally ⚠️ No-regret learning (CFR), not equilibrium solvers (§16 → gametheory-evolutionary-empirical-limits-and-computation)

20.2 Modelling checklist

  • [ ] Who are the players, and is the relevant set actually larger? (§1 → gametheory-framework-nash-and-classic-games)
  • [ ] Are strategies complete contingent plans, or am I listing moves? (§1.1 → gametheory-framework-nash-and-classic-games)
  • [ ] Are payoffs utilities, including non-monetary concerns? (§1.2 → gametheory-framework-nash-and-classic-games)
  • [ ] Perfect vs complete information — which am I assuming? (§2 → gametheory-framework-nash-and-classic-games)
  • [ ] Is this genuinely a PD, or a coordination game? ⚠️ Check the inequalities (§5.1 → gametheory-framework-nash-and-classic-games)
  • [ ] One-shot or repeated, and is the horizon known? (§8 → gametheory-zero-sum-sequential-repeated-and-information)
  • [ ] Does my solution rely on a non-credible threat? (§7 → gametheory-zero-sum-sequential-repeated-and-information)
  • [ ] Is the equilibrium unique — and if not, what selects among them? (§4.3 → gametheory-framework-nash-and-classic-games)
  • [ ] Would real people play this way? (§15 → gametheory-evolutionary-empirical-limits-and-computation)
  • [ ] Could the players actually compute this? (§16 → gametheory-evolutionary-empirical-limits-and-computation)

§21. Method

No searches were run; none would have helped. ⚠️ This is settled mathematics. von Neumann's minimax theorem (1928), von Neumann & Morgenstern (1944), Nash (1950), Arrow (1951), Gale-Shapley (1962), Selten (1965), Harsanyi (1967), Gibbard-Satterthwaite (1973), Maynard Smith & Price (1973), Rubinstein (1982), Myerson-Satterthwaite (1983). ⚠️ The theorems have not changed.

Sources are the references in §19 — chiefly Osborne & Rubinstein and Fudenberg & Tirole for §1–§12 → gametheory-framework-nash-and-classic-games, gametheory-zero-sum-sequential-repeated-and-information, gametheory-bargaining-cooperative-mechanism-design-and-matching, Roth for §13 → gametheory-bargaining-cooperative-mechanism-design-and-matching, Nowak and Maynard Smith for §14 → gametheory-evolutionary-empirical-limits-and-computation, Camerer for §15 → gametheory-evolutionary-empirical-limits-and-computation, and Nisan et al. for §16 → gametheory-evolutionary-empirical-limits-and-computation.

Confidence: high on the mathematics, and ⚠️ I have stated theorems with their hypotheses throughout, because in this field the hypotheses are where the content is — revenue equivalence, the folk theorem, and second-price truthfulness are all routinely invoked outside the conditions under which they hold, and §12 → gametheory-bargaining-cooperative-mechanism-design-and-matching and §17 flag each case.

⚠️ Three places I've taken a position rather than reporting neutrally.

§5.1 → gametheory-framework-nash-and-classic-games's warning about the Prisoner's Dilemma is deliberate and I'd defend it strongly. ⚠️ The PD is the most over-applied model in social science, and the misdiagnosis has practical consequences: a Stag Hunt is a trust problem where the good outcome is already an equilibrium and communication may suffice; a true PD requires enforcement or repetition. Prescribing the wrong remedy follows directly from naming the wrong game, and the inequality test in §5.1 → gametheory-framework-nash-and-classic-games takes thirty seconds.

§8 → gametheory-zero-sum-sequential-repeated-and-information's framing of the folk theorem as bad news is the standard view among theorists and the opposite of how it's usually popularized. ⚠️ "Anything can be sustained in equilibrium" is a statement that the concept has lost its predictive content in that setting, and presenting it as "game theory explains cooperation" overstates it.

§16 → gametheory-evolutionary-empirical-limits-and-computation's complexity concern I treat as substantive rather than technical. ⚠️ If computing an equilibrium is PPAD-complete, the claim that players locate it needs defending, and correlated equilibrium's tractability plus its reachability by simple no-regret learning is a genuine argument for preferring it. The practical successes in §16 → gametheory-evolutionary-empirical-limits-and-computation are no-regret learning methods, not equilibrium solvers — ⚠️ which I'd read as the field's computational results being vindicated in practice rather than worked around.

§15 → gametheory-evolutionary-empirical-limits-and-computation I've tried to state fairly in both directions: the experimental failures are real and large, ⚠️ and they mostly indict the auxiliary assumptions rather than the framework — which is why the behavioural variants fit inside it.

04 同仓库更多 Skills

the-vibey-project
the-vibey-project/vibey

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.

★ 5
the-vibey-project
the-vibey-project/vibey

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.

★ 5
the-vibey-project
the-vibey-project/vibey

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.

★ 5
the-vibey-project
the-vibey-project/vibey

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.

★ 5
the-vibey-project
the-vibey-project/vibey

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.

★ 5
the-vibey-project
the-vibey-project/vibey

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.

★ 5