{"id":"fa3fcec2-b6e1-4457-8f9d-f64bd7176370","entityType":"agent","slug":"clawhub-deciqai-nash-equilibrium","name":"Nash Equilibrium","canonicalUrl":"https://www.xpersona.co/agent/clawhub-deciqai-nash-equilibrium","canonicalPath":"/agent/clawhub-deciqai-nash-equilibrium","generatedAt":"2026-10-11T11:25:43.489Z","source":"CLAWHUB","claimStatus":"UNCLAIMED","verificationTier":"NONE","summary":{"evidence":{"source":"editorial-content","verified":true,"confidence":"high","updatedAt":"2026-10-11T09:18:05.017Z","emptyReason":null},"description":"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep e... Skill: Nash Equilibrium Owner: deciqai Summary: Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep e... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-07-16T18:08:07.042Z | user Description tail link + agents machine-readable metadata line (deciqai.com/s/nash-equilibrium.json) v1.0.4 | 2026-07-09T11:19:05.292Z","descriptionLabel":"Technical summary","evidenceSummary":"Capability contract not published. No trust telemetry is available yet. 1.1K downloads reported by the source. Last updated 10/11/2026.","installCommand":"clawhub skill install s17a4mqcnk515kvaca5ze55d0x88pfpx:nash-equilibrium","sourceUrl":"https://clawhub.ai/deciqai/nash-equilibrium","homepage":"https://clawhub.ai/deciqai/skills/nash-equilibrium","primaryLinks":[{"label":"View on ClawHub","url":"https://clawhub.ai/deciqai/nash-equilibrium","kind":"source"},{"label":"Homepage","url":"https://clawhub.ai/deciqai/skills/nash-equilibrium","kind":"homepage"}],"safetyScore":84,"overallRank":62,"popularityScore":61,"trustScore":null,"claimedByName":null,"isOwner":false,"seoDescription":"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep e..."},"coverage":{"evidence":{"source":"public-profile","verified":false,"confidence":"medium","updatedAt":"2026-10-11T09:18:05.017Z","emptyReason":null},"protocols":[{"protocol":"OPENCLEW","label":"OpenClaw","status":"self-declared","notes":"Declared in the public agent profile."}],"capabilities":[],"verifiedCount":0,"selfDeclaredCount":1,"capabilityMatrix":{"rows":[{"key":"OPENCLEW","type":"protocol","support":"unknown","confidenceSource":"profile","notes":"Listed on profile"}],"flattenedTokens":"protocol:OPENCLEW|unknown|profile"}},"adoption":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-11T09:18:05.017Z","emptyReason":null},"stars":null,"forks":null,"downloads":1102,"packageName":null,"latestVersion":"1.0.5","tractionLabel":"1.1K downloads"},"release":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-11T09:18:05.003Z","emptyReason":null},"lastUpdatedAt":"2026-10-11T09:18:05.017Z","lastCrawledAt":"2026-10-11T09:18:05.003Z","lastIndexedAt":null,"nextCrawlAt":"2026-10-12T09:18:05.003Z","lastVerifiedAt":null,"highlights":[{"version":"1.0.5","createdAt":"2026-07-16T18:08:07.042Z","changelog":"Description tail link + agents machine-readable metadata line (deciqai.com/s/nash-equilibrium.json)","fileCount":7,"zipByteSize":14829},{"version":"1.0.4","createdAt":"2026-07-09T11:19:05.292Z","changelog":"Refresh: 2024-2026 AI-era worked examples added (strategy/leadership + systems/game-theory batch)","fileCount":7,"zipByteSize":14992},{"version":"1.0.3","createdAt":"2026-07-08T11:11:10.444Z","changelog":"Footer now uses /c/<slug> short link (fixes UTM truncation when SKILL.md is read in a terminal)","fileCount":6,"zipByteSize":11619},{"version":"1.0.2","createdAt":"2026-07-08T00:56:05.502Z","changelog":"Refreshed content + GitHub star link in footer","fileCount":6,"zipByteSize":11591},{"version":"1.0.1","createdAt":"2026-07-07T22:29:59.671Z","changelog":"Add catalog categories and topics","fileCount":5,"zipByteSize":9160},{"version":"1.0.0","createdAt":"2026-06-30T11:18:00.313Z","changelog":"Initial publish","fileCount":5,"zipByteSize":9167}]},"execution":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No published capability contract is available yet."},"installCommand":"clawhub skill install s17a4mqcnk515kvaca5ze55d0x88pfpx:nash-equilibrium","setupComplexity":"low","setupSteps":["Setup complexity is classified as HIGH. You must provision dedicated cloud infrastructure or an isolated VM. Do not run this directly on your local workstation.","Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data."],"contract":{"contractStatus":"missing","authModes":[],"requires":[],"forbidden":[],"supportsMcp":false,"supportsA2a":false,"supportsStreaming":false,"inputSchemaRef":null,"outputSchemaRef":null,"dataRegion":null,"contractUpdatedAt":null,"sourceUpdatedAt":null,"freshnessSeconds":null},"invocationGuide":{"preferredApi":{"snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/trust"},"curlExamples":["curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/snapshot\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/contract\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/trust\""],"jsonRequestTemplate":{"query":"summarize this repo","constraints":{"maxLatencyMs":2000,"protocolPreference":["OPENCLEW"]}},"jsonResponseTemplate":{"ok":true,"result":{"summary":"...","confidence":0.9},"meta":{"source":"CLAWHUB","generatedAt":"2026-10-11T11:25:43.482Z"}},"retryPolicy":{"maxAttempts":3,"backoffMs":[500,1500,3500],"retryableConditions":["HTTP_429","HTTP_503","NETWORK_TIMEOUT"]}},"endpoints":{"dossierUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/dossier","snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-deciqai-nash-equilibrium/trust"}},"reliability":{"evidence":{"source":"runtime-metrics","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No trust, reliability, or runtime telemetry is available."},"trust":{"status":"unavailable","handshakeStatus":"UNKNOWN","verificationFreshnessHours":null,"reputationScore":null,"p95LatencyMs":null,"successRate30d":null,"fallbackRate":null,"attempts30d":null,"trustUpdatedAt":null,"trustConfidence":"unknown","sourceUpdatedAt":null,"freshnessSeconds":null},"decisionGuardrails":{"doNotUseIf":["Contract metadata is missing or unavailable for deterministic execution."],"safeUseWhen":[],"riskFlags":["missing_or_unavailable_contract","trust_data_unavailable","schema_references_missing"],"operationalConfidence":"low"},"executionMetrics":{"observedLatencyMsP50":null,"observedLatencyMsP95":null,"estimatedCostUsd":null,"uptime30d":null,"rateLimitRpm":null,"rateLimitBurst":null,"lastVerifiedAt":null,"verificationSource":null},"runtimeMetrics":{"successRate":null,"avgLatencyMs":null,"avgCostUsd":null,"hallucinationRate":null,"retryRate":null,"disputeRate":null,"p50Latency":null,"p95Latency":null,"lastUpdated":null}},"benchmarks":{"evidence":{"source":"no-benchmark-data","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No benchmark suites or observed failure patterns are available."},"suites":[],"failurePatterns":[]},"artifacts":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"high","updatedAt":"2026-10-11T09:18:05.017Z","emptyReason":null},"readme":"Skill: Nash Equilibrium\n\nOwner: deciqai\n\nSummary: Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep e...\n\nTags: latest:1.0.5\n\nVersion history:\n\nv1.0.5 | 2026-07-16T18:08:07.042Z | user\n\nDescription tail link + agents machine-readable metadata line (deciqai.com/s/nash-equilibrium.json)\n\nv1.0.4 | 2026-07-09T11:19:05.292Z | user\n\nRefresh: 2024-2026 AI-era worked examples added (strategy/leadership + systems/game-theory batch)\n\nv1.0.3 | 2026-07-08T11:11:10.444Z | user\n\nFooter now uses /c/<slug> short link (fixes UTM truncation when SKILL.md is read in a terminal)\n\nv1.0.2 | 2026-07-08T00:56:05.502Z | user\n\nRefreshed content + GitHub star link in footer\n\nv1.0.1 | 2026-07-07T22:29:59.671Z | user\n\nAdd catalog categories and topics\n\nv1.0.0 | 2026-06-30T11:18:00.313Z | user\n\nInitial publish\n\nArchive index:\n\nArchive v1.0.5: 7 files, 14829 bytes\n\nFiles: examples/ai-capex-race-among-hyperscalers-2024-2026.md (5956b), examples/nash-1950-51-and-the-fcc-auctions-1994.md (6810b), examples/penalty-kicks-professionals-play-minimax.md (4283b), references/sources.md (2061b), skill-card.md (1988b), SKILL.md (7865b), _meta.json (135b)\n\nFile v1.0.5:SKILL.md\n\n---\nname: nash-equilibrium\ndescription: \"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep ending up in a bad outcome even though everyone prefers better', or is analyzing a strategic situation with multiple rational counterparties (pricing, negotiation, M&A, regulation, platform launch). Do NOT activate when: the decision is essentially solo with no strategic counterparty; the counterparty is clearly irrational or acting on emotion rather than self-interest. More: deciqai.com/c/nash-equilibrium\"\n---\n\n# Nash Equilibrium\n\n## Overview\n\nA **Nash equilibrium** is a stable point in multi-player interaction: a combination of strategies where no player can improve their payoff by unilaterally changing their own strategy, given others hold theirs fixed. Key properties: (1) best-response logic — the equilibrium is a fixed point of mutual best-responses; (2) equilibria can be Pareto-suboptimal (prisoner's dilemma); (3) multiple equilibria are common; (4) equilibrium does not predict the path to get there.\n\nComposes with `prisoners-dilemma`, `repeated-games-reputation`, `signaling-games`, `pricing-strategy`, and `batna-zopa`.\n\n## When to Use\n\n- Designing pricing in a competitive market with rational rivals\n- Negotiating with a sophisticated counterparty; modeling regulatory or legislative outcomes\n- Evaluating M&A or partnership decisions where multiple parties react strategically\n- Designing auctions, marketplaces, or platform mechanisms\n- Assessing an investment or capacity arms race where rivals match each other (e.g., AI capex spending, AI-native competition, matching AI adoption to keep position)\n\n**Not when:** the situation is essentially solo; the counterparty is not rational; modeling cost exceeds the decision's value.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line what-it-is: find the point in a multi-player situation where no one has incentive to deviate from their strategy.\n2. Check fit: no rational counterparty? Nash adds less value than other frameworks.\n3. Elicit their real case: who are the players, what can each do, what does each gain or lose?\n> **[WAIT — do not advance until user responds]**\n4. Run The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close by naming the equilibrium and recommended action; if equilibrium is bad, name a redesign option.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Specify the game:** Players · Actions per player · Payoff matrix or game tree · Information structure (full vs. private) · Sequential or simultaneous.\n\n**Step 2 — Best-response analysis:** For each player, find the optimal action given each combination of others' strategies. The combination where everyone is best-responding is a Nash equilibrium.\n\n**Step 3 — Identify all equilibria:** Pure-strategy (deterministic) and mixed-strategy (randomized). If multiple equilibria, identify the most plausible focal point.\n\n**Step 4 — Evaluate quality:** Pareto-optimal? If not, what better collective outcome does the structure prevent?\n\n**Step 5 — Act or redesign:** Play equilibrium strategy if acceptable. If bad: change rules, add repeated interaction, change payoffs, add transparency — or walk away.\n\n**Step 6 — Plan counterparty response:** What do others do once you act? Contingency if they deviate from equilibrium.\n\n## Output Template\n\n```markdown\n# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate\n```\n\n*→ Method in Action: [Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md) · [Penalty Kicks and the Minimax Test](examples/penalty-kicks-professionals-play-minimax.md)*\n*→ 2026 lens: [The AI-Capex Race Among Hyperscalers (2024–2026)](examples/ai-capex-race-among-hyperscalers-2024-2026.md)*\n\n## Pack: Application Patterns\n\n| Domain | Typical equilibrium | Common pitfall |\n|---|---|---|\n| Pricing in oligopoly | Cournot markup or Bertrand marginal-cost | Assuming price-cuts go unanswered |\n| Auction bidding | Bid to value (English) or shaded bid (sealed) | Ignoring winner's curse in common-value auctions |\n| Platform launch | Multiple equilibria; focal point determines winner | Treating launch as solo, not a coordination game |\n| M&A bidding | Strategic incumbent often overpays | Failing to model how target plays bidders against each other |\n| Cartel formation | Cooperation possible if discount rate is low | Underestimating defection incentives and detection lag |\n\n## Applying It Well\n\nModel the game explicitly before acting. Equilibrium is not optimum — if the equilibrium is bad, redesign the game (add repetition, change payoffs, build trust). When multiple equilibria exist, use focal points or commitment devices to coordinate on the one you want.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We'll just do the right thing; no need for game theory\" | The right thing depends on what others do. Without modeling reactions, you'll be surprised. |\n| [D] \"Game theory is too abstract for real decisions\" | The 1994 FCC auction raised $617M using Nash modeling. Auction design, antitrust, military strategy are operational applications. |\n| [D] \"Counterparties aren't rational; theory doesn't apply\" | They're often more rational than assumed. Use Nash as a baseline and adjust for behavioral deviations. |\n| [D] \"Multiple equilibria mean Nash isn't useful here\" | Multiple equilibria mean selection matters more — use focal points, communication, commitment devices. |\n| [D] \"First mover always wins\" | Often false. Equilibrium analysis tells you when first move helps and when it hurts. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Making a strategic decision without modeling how counterparties will respond\n- Assuming counterparties will react the way you'd react (mirror-imaging)\n- The equilibrium is bad and you haven't considered redesigning the game\n- Playing a repeated game using single-round logic\n- Multiple equilibria exist and you haven't considered which will emerge\n\n## Verification\n\n- [ ] Players, actions, and payoffs explicitly specified\n- [ ] Best-response analysis performed for each player\n- [ ] Nash equilibria identified; Pareto-optimality evaluated\n- [ ] If multiple equilibria, equilibrium-selection considered\n- [ ] If equilibrium is bad, game-redesign options considered\n- [ ] Counterparty response plan exists for equilibrium play and deviation\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 227 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/nash-equilibrium** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\n*Agents: latest version & machine-readable metadata → https://www.deciqai.com/s/nash-equilibrium.json*\n\nFile v1.0.5:_meta.json\n\n{\n  \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n  \"slug\": \"nash-equilibrium\",\n  \"version\": \"1.0.5\",\n  \"publishedAt\": 1784225287042\n}\n\nFile v1.0.5:references/sources.md\n\n# Sources — nash-equilibrium\n\n> *Primary sources for the [nash-equilibrium](../SKILL.md) skill.*\n\n- Nash, J. F. (1950). \"Equilibrium Points in N-Person Games.\" *Proceedings of the National Academy of Sciences*, 36(1), 48-49. The foundational two-page proof.\n- Nash, J. F. (1951). \"Non-Cooperative Games.\" *Annals of Mathematics*, 54(2), 286-295. The extended treatment.\n- von Neumann, J. & Morgenstern, O. (1944). *Theory of Games and Economic Behavior.* Princeton University Press. The precursor; covers zero-sum games.\n- Cournot, A. A. (1838). *Recherches sur les principes mathématiques de la théorie des richesses.* Paris: Hachette. The duopoly precursor.\n- Milgrom, P. (2004). *Putting Auction Theory to Work.* Cambridge University Press. ISBN 978-0521536729. The FCC auction case study.\n- Camerer, C. F. (2003). *Behavioral Game Theory: Experiments in Strategic Interaction.* Princeton University Press. ISBN 978-0691090399. Empirical deviations from Nash predictions.\n- Palacios-Huerta, I. (2003). \"Professionals Play Minimax.\" *Review of Economic Studies*, 70(2), 395-415. Field test of mixed-strategy equilibrium on professional penalty kicks.\n- Roth, A. E. (2002). \"The Economist as Engineer.\" *Econometrica*, 70(4), 1341-1378. Mechanism-design applications.\n- Schelling, T. C. (1960). *The Strategy of Conflict.* Harvard University Press. ISBN 978-0674840317. Strategic applications to security.\n- Tirole, J. (1988). *The Theory of Industrial Organization.* MIT Press. ISBN 978-0262200714. Industrial-organization applications.\n- Microsoft, Alphabet, Amazon, and Meta (2024–2026). Quarterly earnings calls and investor disclosures on AI-related capital expenditure guidance. Contemporary example of investment competition sustaining a high-capex equilibrium among rational rivals.\n- Investor commentary and reporting on the hyperscaler AI-capex build-out (2024–2026), e.g., quarterly earnings coverage in the financial press. Documents the \"risk of under-investing exceeds risk of over-investing\" framing that signals best-response spending.\n\nFile v1.0.5:examples/ai-capex-race-among-hyperscalers-2024-2026.md\n\n# Method in Action: The AI-Capex Race Among Hyperscalers (2024–2026)\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\nAcross 2024 and 2025 (and into 2026), the largest U.S. cloud-and-platform companies — Microsoft, Alphabet (Google), Amazon, and Meta — raised their capital expenditure to record levels, the bulk of it directed at AI data centers, accelerators, and power. On successive earnings calls, each firm's leadership publicly framed the risk of *under*-investing in AI capacity as outweighing the risk of *over*-investing. That is a telltale sign of a Nash equilibrium: not four firms independently arriving at the same plan, but four firms each best-responding to what the others are doing. This example runs the case through the skill's six-step Process.\n\n**Step 1 — Specify the game.** *Players:* the major hyperscalers (Microsoft, Alphabet, Amazon, Meta), plus adjacent capacity buyers. *Actions per player:* along a spectrum from \"spend aggressively on AI compute\" to \"hold capex flat and harvest cash.\" *Payoffs:* market position in AI-driven cloud and consumer products, which depends on relative capacity, not just absolute spend. *Information:* largely public — capex guidance is disclosed on quarterly earnings calls, so each player observes rivals' commitments with a short lag. *Timing:* effectively simultaneous and repeated, quarter after quarter, since all players revise guidance on a similar cadence.\n\n**Step 2 — Best-response analysis.** Consider any single firm's choice given the others are spending heavily. If it also spends, it holds its position and shares in whatever AI demand materializes. If it unilaterally pulls back, it frees cash and lifts near-term margins — but it risks ceding compute capacity, model quality, and enterprise-cloud share to rivals who kept building, in a market where capacity is a binding constraint and lead times for chips and power are long. Given rivals spending, the best response is to keep spending. The same logic holds for each player. The mutual-best-response fixed point is \"everyone spends heavily.\"\n\n**Step 3 — Identify all equilibria.** The high-capex profile is a pure-strategy Nash equilibrium: no firm can improve its position by unilaterally cutting while the others build. A low-capex \"everyone restrains\" profile would be more profitable collectively in the short run, but it is *not* an equilibrium — from it, any single firm gains by defecting and out-building the restrained rivals to capture position, so it unravels. This is the structural signature of a prisoner's-dilemma-type game: the cooperative outcome is not self-enforcing. The most plausible focal point is therefore the high-spend equilibrium, which is what the public capex trajectory through this period reflects.\n\n**Step 4 — Evaluate quality.** The equilibrium is not Pareto-optimal for the firms as a group: collectively they would earn higher near-term free cash flow spending less, if all could credibly commit to restraint. The game structure prevents that outcome because commitments are not binding and defection is rewarded by relative position — exactly the mechanism the [`prisoners-dilemma`](../prisoners-dilemma/SKILL.md) skill formalizes. (Note the distributional caveat: what is Pareto-suboptimal for the *spenders* can be highly beneficial for suppliers such as accelerator vendors and for users who receive rapidly cheapening AI capability.)\n\n**Step 5 — Act or redesign.** For an individual hyperscaler, the prescription is to play the equilibrium — keep investing to hold position — while managing the risk that demand disappoints. Redesign options that could relax the arms race are mostly unavailable or illegal: explicit coordination to jointly cut capex would be antitrust-exposed, and no player controls the payoffs. The realistic levers are unilateral efficiency moves that change one's own cost curve without conceding position — custom silicon, better utilization, cheaper power — which shift the payoffs of spending rather than the decision to spend. A firm *outside* the arms race (a smaller player or investor) should instead avoid competing head-on on raw capacity and look for positions the equilibrium leaves open.\n\n**Step 6 — Plan counterparty response.** Once you commit capex, expect rivals to match rather than retreat; guidance raises tend to be met with guidance raises, not with a rival standing down. The contingency that matters is deviation on the *demand* side: if AI monetization underdelivers, the equilibrium can shift toward restraint, but only if enough players cut roughly together — a single firm cutting first bears the position loss alone. So the watch-item is a coordinated signal (softening guidance across multiple players at once), not any one firm's move.\n\nThe mapped steps:\n1. Specify the game: hyperscalers choosing spend levels, public quarterly guidance, repeated and roughly simultaneous, payoff = relative AI/cloud position\n2. Best-response analysis: given rivals spending, each firm's best response is to keep spending; pulling back cedes position\n3. Equilibria: high-capex profile is a pure-strategy Nash equilibrium; the collectively cheaper \"all restrain\" profile is not self-enforcing\n4. Evaluate quality: Pareto-suboptimal for the spenders (a prisoner's-dilemma structure), yet often beneficial to suppliers and users\n5. Act or redesign: play the equilibrium; joint restraint is antitrust-barred, so compete on cost curve instead\n6. Counterparty response: expect matching, not retreat; the real trigger for change is roughly simultaneous demand-side softening across players\n\n*Sources: Company investor disclosures and earnings-call capex guidance, Microsoft, Alphabet, Amazon, and Meta (2024–2026); Nash, J. F. (1951), \"Non-Cooperative Games,\" Annals of Mathematics 54(2), 286–295; Tirole, J. (1988), The Theory of Industrial Organization, MIT Press (oligopoly investment competition).*\n\nFile v1.0.5:examples/nash-1950-51-and-the-fcc-auctions-1994.md\n\n# Method in Action: Nash 1950-51 and the FCC Auctions 1994\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\n**John Forbes Nash Jr.** (1928-2015) wrote his doctoral dissertation at Princeton in 1950, at age 21. The thesis was 28 pages. From it came the 1950 PNAS paper (two pages) and the 1951 *Annals of Mathematics* paper (10 pages). For these two papers — establishing the equilibrium concept that bears his name — Nash received the 1994 Nobel Memorial Prize in Economic Sciences (shared with Reinhard Selten and John Harsanyi for the broader development of non-cooperative game theory).\n\nThe theorem's mathematical proof used the Kakutani fixed-point theorem (Nash 1951). The proof's essence: in the space of mixed strategies, the best-response correspondence is upper-hemicontinuous and the strategy space is compact and convex; therefore by Kakutani, there exists a fixed point — i.e., a strategy combination where every player is best-responding. That fixed point is the equilibrium.\n\nWhat made Nash's contribution revolutionary was not the equilibrium concept itself (which had been informally articulated by Cournot in 1838 for the duopoly case) but the *generality* — the proof that *every* finite n-person game has at least one equilibrium. Before Nash, game theory could analyze only special cases. After Nash, game theory could analyze any finite strategic interaction. This is why Nash's two short papers are considered among the most important in 20th-century mathematics and economics.\n\nThe framework took decades to translate into operational policy. The breakthrough moment was the **1994 FCC spectrum auction**.\n\nThe U.S. Federal Communications Commission had historically allocated radio spectrum via \"comparative hearings\" (essentially bureaucratic beauty contests) and lotteries. Both methods had clear flaws: hearings were corruptible and slow; lotteries gave spectrum to entities who promptly resold it for windfall profits to actual operators. By the early 1990s, the value of spectrum was understood to be enormous and growing (especially for cellular telephony), and pressure was rising for a better mechanism.\n\nPaul Milgrom and Robert Wilson at Stanford were retained as consultants to design the new auction. Their core insight: spectrum auctions are complex strategic games with multiple bidders who hold private valuations for combinations of licenses. The auction rules must be designed so that bidders' Nash-equilibrium strategies — what each bidder rationally does given what every other bidder rationally does — produce efficient allocations (the spectrum ends up with the operators who value it most) and reasonable revenue.\n\nMilgrom and Wilson designed the **simultaneous multiple-round (SMR) auction**: licenses are auctioned simultaneously, in multiple rounds with rising bids, until no bidder is willing to raise any bid. This rule produces a Nash equilibrium in which bidders coordinate on combinations of licenses they want, while preventing bidders from being \"exposed\" (winning licenses individually but not the package they wanted). The design explicitly modeled bidder Nash strategies and chose rules to push the equilibrium toward efficient outcomes.\n\nFrom Milgrom's retrospective:\n\n> \"The SMR design rests on game-theoretic reasoning at every step. The rule structure determines what bidder strategies are available; the strategies determine the Nash equilibria; the equilibria determine the auction outcome. The FCC's 1994 auction raised $617 million for 99 licenses, against pre-auction estimates of $400 million for the prior beauty-contest method. The equilibrium logic worked: bidders coordinated through the rounds on packages of licenses that reflected their actual operational value, prices reflected willingness-to-pay, and the spectrum went to operators who could best use it. The subsequent FCC auctions — 87 of them through 2020 — have raised over $200 billion. The cumulative welfare gain to the U.S. economy is in the trillions. Game theory's policy-impact case rests substantially on this single application.\"\n>\n> — Milgrom (2004). *Putting Auction Theory to Work.* Cambridge University Press, ch. 1.\n\nMilgrom and Wilson received the 2020 Nobel Memorial Prize in Economic Sciences for their auction theory work, with the FCC case as the canonical example.\n\nThe framework has been applied at industrial scale beyond auctions:\n\n**Antitrust analysis.** Modern antitrust evaluation of mergers, vertical restraints, and platform conduct uses Nash-equilibrium modeling to predict how markets will behave under proposed changes. The DOJ and FTC employ economists who specialize in game-theoretic analysis of merger effects.\n\n**Mechanism design.** The broader field that grew from Nash's work — designing the rules of a game to produce desired equilibria — has applications in matching markets (kidney exchange, medical residency placement; Roth 2002, 2007 Nobel work), voting systems, school choice, and platform design.\n\n**Regulatory and policy analysis.** Setting taxes, designing regulations, structuring international trade agreements — all use Nash-equilibrium analysis to predict how strategic actors (firms, countries, voters) will respond.\n\n**Military and security strategy.** Cold War nuclear deterrence theory (Schelling 1960; Schelling 2005 Nobel for related work) is foundational Nash-equilibrium analysis. Modern cybersecurity, terrorism response, and great-power strategic competition all use the framework.\n\n**Corporate strategy.** Competitive interaction (pricing, capacity, R&D, marketing) in oligopolistic markets is Nash-equilibrium analysis. Sophisticated strategy departments at major firms (Walmart, Amazon, ExxonMobil, major banks) employ game theorists.\n\nThree operational lessons:\n\n**First, Nash equilibrium is the right tool when counterparties are rational and react to your moves.** If you don't model their reactions, you'll be systematically surprised by their behavior. The discipline of \"what's the best response to my move?\" prevents many strategic errors.\n\n**Second, equilibrium ≠ optimum.** The prisoner's dilemma is the canonical case: the equilibrium is bad for everyone, but the game's structure prevents the better outcome. Recognizing this is the prerequisite to redesigning the game (e.g., adding repeated interaction, changing payoffs, providing escrow).\n\n**Third, equilibrium-selection is often the operationally hardest problem.** Many games have multiple Nash equilibria, and the framework alone doesn't tell you which will emerge. Empirical, behavioral, and focal-point considerations matter. Treating Nash equilibrium as a complete theory of strategic behavior overstates its predictive power; treating it as one important component of strategic analysis is calibrated.\n\nFile v1.0.5:examples/penalty-kicks-professionals-play-minimax.md\n\n# Method in Action: Penalty Kicks and the Minimax Test (1995-2003)\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\nMixed-strategy Nash equilibrium makes a strange prediction: in games with no stable pure strategy, rational players should randomize — and randomize in exactly the proportions that make their opponent indifferent between responses. For decades this prediction fared poorly in laboratory experiments, where subjects failed to randomize properly. The economist Ignacio Palacios-Huerta asked a sharper question: do *professionals*, playing for real stakes at the top of their craft, play the equilibrium? He tested it on soccer penalty kicks.\n\n**Specify the game (Step 1).** Two players: kicker and goalkeeper. A penalty kick travels to the goal in roughly a third of a second — too fast for the keeper to react to the ball — so both players effectively choose simultaneously: kicker aims to his natural side or the opposite side; keeper dives one way or the other. Payoffs are opposed: the kicker wants to score, the keeper wants to save. Information is public — professional teams scout each other's penalty histories.\n\n**Best-response analysis (Step 2).** No pure strategy survives: if the kicker always shoots to his natural side, the keeper's best response is to always dive there, at which point the kicker's best response flips — and so on, forever. The best-response cycle never settles on a deterministic pair.\n\n**Identify the equilibrium (Step 3).** The game has a unique mixed-strategy Nash equilibrium (the minimax solution, since the game is zero-sum). Each player randomizes with probabilities that make the opponent indifferent. The equilibrium yields two testable predictions: (a) each player's success rate must be *equal* across his own actions — if kicking left scored more often than kicking right, he should kick left more, contradicting equilibrium; (b) each player's sequence of choices must be *serially independent* — any detectable pattern could be exploited.\n\n**Evaluate against data (Step 4).** Palacios-Huerta assembled roughly 1,400 penalty kicks from professional league matches in Spain, Italy, and England (1995-2000), recording kicker, keeper, direction, and outcome. Both predictions held. Scoring probabilities were statistically indistinguishable across a kicker's directions (in the neighborhood of 80% overall), and keepers' save rates were likewise equal across their choices. Player-by-player tests could not reject equality of payoffs for the overwhelming majority of individuals. And unlike laboratory subjects — who notoriously over-alternate when asked to randomize — professionals' direction choices passed standard tests of serial independence. Professionals play minimax.\n\n**Act or redesign (Step 5).** For the players there is nothing to redesign: the equilibrium *is* the optimal policy, and any deviation from it is exploitable. The operational lesson runs the other way — if you face a rival in a genuinely opposed, repeated interaction (competitive bidding, audit scheduling, security patrols), the equilibrium prescribes calibrated unpredictability, and your rival's data on you is the test of whether you achieved it.\n\n**Plan counterparty response (Step 6).** A kicker with a detectable bias hands the keeper a profitable deviation; scouting departments exist precisely to find such biases. Equilibrium play is the only strategy that leaves no pattern for the counterparty to exploit — the contingency plan is built into the randomization itself.\n\nThe mapped steps:\n1. Specify the game: kicker vs. keeper, simultaneous moves, opposed payoffs, public histories\n2. Best-response analysis: no pure strategy is a mutual best response; the cycle never settles\n3. Equilibria: a unique mixed-strategy (minimax) equilibrium with equal payoffs across own actions and serially independent choices\n4. Evaluate quality: field data on ~1,400 professional kicks matched both equilibrium predictions\n5. Act or redesign: equilibrium randomization is the optimal, unexploitable policy\n6. Counterparty response: any bias is scouted and punished; unpredictability is the defense\n\nPrimary source: Palacios-Huerta, I. (2003). \"Professionals Play Minimax.\" *Review of Economic Studies*, 70(2), 395-415.\n\nFile v1.0.5:skill-card.md\n\n## Description:\n\nGuides agents through Nash equilibrium analysis for strategic situations involving rational counterparties, including pricing, negotiation, auctions, regulation, platforms, M&A, and capacity races.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[deciqai](https://clawhub.ai/user/deciqai)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nDevelopers, strategists, and business operators use this skill to model players, actions, payoffs, best responses, equilibria, Pareto quality, and counterparty responses before making strategic decisions.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: Reusable observed-use notes could retain private business details or sensitive case facts.\n\nMitigation: Keep reusable notes sanitized or anonymized unless the information is intentionally meant to persist in the skill content.\n\nRisk: Strategic recommendations can be misleading when players, actions, payoffs, information structure, or counterparty rationality are underspecified.\n\nMitigation: Use the skill's verification checklist to confirm the game setup, equilibria, Pareto assessment, and counterparty response plan before acting.\n\n## Reference(s):\n\n- [Sources - nash-equilibrium](references/sources.md)\n- [Nash Equilibrium Skill Page](https://clawhub.ai/deciqai/skills/nash-equilibrium)\n\n## Skill Output:\n\n**Output Type(s):** [text, markdown, guidance]\n\n**Output Format:** [Markdown analysis with structured decision guidance]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [May ask step-by-step clarification questions before producing a final Nash analysis.]\n\n## Skill Version(s):\n\n1.0.5 (source: server release evidence)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.\n\nArchive v1.0.4: 7 files, 14992 bytes\n\nFiles: examples/ai-capex-race-among-hyperscalers-2024-2026.md (5956b), examples/nash-1950-51-and-the-fcc-auctions-1994.md (6810b), examples/penalty-kicks-professionals-play-minimax.md (4283b), references/sources.md (2061b), skill-card.md (2536b), SKILL.md (7722b), _meta.json (135b)\n\nFile v1.0.4:SKILL.md\n\n---\nname: nash-equilibrium\ndescription: \"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep ending up in a bad outcome even though everyone prefers better', or is analyzing a strategic situation with multiple rational counterparties (pricing, negotiation, M&A, regulation, platform launch). Do NOT activate when: the decision is essentially solo with no strategic counterparty; the counterparty is clearly irrational or acting on emotion rather than self-interest.\"\n---\n\n# Nash Equilibrium\n\n## Overview\n\nA **Nash equilibrium** is a stable point in multi-player interaction: a combination of strategies where no player can improve their payoff by unilaterally changing their own strategy, given others hold theirs fixed. Key properties: (1) best-response logic — the equilibrium is a fixed point of mutual best-responses; (2) equilibria can be Pareto-suboptimal (prisoner's dilemma); (3) multiple equilibria are common; (4) equilibrium does not predict the path to get there.\n\nComposes with `prisoners-dilemma`, `repeated-games-reputation`, `signaling-games`, `pricing-strategy`, and `batna-zopa`.\n\n## When to Use\n\n- Designing pricing in a competitive market with rational rivals\n- Negotiating with a sophisticated counterparty; modeling regulatory or legislative outcomes\n- Evaluating M&A or partnership decisions where multiple parties react strategically\n- Designing auctions, marketplaces, or platform mechanisms\n- Assessing an investment or capacity arms race where rivals match each other (e.g., AI capex spending, AI-native competition, matching AI adoption to keep position)\n\n**Not when:** the situation is essentially solo; the counterparty is not rational; modeling cost exceeds the decision's value.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line what-it-is: find the point in a multi-player situation where no one has incentive to deviate from their strategy.\n2. Check fit: no rational counterparty? Nash adds less value than other frameworks.\n3. Elicit their real case: who are the players, what can each do, what does each gain or lose?\n> **[WAIT — do not advance until user responds]**\n4. Run The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close by naming the equilibrium and recommended action; if equilibrium is bad, name a redesign option.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Specify the game:** Players · Actions per player · Payoff matrix or game tree · Information structure (full vs. private) · Sequential or simultaneous.\n\n**Step 2 — Best-response analysis:** For each player, find the optimal action given each combination of others' strategies. The combination where everyone is best-responding is a Nash equilibrium.\n\n**Step 3 — Identify all equilibria:** Pure-strategy (deterministic) and mixed-strategy (randomized). If multiple equilibria, identify the most plausible focal point.\n\n**Step 4 — Evaluate quality:** Pareto-optimal? If not, what better collective outcome does the structure prevent?\n\n**Step 5 — Act or redesign:** Play equilibrium strategy if acceptable. If bad: change rules, add repeated interaction, change payoffs, add transparency — or walk away.\n\n**Step 6 — Plan counterparty response:** What do others do once you act? Contingency if they deviate from equilibrium.\n\n## Output Template\n\n```markdown\n# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate\n```\n\n*→ Method in Action: [Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md) · [Penalty Kicks and the Minimax Test](examples/penalty-kicks-professionals-play-minimax.md)*\n*→ 2026 lens: [The AI-Capex Race Among Hyperscalers (2024–2026)](examples/ai-capex-race-among-hyperscalers-2024-2026.md)*\n\n## Pack: Application Patterns\n\n| Domain | Typical equilibrium | Common pitfall |\n|---|---|---|\n| Pricing in oligopoly | Cournot markup or Bertrand marginal-cost | Assuming price-cuts go unanswered |\n| Auction bidding | Bid to value (English) or shaded bid (sealed) | Ignoring winner's curse in common-value auctions |\n| Platform launch | Multiple equilibria; focal point determines winner | Treating launch as solo, not a coordination game |\n| M&A bidding | Strategic incumbent often overpays | Failing to model how target plays bidders against each other |\n| Cartel formation | Cooperation possible if discount rate is low | Underestimating defection incentives and detection lag |\n\n## Applying It Well\n\nModel the game explicitly before acting. Equilibrium is not optimum — if the equilibrium is bad, redesign the game (add repetition, change payoffs, build trust). When multiple equilibria exist, use focal points or commitment devices to coordinate on the one you want.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We'll just do the right thing; no need for game theory\" | The right thing depends on what others do. Without modeling reactions, you'll be surprised. |\n| [D] \"Game theory is too abstract for real decisions\" | The 1994 FCC auction raised $617M using Nash modeling. Auction design, antitrust, military strategy are operational applications. |\n| [D] \"Counterparties aren't rational; theory doesn't apply\" | They're often more rational than assumed. Use Nash as a baseline and adjust for behavioral deviations. |\n| [D] \"Multiple equilibria mean Nash isn't useful here\" | Multiple equilibria mean selection matters more — use focal points, communication, commitment devices. |\n| [D] \"First mover always wins\" | Often false. Equilibrium analysis tells you when first move helps and when it hurts. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Making a strategic decision without modeling how counterparties will respond\n- Assuming counterparties will react the way you'd react (mirror-imaging)\n- The equilibrium is bad and you haven't considered redesigning the game\n- Playing a repeated game using single-round logic\n- Multiple equilibria exist and you haven't considered which will emerge\n\n## Verification\n\n- [ ] Players, actions, and payoffs explicitly specified\n- [ ] Best-response analysis performed for each player\n- [ ] Nash equilibria identified; Pareto-optimality evaluated\n- [ ] If multiple equilibria, equilibrium-selection considered\n- [ ] If equilibrium is bad, game-redesign options considered\n- [ ] Counterparty response plan exists for equilibrium play and deviation\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 189 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/nash-equilibrium** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\nFile v1.0.4:_meta.json\n\n{\n  \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n  \"slug\": \"nash-equilibrium\",\n  \"version\": \"1.0.4\",\n  \"publishedAt\": 1783595945292\n}\n\nFile v1.0.4:references/sources.md\n\n# Sources — nash-equilibrium\n\n> *Primary sources for the [nash-equilibrium](../SKILL.md) skill.*\n\n- Nash, J. F. (1950). \"Equilibrium Points in N-Person Games.\" *Proceedings of the National Academy of Sciences*, 36(1), 48-49. The foundational two-page proof.\n- Nash, J. F. (1951). \"Non-Cooperative Games.\" *Annals of Mathematics*, 54(2), 286-295. The extended treatment.\n- von Neumann, J. & Morgenstern, O. (1944). *Theory of Games and Economic Behavior.* Princeton University Press. The precursor; covers zero-sum games.\n- Cournot, A. A. (1838). *Recherches sur les principes mathématiques de la théorie des richesses.* Paris: Hachette. The duopoly precursor.\n- Milgrom, P. (2004). *Putting Auction Theory to Work.* Cambridge University Press. ISBN 978-0521536729. The FCC auction case study.\n- Camerer, C. F. (2003). *Behavioral Game Theory: Experiments in Strategic Interaction.* Princeton University Press. ISBN 978-0691090399. Empirical deviations from Nash predictions.\n- Palacios-Huerta, I. (2003). \"Professionals Play Minimax.\" *Review of Economic Studies*, 70(2), 395-415. Field test of mixed-strategy equilibrium on professional penalty kicks.\n- Roth, A. E. (2002). \"The Economist as Engineer.\" *Econometrica*, 70(4), 1341-1378. Mechanism-design applications.\n- Schelling, T. C. (1960). *The Strategy of Conflict.* Harvard University Press. ISBN 978-0674840317. Strategic applications to security.\n- Tirole, J. (1988). *The Theory of Industrial Organization.* MIT Press. ISBN 978-0262200714. Industrial-organization applications.\n- Microsoft, Alphabet, Amazon, and Meta (2024–2026). Quarterly earnings calls and investor disclosures on AI-related capital expenditure guidance. Contemporary example of investment competition sustaining a high-capex equilibrium among rational rivals.\n- Investor commentary and reporting on the hyperscaler AI-capex build-out (2024–2026), e.g., quarterly earnings coverage in the financial press. Documents the \"risk of under-investing exceeds risk of over-investing\" framing that signals best-response spending.\n\nFile v1.0.4:examples/ai-capex-race-among-hyperscalers-2024-2026.md\n\n# Method in Action: The AI-Capex Race Among Hyperscalers (2024–2026)\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\nAcross 2024 and 2025 (and into 2026), the largest U.S. cloud-and-platform companies — Microsoft, Alphabet (Google), Amazon, and Meta — raised their capital expenditure to record levels, the bulk of it directed at AI data centers, accelerators, and power. On successive earnings calls, each firm's leadership publicly framed the risk of *under*-investing in AI capacity as outweighing the risk of *over*-investing. That is a telltale sign of a Nash equilibrium: not four firms independently arriving at the same plan, but four firms each best-responding to what the others are doing. This example runs the case through the skill's six-step Process.\n\n**Step 1 — Specify the game.** *Players:* the major hyperscalers (Microsoft, Alphabet, Amazon, Meta), plus adjacent capacity buyers. *Actions per player:* along a spectrum from \"spend aggressively on AI compute\" to \"hold capex flat and harvest cash.\" *Payoffs:* market position in AI-driven cloud and consumer products, which depends on relative capacity, not just absolute spend. *Information:* largely public — capex guidance is disclosed on quarterly earnings calls, so each player observes rivals' commitments with a short lag. *Timing:* effectively simultaneous and repeated, quarter after quarter, since all players revise guidance on a similar cadence.\n\n**Step 2 — Best-response analysis.** Consider any single firm's choice given the others are spending heavily. If it also spends, it holds its position and shares in whatever AI demand materializes. If it unilaterally pulls back, it frees cash and lifts near-term margins — but it risks ceding compute capacity, model quality, and enterprise-cloud share to rivals who kept building, in a market where capacity is a binding constraint and lead times for chips and power are long. Given rivals spending, the best response is to keep spending. The same logic holds for each player. The mutual-best-response fixed point is \"everyone spends heavily.\"\n\n**Step 3 — Identify all equilibria.** The high-capex profile is a pure-strategy Nash equilibrium: no firm can improve its position by unilaterally cutting while the others build. A low-capex \"everyone restrains\" profile would be more profitable collectively in the short run, but it is *not* an equilibrium — from it, any single firm gains by defecting and out-building the restrained rivals to capture position, so it unravels. This is the structural signature of a prisoner's-dilemma-type game: the cooperative outcome is not self-enforcing. The most plausible focal point is therefore the high-spend equilibrium, which is what the public capex trajectory through this period reflects.\n\n**Step 4 — Evaluate quality.** The equilibrium is not Pareto-optimal for the firms as a group: collectively they would earn higher near-term free cash flow spending less, if all could credibly commit to restraint. The game structure prevents that outcome because commitments are not binding and defection is rewarded by relative position — exactly the mechanism the [`prisoners-dilemma`](../prisoners-dilemma/SKILL.md) skill formalizes. (Note the distributional caveat: what is Pareto-suboptimal for the *spenders* can be highly beneficial for suppliers such as accelerator vendors and for users who receive rapidly cheapening AI capability.)\n\n**Step 5 — Act or redesign.** For an individual hyperscaler, the prescription is to play the equilibrium — keep investing to hold position — while managing the risk that demand disappoints. Redesign options that could relax the arms race are mostly unavailable or illegal: explicit coordination to jointly cut capex would be antitrust-exposed, and no player controls the payoffs. The realistic levers are unilateral efficiency moves that change one's own cost curve without conceding position — custom silicon, better utilization, cheaper power — which shift the payoffs of spending rather than the decision to spend. A firm *outside* the arms race (a smaller player or investor) should instead avoid competing head-on on raw capacity and look for positions the equilibrium leaves open.\n\n**Step 6 — Plan counterparty response.** Once you commit capex, expect rivals to match rather than retreat; guidance raises tend to be met with guidance raises, not with a rival standing down. The contingency that matters is deviation on the *demand* side: if AI monetization underdelivers, the equilibrium can shift toward restraint, but only if enough players cut roughly together — a single firm cutting first bears the position loss alone. So the watch-item is a coordinated signal (softening guidance across multiple players at once), not any one firm's move.\n\nThe mapped steps:\n1. Specify the game: hyperscalers choosing spend levels, public quarterly guidance, repeated and roughly simultaneous, payoff = relative AI/cloud position\n2. Best-response analysis: given rivals spending, each firm's best response is to keep spending; pulling back cedes position\n3. Equilibria: high-capex profile is a pure-strategy Nash equilibrium; the collectively cheaper \"all restrain\" profile is not self-enforcing\n4. Evaluate quality: Pareto-suboptimal for the spenders (a prisoner's-dilemma structure), yet often beneficial to suppliers and users\n5. Act or redesign: play the equilibrium; joint restraint is antitrust-barred, so compete on cost curve instead\n6. Counterparty response: expect matching, not retreat; the real trigger for change is roughly simultaneous demand-side softening across players\n\n*Sources: Company investor disclosures and earnings-call capex guidance, Microsoft, Alphabet, Amazon, and Meta (2024–2026); Nash, J. F. (1951), \"Non-Cooperative Games,\" Annals of Mathematics 54(2), 286–295; Tirole, J. (1988), The Theory of Industrial Organization, MIT Press (oligopoly investment competition).*\n\nFile v1.0.4:examples/nash-1950-51-and-the-fcc-auctions-1994.md\n\n# Method in Action: Nash 1950-51 and the FCC Auctions 1994\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\n**John Forbes Nash Jr.** (1928-2015) wrote his doctoral dissertation at Princeton in 1950, at age 21. The thesis was 28 pages. From it came the 1950 PNAS paper (two pages) and the 1951 *Annals of Mathematics* paper (10 pages). For these two papers — establishing the equilibrium concept that bears his name — Nash received the 1994 Nobel Memorial Prize in Economic Sciences (shared with Reinhard Selten and John Harsanyi for the broader development of non-cooperative game theory).\n\nThe theorem's mathematical proof used the Kakutani fixed-point theorem (Nash 1951). The proof's essence: in the space of mixed strategies, the best-response correspondence is upper-hemicontinuous and the strategy space is compact and convex; therefore by Kakutani, there exists a fixed point — i.e., a strategy combination where every player is best-responding. That fixed point is the equilibrium.\n\nWhat made Nash's contribution revolutionary was not the equilibrium concept itself (which had been informally articulated by Cournot in 1838 for the duopoly case) but the *generality* — the proof that *every* finite n-person game has at least one equilibrium. Before Nash, game theory could analyze only special cases. After Nash, game theory could analyze any finite strategic interaction. This is why Nash's two short papers are considered among the most important in 20th-century mathematics and economics.\n\nThe framework took decades to translate into operational policy. The breakthrough moment was the **1994 FCC spectrum auction**.\n\nThe U.S. Federal Communications Commission had historically allocated radio spectrum via \"comparative hearings\" (essentially bureaucratic beauty contests) and lotteries. Both methods had clear flaws: hearings were corruptible and slow; lotteries gave spectrum to entities who promptly resold it for windfall profits to actual operators. By the early 1990s, the value of spectrum was understood to be enormous and growing (especially for cellular telephony), and pressure was rising for a better mechanism.\n\nPaul Milgrom and Robert Wilson at Stanford were retained as consultants to design the new auction. Their core insight: spectrum auctions are complex strategic games with multiple bidders who hold private valuations for combinations of licenses. The auction rules must be designed so that bidders' Nash-equilibrium strategies — what each bidder rationally does given what every other bidder rationally does — produce efficient allocations (the spectrum ends up with the operators who value it most) and reasonable revenue.\n\nMilgrom and Wilson designed the **simultaneous multiple-round (SMR) auction**: licenses are auctioned simultaneously, in multiple rounds with rising bids, until no bidder is willing to raise any bid. This rule produces a Nash equilibrium in which bidders coordinate on combinations of licenses they want, while preventing bidders from being \"exposed\" (winning licenses individually but not the package they wanted). The design explicitly modeled bidder Nash strategies and chose rules to push the equilibrium toward efficient outcomes.\n\nFrom Milgrom's retrospective:\n\n> \"The SMR design rests on game-theoretic reasoning at every step. The rule structure determines what bidder strategies are available; the strategies determine the Nash equilibria; the equilibria determine the auction outcome. The FCC's 1994 auction raised $617 million for 99 licenses, against pre-auction estimates of $400 million for the prior beauty-contest method. The equilibrium logic worked: bidders coordinated through the rounds on packages of licenses that reflected their actual operational value, prices reflected willingness-to-pay, and the spectrum went to operators who could best use it. The subsequent FCC auctions — 87 of them through 2020 — have raised over $200 billion. The cumulative welfare gain to the U.S. economy is in the trillions. Game theory's policy-impact case rests substantially on this single application.\"\n>\n> — Milgrom (2004). *Putting Auction Theory to Work.* Cambridge University Press, ch. 1.\n\nMilgrom and Wilson received the 2020 Nobel Memorial Prize in Economic Sciences for their auction theory work, with the FCC case as the canonical example.\n\nThe framework has been applied at industrial scale beyond auctions:\n\n**Antitrust analysis.** Modern antitrust evaluation of mergers, vertical restraints, and platform conduct uses Nash-equilibrium modeling to predict how markets will behave under proposed changes. The DOJ and FTC employ economists who specialize in game-theoretic analysis of merger effects.\n\n**Mechanism design.** The broader field that grew from Nash's work — designing the rules of a game to produce desired equilibria — has applications in matching markets (kidney exchange, medical residency placement; Roth 2002, 2007 Nobel work), voting systems, school choice, and platform design.\n\n**Regulatory and policy analysis.** Setting taxes, designing regulations, structuring international trade agreements — all use Nash-equilibrium analysis to predict how strategic actors (firms, countries, voters) will respond.\n\n**Military and security strategy.** Cold War nuclear deterrence theory (Schelling 1960; Schelling 2005 Nobel for related work) is foundational Nash-equilibrium analysis. Modern cybersecurity, terrorism response, and great-power strategic competition all use the framework.\n\n**Corporate strategy.** Competitive interaction (pricing, capacity, R&D, marketing) in oligopolistic markets is Nash-equilibrium analysis. Sophisticated strategy departments at major firms (Walmart, Amazon, ExxonMobil, major banks) employ game theorists.\n\nThree operational lessons:\n\n**First, Nash equilibrium is the right tool when counterparties are rational and react to your moves.** If you don't model their reactions, you'll be systematically surprised by their behavior. The discipline of \"what's the best response to my move?\" prevents many strategic errors.\n\n**Second, equilibrium ≠ optimum.** The prisoner's dilemma is the canonical case: the equilibrium is bad for everyone, but the game's structure prevents the better outcome. Recognizing this is the prerequisite to redesigning the game (e.g., adding repeated interaction, changing payoffs, providing escrow).\n\n**Third, equilibrium-selection is often the operationally hardest problem.** Many games have multiple Nash equilibria, and the framework alone doesn't tell you which will emerge. Empirical, behavioral, and focal-point considerations matter. Treating Nash equilibrium as a complete theory of strategic behavior overstates its predictive power; treating it as one important component of strategic analysis is calibrated.\n\nFile v1.0.4:examples/penalty-kicks-professionals-play-minimax.md\n\n# Method in Action: Penalty Kicks and the Minimax Test (1995-2003)\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\nMixed-strategy Nash equilibrium makes a strange prediction: in games with no stable pure strategy, rational players should randomize — and randomize in exactly the proportions that make their opponent indifferent between responses. For decades this prediction fared poorly in laboratory experiments, where subjects failed to randomize properly. The economist Ignacio Palacios-Huerta asked a sharper question: do *professionals*, playing for real stakes at the top of their craft, play the equilibrium? He tested it on soccer penalty kicks.\n\n**Specify the game (Step 1).** Two players: kicker and goalkeeper. A penalty kick travels to the goal in roughly a third of a second — too fast for the keeper to react to the ball — so both players effectively choose simultaneously: kicker aims to his natural side or the opposite side; keeper dives one way or the other. Payoffs are opposed: the kicker wants to score, the keeper wants to save. Information is public — professional teams scout each other's penalty histories.\n\n**Best-response analysis (Step 2).** No pure strategy survives: if the kicker always shoots to his natural side, the keeper's best response is to always dive there, at which point the kicker's best response flips — and so on, forever. The best-response cycle never settles on a deterministic pair.\n\n**Identify the equilibrium (Step 3).** The game has a unique mixed-strategy Nash equilibrium (the minimax solution, since the game is zero-sum). Each player randomizes with probabilities that make the opponent indifferent. The equilibrium yields two testable predictions: (a) each player's success rate must be *equal* across his own actions — if kicking left scored more often than kicking right, he should kick left more, contradicting equilibrium; (b) each player's sequence of choices must be *serially independent* — any detectable pattern could be exploited.\n\n**Evaluate against data (Step 4).** Palacios-Huerta assembled roughly 1,400 penalty kicks from professional league matches in Spain, Italy, and England (1995-2000), recording kicker, keeper, direction, and outcome. Both predictions held. Scoring probabilities were statistically indistinguishable across a kicker's directions (in the neighborhood of 80% overall), and keepers' save rates were likewise equal across their choices. Player-by-player tests could not reject equality of payoffs for the overwhelming majority of individuals. And unlike laboratory subjects — who notoriously over-alternate when asked to randomize — professionals' direction choices passed standard tests of serial independence. Professionals play minimax.\n\n**Act or redesign (Step 5).** For the players there is nothing to redesign: the equilibrium *is* the optimal policy, and any deviation from it is exploitable. The operational lesson runs the other way — if you face a rival in a genuinely opposed, repeated interaction (competitive bidding, audit scheduling, security patrols), the equilibrium prescribes calibrated unpredictability, and your rival's data on you is the test of whether you achieved it.\n\n**Plan counterparty response (Step 6).** A kicker with a detectable bias hands the keeper a profitable deviation; scouting departments exist precisely to find such biases. Equilibrium play is the only strategy that leaves no pattern for the counterparty to exploit — the contingency plan is built into the randomization itself.\n\nThe mapped steps:\n1. Specify the game: kicker vs. keeper, simultaneous moves, opposed payoffs, public histories\n2. Best-response analysis: no pure strategy is a mutual best response; the cycle never settles\n3. Equilibria: a unique mixed-strategy (minimax) equilibrium with equal payoffs across own actions and serially independent choices\n4. Evaluate quality: field data on ~1,400 professional kicks matched both equilibrium predictions\n5. Act or redesign: equilibrium randomization is the optimal, unexploitable policy\n6. Counterparty response: any bias is scouted and punished; unpredictability is the defense\n\nPrimary source: Palacios-Huerta, I. (2003). \"Professionals Play Minimax.\" *Review of Economic Studies*, 70(2), 395-415.\n\nFile v1.0.4:skill-card.md\n\n## Description: <br>\nGuides agents through Nash equilibrium analysis for strategic situations with rational counterparties, including pricing, negotiation, auctions, M&A, regulation, platform launches, and capacity races. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[deciqai](https://clawhub.ai/user/deciqai) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nEmployees, external users, developers, and business analysts use this skill to model multi-player strategic decisions, identify best responses and equilibria, evaluate whether the equilibrium is desirable, and plan action or game redesign. <br>\n\n### Deployment Geography for Use: <br>\nGlobal <br>\n\n## Known Risks and Mitigations: <br>\nRisk: The skill can shape advice about pricing, auctions, negotiation, M&A, regulation, and other high-stakes strategic decisions. <br>\nMitigation: Use outputs as decision support and obtain appropriate legal, financial, antitrust, or domain review before acting on real business decisions. <br>\nRisk: A Nash model can be misleading when players, actions, payoffs, information structure, or counterparty rationality are specified incorrectly. <br>\nMitigation: Require explicit players, actions, payoffs, information structure, equilibrium checks, and a counterparty response plan before relying on the analysis. <br>\n\n\n## Reference(s): <br>\n- [Sources - nash-equilibrium](references/sources.md) <br>\n- [Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md) <br>\n- [Penalty Kicks and the Minimax Test](examples/penalty-kicks-professionals-play-minimax.md) <br>\n- [The AI-Capex Race Among Hyperscalers (2024-2026)](examples/ai-capex-race-among-hyperscalers-2024-2026.md) <br>\n- [ClawHub skill page](https://clawhub.ai/deciqai/skills/nash-equilibrium) <br>\n\n\n## Skill Output: <br>\n**Output Type(s):** [analysis, markdown, guidance] <br>\n**Output Format:** [Markdown analysis using a structured Nash analysis template] <br>\n**Output Parameters:** [1D] <br>\n**Other Properties Related to Output:** [May ask step-by-step clarification questions in coach mode before producing the final analysis.] <br>\n\n## Skill Version(s): <br>\n1.0.4 (source: server release evidence) <br>\n\n## Ethical Considerations: <br>\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment. <br>\n\nArchive v1.0.3: 6 files, 11619 bytes\n\nFiles: examples/nash-1950-51-and-the-fcc-auctions-1994.md (6810b), examples/penalty-kicks-professionals-play-minimax.md (4283b), references/sources.md (1543b), skill-card.md (2528b), SKILL.md (7430b), _meta.json (135b)\n\nFile v1.0.3:SKILL.md\n\n---\nname: nash-equilibrium\ndescription: \"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep ending up in a bad outcome even though everyone prefers better', or is analyzing a strategic situation with multiple rational counterparties (pricing, negotiation, M&A, regulation, platform launch). Do NOT activate when: the decision is essentially solo with no strategic counterparty; the counterparty is clearly irrational or acting on emotion rather than self-interest.\"\n---\n\n# Nash Equilibrium\n\n## Overview\n\nA **Nash equilibrium** is a stable point in multi-player interaction: a combination of strategies where no player can improve their payoff by unilaterally changing their own strategy, given others hold theirs fixed. Key properties: (1) best-response logic — the equilibrium is a fixed point of mutual best-responses; (2) equilibria can be Pareto-suboptimal (prisoner's dilemma); (3) multiple equilibria are common; (4) equilibrium does not predict the path to get there.\n\nComposes with `prisoners-dilemma`, `repeated-games-reputation`, `signaling-games`, `pricing-strategy`, and `batna-zopa`.\n\n## When to Use\n\n- Designing pricing in a competitive market with rational rivals\n- Negotiating with a sophisticated counterparty; modeling regulatory or legislative outcomes\n- Evaluating M&A or partnership decisions where multiple parties react strategically\n- Designing auctions, marketplaces, or platform mechanisms\n\n**Not when:** the situation is essentially solo; the counterparty is not rational; modeling cost exceeds the decision's value.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line what-it-is: find the point in a multi-player situation where no one has incentive to deviate from their strategy.\n2. Check fit: no rational counterparty? Nash adds less value than other frameworks.\n3. Elicit their real case: who are the players, what can each do, what does each gain or lose?\n> **[WAIT — do not advance until user responds]**\n4. Run The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close by naming the equilibrium and recommended action; if equilibrium is bad, name a redesign option.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Specify the game:** Players · Actions per player · Payoff matrix or game tree · Information structure (full vs. private) · Sequential or simultaneous.\n\n**Step 2 — Best-response analysis:** For each player, find the optimal action given each combination of others' strategies. The combination where everyone is best-responding is a Nash equilibrium.\n\n**Step 3 — Identify all equilibria:** Pure-strategy (deterministic) and mixed-strategy (randomized). If multiple equilibria, identify the most plausible focal point.\n\n**Step 4 — Evaluate quality:** Pareto-optimal? If not, what better collective outcome does the structure prevent?\n\n**Step 5 — Act or redesign:** Play equilibrium strategy if acceptable. If bad: change rules, add repeated interaction, change payoffs, add transparency — or walk away.\n\n**Step 6 — Plan counterparty response:** What do others do once you act? Contingency if they deviate from equilibrium.\n\n## Output Template\n\n```markdown\n# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate\n```\n\n*→ Method in Action: [Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md) · [Penalty Kicks and the Minimax Test](examples/penalty-kicks-professionals-play-minimax.md)*\n\n## Pack: Application Patterns\n\n| Domain | Typical equilibrium | Common pitfall |\n|---|---|---|\n| Pricing in oligopoly | Cournot markup or Bertrand marginal-cost | Assuming price-cuts go unanswered |\n| Auction bidding | Bid to value (English) or shaded bid (sealed) | Ignoring winner's curse in common-value auctions |\n| Platform launch | Multiple equilibria; focal point determines winner | Treating launch as solo, not a coordination game |\n| M&A bidding | Strategic incumbent often overpays | Failing to model how target plays bidders against each other |\n| Cartel formation | Cooperation possible if discount rate is low | Underestimating defection incentives and detection lag |\n\n## Applying It Well\n\nModel the game explicitly before acting. Equilibrium is not optimum — if the equilibrium is bad, redesign the game (add repetition, change payoffs, build trust). When multiple equilibria exist, use focal points or commitment devices to coordinate on the one you want.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We'll just do the right thing; no need for game theory\" | The right thing depends on what others do. Without modeling reactions, you'll be surprised. |\n| [D] \"Game theory is too abstract for real decisions\" | The 1994 FCC auction raised $617M using Nash modeling. Auction design, antitrust, military strategy are operational applications. |\n| [D] \"Counterparties aren't rational; theory doesn't apply\" | They're often more rational than assumed. Use Nash as a baseline and adjust for behavioral deviations. |\n| [D] \"Multiple equilibria mean Nash isn't useful here\" | Multiple equilibria mean selection matters more — use focal points, communication, commitment devices. |\n| [D] \"First mover always wins\" | Often false. Equilibrium analysis tells you when first move helps and when it hurts. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Making a strategic decision without modeling how counterparties will respond\n- Assuming counterparties will react the way you'd react (mirror-imaging)\n- The equilibrium is bad and you haven't considered redesigning the game\n- Playing a repeated game using single-round logic\n- Multiple equilibria exist and you haven't considered which will emerge\n\n## Verification\n\n- [ ] Players, actions, and payoffs explicitly specified\n- [ ] Best-response analysis performed for each player\n- [ ] Nash equilibria identified; Pareto-optimality evaluated\n- [ ] If multiple equilibria, equilibrium-selection considered\n- [ ] If equilibrium is bad, game-redesign options considered\n- [ ] Counterparty response plan exists for equilibrium play and deviation\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 164 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/nash-equilibrium** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\nFile v1.0.3:_meta.json\n\n{\n  \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n  \"slug\": \"nash-equilibrium\",\n  \"version\": \"1.0.3\",\n  \"publishedAt\": 1783509070444\n}\n\nFile v1.0.3:references/sources.md\n\n# Sources — nash-equilibrium\n\n> *Primary sources for the [nash-equilibrium](../SKILL.md) skill.*\n\n- Nash, J. F. (1950). \"Equilibrium Points in N-Person Games.\" *Proceedings of the National Academy of Sciences*, 36(1), 48-49. The foundational two-page proof.\n- Nash, J. F. (1951). \"Non-Cooperative Games.\" *Annals of Mathematics*, 54(2), 286-295. The extended treatment.\n- von Neumann, J. & Morgenstern, O. (1944). *Theory of Games and Economic Behavior.* Princeton University Press. The precursor; covers zero-sum games.\n- Cournot, A. A. (1838). *Recherches sur les principes mathématiques de la théorie des richesses.* Paris: Hachette. The duopoly precursor.\n- Milgrom, P. (2004). *Putting Auction Theory to Work.* Cambridge University Press. ISBN 978-0521536729. The FCC auction case study.\n- Camerer, C. F. (2003). *Behavioral Game Theory: Experiments in Strategic Interaction.* Princeton University Press. ISBN 978-0691090399. Empirical deviations from Nash predictions.\n- Palacios-Huerta, I. (2003). \"Professionals Play Minimax.\" *Review of Economic Studies*, 70(2), 395-415. Field test of mixed-strategy equilibrium on professional penalty kicks.\n- Roth, A. E. (2002). \"The Economist as Engineer.\" *Econometrica*, 70(4), 1341-1378. Mechanism-design applications.\n- Schelling, T. C. (1960). *The Strategy of Conflict.* Harvard University Press. ISBN 978-0674840317. Strategic applications to security.\n- Tirole, J. (1988). *The Theory of Industrial Organization.* MIT Press. ISBN 978-0262200714. Industrial-organization applications.\n\nFile v1.0.3:examples/nash-1950-51-and-the-fcc-auctions-1994.md\n\n# Method in Action: Nash 1950-51 and the FCC Auctions 1994\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\n**John Forbes Nash Jr.** (1928-2015) wrote his doctoral dissertation at Princeton in 1950, at age 21. The thesis was 28 pages. From it came the 1950 PNAS paper (two pages) and the 1951 *Annals of Mathematics* paper (10 pages). For these two papers — establishing the equilibrium concept that bears his name — Nash received the 1994 Nobel Memorial Prize in Economic Sciences (shared with Reinhard Selten and John Harsanyi for the broader development of non-cooperative game theory).\n\nThe theorem's mathematical proof used the Kakutani fixed-point theorem (Nash 1951). The proof's essence: in the space of mixed strategies, the best-response correspondence is upper-hemicontinuous and the strategy space is compact and convex; therefore by Kakutani, there exists a fixed point — i.e., a strategy combination where every player is best-responding. That fixed point is the equilibrium.\n\nWhat made Nash's contribution revolutionary was not the equilibrium concept itself (which had been informally articulated by Cournot in 1838 for the duopoly case) but the *generality* — the proof that *every* finite n-person game has at least one equilibrium. Before Nash, game theory could analyze only special cases. After Nash, game theory could analyze any finite strategic interaction. This is why Nash's two short papers are considered among the most important in 20th-century mathematics and economics.\n\nThe framework took decades to translate into operational policy. The breakthrough moment was the **1994 FCC spectrum auction**.\n\nThe U.S. Federal Communications Commission had historically allocated radio spectrum via \"comparative hearings\" (essentially bureaucratic beauty contests) and lotteries. Both methods had clear flaws: hearings were corruptible and slow; lotteries gave spectrum to entities who promptly resold it for windfall profits to actual operators. By the early 1990s, the value of spectrum was understood to be enormous and growing (especially for cellular telephony), and pressure was rising for a better mechanism.\n\nPaul Milgrom and Robert Wilson at Stanford were retained as consultants to design the new auction. Their core insight: spectrum auctions are complex strategic games with multiple bidders who hold private valuations for combinations of licenses. The auction rules must be designed so that bidders' Nash-equilibrium strategies — what each bidder rationally does given what every other bidder rationally does — produce efficient allocations (the spectrum ends up with the operators who value it most) and reasonable revenue.\n\nMilgrom and Wilson designed the **simultaneous multiple-round (SMR) auction**: licenses are auctioned simultaneously, in multiple rounds with rising bids, until no bidder is willing to raise any bid. This rule produces a Nash equilibrium in which bidders coordinate on combinations of licenses they want, while preventing bidders from being \"exposed\" (winning licenses individually but not the package they wanted). The design explicitly modeled bidder Nash strategies and chose rules to push the equilibrium toward efficient outcomes.\n\nFrom Milgrom's retrospective:\n\n> \"The SMR design rests on game-theoretic reasoning at every step. The rule structure determines what bidder strategies are available; the strategies determine the Nash equilibria; the equilibria determine the auction outcome. The FCC's 1994 auction raised $617 million for 99 licenses, against pre-auction estimates of $400 million for the prior beauty-contest method. The equilibrium logic worked: bidders coordinated through the rounds on packages of licenses that reflected their actual operational value, prices reflected willingness-to-pay, and the spectrum went to operators who could best use it. The subsequent FCC auctions — 87 of them through 2020 — have raised over $200 billion. The cumulative welfare gain to the U.S. economy is in the trillions. Game theory's policy-impact case rests substantially on this single application.\"\n>\n> — Milgrom (2004). *Putting Auction Theory to Work.* Cambridge University Press, ch. 1.\n\nMilgrom and Wilson received the 2020 Nobel Memorial Prize in Economic Sciences for their auction theory work, with the FCC case as the canonical example.\n\nThe framework has been applied at industrial scale beyond auctions:\n\n**Antitrust analysis.** Modern antitrust evaluation of mergers, vertical restraints, and platform conduct uses Nash-equilibrium modeling to predict how markets will behave under proposed changes. The DOJ and FTC employ economists who specialize in game-theoretic analysis of merger effects.\n\n**Mechanism design.** The broader field that grew from Nash's work — designing the rules of a game to produce desired equilibria — has applications in matching markets (kidney exchange, medical residency placement; Roth 2002, 2007 Nobel work), voting systems, school choice, and platform design.\n\n**Regulatory and policy analysis.** Setting taxes, designing regulations, structuring international trade agreements — all use Nash-equilibrium analysis to predict how strategic actors (firms, countries, voters) will respond.\n\n**Military and security strategy.** Cold War nuclear deterrence theory (Schelling 1960; Schelling 2005 Nobel for related work) is foundational Nash-equilibrium analysis. Modern cybersecurity, terrorism response, and great-power strategic competition all use the framework.\n\n**Corporate strategy.** Competitive interaction (pricing, capacity, R&D, marketing) in oligopolistic markets is Nash-equilibrium analysis. Sophisticated strategy departments at major firms (Walmart, Amazon, ExxonMobil, major banks) employ game theorists.\n\nThree operational lessons:\n\n**First, Nash equilibrium is the right tool when counterparties are rational and react to your moves.** If you don't model their reactions, you'll be systematically surprised by their behavior. The discipline of \"what's the best response to my move?\" prevents many strategic errors.\n\n**Second, equilibrium ≠ optimum.** The prisoner's dilemma is the canonical case: the equilibrium is bad for everyone, but the game's structure prevents the better outcome. Recognizing this is the prerequisite to redesigning the game (e.g., adding repeated interaction, changing payoffs, providing escrow).\n\n**Third, equilibrium-selection is often the operationally hardest problem.** Many games have multiple Nash equilibria, and the framework alone doesn't tell you which will emerge. Empirical, behavioral, and focal-point considerations matter. Treating Nash equilibrium as a complete theory of strategic behavior overstates its predictive power; treating it as one important component of strategic analysis is calibrated.\n\nFile v1.0.3:examples/penalty-kicks-professionals-play-minimax.md\n\n# Method in Action: Penalty Kicks and the Minimax Test (1995-2003)\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\nMixed-strategy Nash equilibrium makes a strange prediction: in games with no stable pure strategy, rational players should randomize — and randomize in exactly the proportions that make their opponent indifferent between responses. For decades this prediction fared poorly in laboratory experiments, where subjects failed to randomize properly. The economist Ignacio Palacios-Huerta asked a sharper question: do *professionals*, playing for real stakes at the top of their craft, play the equilibrium? He tested it on soccer penalty kicks.\n\n**Specify the game (Step 1).** Two players: kicker and goalkeeper. A penalty kick travels to the goal in roughly a third of a second — too fast for the keeper to react to the ball — so both players effectively choose simultaneously: kicker aims to his natural side or the opposite side; keeper dives one way or the other. Payoffs are opposed: the kicker wants to score, the keeper wants to save. Information is public — professional teams scout each other's penalty histories.\n\n**Best-response analysis (Step 2).** No pure strategy survives: if the kicker always shoots to his natural side, the keeper's best response is to always dive there, at which point the kicker's best response flips — and so on, forever. The best-response cycle never settles on a deterministic pair.\n\n**Identify the equilibrium (Step 3).** The game has a unique mixed-strategy Nash equilibrium (the minimax solution, since the game is zero-sum). Each player randomizes with probabilities that make the opponent indifferent. The equilibrium yields two testable predictions: (a) each player's success rate must be *equal* across his own actions — if kicking left scored more often than kicking right, he should kick left more, contradicting equilibrium; (b) each player's sequence of choices must be *serially independent* — any detectable pattern could be exploited.\n\n**Evaluate against data (Step 4).** Palacios-Huerta assembled roughly 1,400 penalty kicks from professional league matches in Spain, Italy, and England (1995-2000), recording kicker, keeper, direction, and outcome. Both predictions held. Scoring probabilities were statistically indistinguishable across a kicker's directions (in the neighborhood of 80% overall), and keepers' save rates were likewise equal across their choices. Player-by-player tests could not reject equality of payoffs for the overwhelming majority of individuals. And unlike laboratory subjects — who notoriously over-alternate when asked to randomize — professionals' direction choices passed standard tests of serial independence. Professionals play minimax.\n\n**Act or redesign (Step 5).** For the players there is nothing to redesign: the equilibrium *is* the optimal policy, and any deviation from it is exploitable. The operational lesson runs the other way — if you face a rival in a genuinely opposed, repeated interaction (competitive bidding, audit scheduling, security patrols), the equilibrium prescribes calibrated unpredictability, and your rival's data on you is the test of whether you achieved it.\n\n**Plan counterparty response (Step 6).** A kicker with a detectable bias hands the keeper a profitable deviation; scouting departments exist precisely to find such biases. Equilibrium play is the only strategy that leaves no pattern for the counterparty to exploit — the contingency plan is built into the randomization itself.\n\nThe mapped steps:\n1. Specify the game: kicker vs. keeper, simultaneous moves, opposed payoffs, public histories\n2. Best-response analysis: no pure strategy is a mutual best response; the cycle never settles\n3. Equilibria: a unique mixed-strategy (minimax) equilibrium with equal payoffs across own actions and serially independent choices\n4. Evaluate quality: field data on ~1,400 professional kicks matched both equilibrium predictions\n5. Act or redesign: equilibrium randomization is the optimal, unexploitable policy\n6. Counterparty response: any bias is scouted and punished; unpredictability is the defense\n\nPrimary source: Palacios-Huerta, I. (2003). \"Professionals Play Minimax.\" *Review of Economic Studies*, 70(2), 395-415.\n\nFile v1.0.3:skill-card.md\n\n## Description: <br>\nHelps agents apply Nash equilibrium analysis to strategic situations involving rational counterparties, including pricing, negotiation, M&A, regulation, platform launches, and auction or mechanism design. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[deciqai](https://clawhub.ai/user/deciqai) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nDevelopers, operators, and strategy teams use this skill to structure multi-player decisions, identify pure or mixed equilibria, evaluate whether outcomes are Pareto-optimal, and plan whether to play the equilibrium or redesign the game. <br>\n\n### Deployment Geography for Use: <br>\nGlobal <br>\n\n## Known Risks and Mitigations: <br>\nRisk: The skill provides advisory decision-analysis guidance and does not guarantee correct business strategy. <br>\nMitigation: Review the modeled players, actions, payoffs, and assumptions before relying on the recommendation. <br>\nRisk: Generated analysis may be misleading if the situation lacks rational counterparties or if strategic assumptions are incomplete. <br>\nMitigation: Use the skill only when the decision involves strategic counterparties, and validate conclusions against domain expertise and current evidence. <br>\nRisk: The artifact includes external source and promotional links. <br>\nMitigation: Treat links as ordinary Markdown references and review them under local browsing and source-trust policies before use. <br>\n\n\n## Reference(s): <br>\n- [Nash Equilibrium on ClawHub](https://clawhub.ai/deciqai/skills/nash-equilibrium) <br>\n- [Primary Sources](references/sources.md) <br>\n- [Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md) <br>\n- [Penalty Kicks and the Minimax Test](examples/penalty-kicks-professionals-play-minimax.md) <br>\n\n\n## Skill Output: <br>\n**Output Type(s):** [Guidance, Analysis, Markdown] <br>\n**Output Format:** [Markdown analysis with structured sections and checklists] <br>\n**Output Parameters:** [1D] <br>\n**Other Properties Related to Output:** [No code execution, shell commands, or configuration changes are produced by the skill.] <br>\n\n## Skill Version(s): <br>\n1.0.3 (source: release evidence) <br>\n\n## Ethical Considerations: <br>\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment. <br>\n\nArchive v1.0.2: 6 files, 11591 bytes\n\nFiles: examples/nash-1950-51-and-the-fcc-auctions-1994.md (6810b), examples/penalty-kicks-professionals-play-minimax.md (4283b), references/sources.md (1543b), skill-card.md (2381b), SKILL.md (7536b), _meta.json (135b)\n\nFile v1.0.2:SKILL.md\n\n---\nname: nash-equilibrium\ndescription: \"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep ending up in a bad outcome even though everyone prefers better', or is analyzing a strategic situation with multiple rational counterparties (pricing, negotiation, M&A, regulation, platform launch). Do NOT activate when: the decision is essentially solo with no strategic counterparty; the counterparty is clearly irrational or acting on emotion rather than self-interest.\"\n---\n\n# Nash Equilibrium\n\n## Overview\n\nA **Nash equilibrium** is a stable point in multi-player interaction: a combination of strategies where no player can improve their payoff by unilaterally changing their own strategy, given others hold theirs fixed. Key properties: (1) best-response logic — the equilibrium is a fixed point of mutual best-responses; (2) equilibria can be Pareto-suboptimal (prisoner's dilemma); (3) multiple equilibria are common; (4) equilibrium does not predict the path to get there.\n\nComposes with `prisoners-dilemma`, `repeated-games-reputation`, `signaling-games`, `pricing-strategy`, and `batna-zopa`.\n\n## When to Use\n\n- Designing pricing in a competitive market with rational rivals\n- Negotiating with a sophisticated counterparty; modeling regulatory or legislative outcomes\n- Evaluating M&A or partnership decisions where multiple parties react strategically\n- Designing auctions, marketplaces, or platform mechanisms\n\n**Not when:** the situation is essentially solo; the counterparty is not rational; modeling cost exceeds the decision's value.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line what-it-is: find the point in a multi-player situation where no one has incentive to deviate from their strategy.\n2. Check fit: no rational counterparty? Nash adds less value than other frameworks.\n3. Elicit their real case: who are the players, what can each do, what does each gain or lose?\n> **[WAIT — do not advance until user responds]**\n4. Run The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close by naming the equilibrium and recommended action; if equilibrium is bad, name a redesign option.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Specify the game:** Players · Actions per player · Payoff matrix or game tree · Information structure (full vs. private) · Sequential or simultaneous.\n\n**Step 2 — Best-response analysis:** For each player, find the optimal action given each combination of others' strategies. The combination where everyone is best-responding is a Nash equilibrium.\n\n**Step 3 — Identify all equilibria:** Pure-strategy (deterministic) and mixed-strategy (randomized). If multiple equilibria, identify the most plausible focal point.\n\n**Step 4 — Evaluate quality:** Pareto-optimal? If not, what better collective outcome does the structure prevent?\n\n**Step 5 — Act or redesign:** Play equilibrium strategy if acceptable. If bad: change rules, add repeated interaction, change payoffs, add transparency — or walk away.\n\n**Step 6 — Plan counterparty response:** What do others do once you act? Contingency if they deviate from equilibrium.\n\n## Output Template\n\n```markdown\n# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate\n```\n\n*→ Method in Action: [Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md) · [Penalty Kicks and the Minimax Test](examples/penalty-kicks-professionals-play-minimax.md)*\n\n## Pack: Application Patterns\n\n| Domain | Typical equilibrium | Common pitfall |\n|---|---|---|\n| Pricing in oligopoly | Cournot markup or Bertrand marginal-cost | Assuming price-cuts go unanswered |\n| Auction bidding | Bid to value (English) or shaded bid (sealed) | Ignoring winner's curse in common-value auctions |\n| Platform launch | Multiple equilibria; focal point determines winner | Treating launch as solo, not a coordination game |\n| M&A bidding | Strategic incumbent often overpays | Failing to model how target plays bidders against each other |\n| Cartel formation | Cooperation possible if discount rate is low | Underestimating defection incentives and detection lag |\n\n## Applying It Well\n\nModel the game explicitly before acting. Equilibrium is not optimum — if the equilibrium is bad, redesign the game (add repetition, change payoffs, build trust). When multiple equilibria exist, use focal points or commitment devices to coordinate on the one you want.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We'll just do the right thing; no need for game theory\" | The right thing depends on what others do. Without modeling reactions, you'll be surprised. |\n| [D] \"Game theory is too abstract for real decisions\" | The 1994 FCC auction raised $617M using Nash modeling. Auction design, antitrust, military strategy are operational applications. |\n| [D] \"Counterparties aren't rational; theory doesn't apply\" | They're often more rational than assumed. Use Nash as a baseline and adjust for behavioral deviations. |\n| [D] \"Multiple equilibria mean Nash isn't useful here\" | Multiple equilibria mean selection matters more — use focal points, communication, commitment devices. |\n| [D] \"First mover always wins\" | Often false. Equilibrium analysis tells you when first move helps and when it hurts. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Making a strategic decision without modeling how counterparties will respond\n- Assuming counterparties will react the way you'd react (mirror-imaging)\n- The equilibrium is bad and you haven't considered redesigning the game\n- Playing a repeated game using single-round logic\n- Multiple equilibria exist and you haven't considered which will emerge\n\n## Verification\n\n- [ ] Players, actions, and payoffs explicitly specified\n- [ ] Best-response analysis performed for each player\n- [ ] Nash equilibria identified; Pareto-optimality evaluated\n- [ ] If multiple equilibria, equilibrium-selection considered\n- [ ] If equilibrium is bad, game-redesign options considered\n- [ ] Counterparty response plan exists for equilibrium play and deviation\n\n---\n\n*Part of **deciqAI Knowledge Skills** — 163 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/skills/nash-equilibrium?utm_source=clawhub&utm_medium=marketplace&utm_campaign=knowledge-skills&utm_content=nash-equilibrium** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*\n\nFile v1.0.2:_meta.json\n\n{\n  \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n  \"slug\": \"nash-equilibrium\",\n  \"version\": \"1.0.2\",\n  \"publishedAt\": 1783472165502\n}\n\nFile v1.0.2:references/sources.md\n\n# Sources — nash-equilibrium\n\n> *Primary sources for the [nash-equilibrium](../SKILL.md) skill.*\n\n- Nash, J. F. (1950). \"Equilibrium Points in N-Person Games.\" *Proceedings of the National Academy of Sciences*, 36(1), 48-49. The foundational two-page proof.\n- Nash, J. F. (1951). \"Non-Cooperative Games.\" *Annals of Mathematics*, 54(2), 286-295. The extended treatment.\n- von Neumann, J. & Morgenstern, O. (1944). *Theory of Games and Economic Behavior.* Princeton University Press. The precursor; covers zero-sum games.\n- Cournot, A. A. (1838). *Recherches sur les principes mathématiques de la théorie des richesses.* Paris: Hachette. The duopoly precursor.\n- Milgrom, P. (2004). *Putting Auction Theory to Work.* Cambridge University Press. ISBN 978-0521536729. The FCC auction case study.\n- Camerer, C. F. (2003). *Behavioral Game Theory: Experiments in Strategic Interaction.* Princeton University Press. ISBN 978-0691090399. Empirical deviations from Nash predictions.\n- Palacios-Huerta, I. (2003). \"Professionals Play Minimax.\" *Review of Economic Studies*, 70(2), 395-415. Field test of mixed-strategy equilibrium on professional penalty kicks.\n- Roth, A. E. (2002). \"The Economist as Engineer.\" *Econometrica*, 70(4), 1341-1378. Mechanism-design applications.\n- Schelling, T. C. (1960). *The Strategy of Conflict.* Harvard University Press. ISBN 978-0674840317. Strategic applications to security.\n- Tirole, J. (1988). *The Theory of Industrial Organization.* MIT Press. ISBN 978-0262200714. Industrial-organization applications.\n\nFile v1.0.2:examples/nash-1950-51-and-the-fcc-auctions-1994.md\n\n# Method in Action: Nash 1950-51 and the FCC Auctions 1994\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\n**John Forbes Nash Jr.** (1928-2015) wrote his doctoral dissertation at Princeton in 1950, at age 21. The thesis was 28 pages. From it came the 1950 PNAS paper (two pages) and the 1951 *Annals of Mathematics* paper (10 pages). For these two papers — establishing the equilibrium concept that bears his name — Nash received the 1994 Nobel Memorial Prize in Economic Sciences (shared with Reinhard Selten and John Harsanyi for the broader development of non-cooperative game theory).\n\nThe theorem's mathematical proof used the Kakutani fixed-point theorem (Nash 1951). The proof's essence: in the space of mixed strategies, the best-response correspondence is upper-hemicontinuous and the strategy space is compact and convex; therefore by Kakutani, there exists a fixed point — i.e., a strategy combination where every player is best-responding. That fixed point is the equilibrium.\n\nWhat made Nash's contribution revolutionary was not the equilibrium concept itself (which had been informally articulated by Cournot in 1838 for the duopoly case) but the *generality* — the proof that *every* finite n-person game has at least one equilibrium. Before Nash, game theory could analyze only special cases. After Nash, game theory could analyze any finite strategic interaction. This is why Nash's two short papers are considered among the most important in 20th-century mathematics and economics.\n\nThe framework took decades to translate into operational policy. The breakthrough moment was the **1994 FCC spectrum auction**.\n\nThe U.S. Federal Communications Commission had historically allocated radio spectrum via \"comparative hearings\" (essentially bureaucratic beauty contests) and lotteries. Both methods had clear flaws: hearings were corruptible and slow; lotteries gave spectrum to entities who promptly resold it for windfall profits to actual operators. By the early 1990s, the value of spectrum was understood to be enormous and growing (especially for cellular telephony), and pressure was rising for a better mechanism.\n\nPaul Milgrom and Robert Wilson at Stanford were retained as consultants to design the new auction. Their core insight: spectrum auctions are complex strategic games with multiple bidders who hold private valuations for combinations of licenses. The auction rules must be designed so that bidders' Nash-equilibrium strategies — what each bidder rationally does given what every other bidder rationally does — produce efficient allocations (the spectrum ends up with the operators who value it most) and reasonable revenue.\n\nMilgrom and Wilson designed the **simultaneous multiple-round (SMR) auction**: licenses are auctioned simultaneously, in multiple rounds with rising bids, until no bidder is willing to raise any bid. This rule produces a Nash equilibrium in which bidders coordinate on combinations of licenses they want, while preventing bidders from being \"exposed\" (winning licenses individually but not the package they wanted). The design explicitly modeled bidder Nash strategies and chose rules to push the equilibrium toward efficient outcomes.\n\nFrom Milgrom's retrospective:\n\n> \"The SMR design rests on game-theoretic reasoning at every step. The rule structure determines what bidder strategies are available; the strategies determine the Nash equilibria; the equilibria determine the auction outcome. The FCC's 1994 auction raised $617 million for 99 licenses, against pre-auction estimates of $400 million for the prior beauty-contest method. The equilibrium logic worked: bidders coordinated through the rounds on packages of licenses that reflected their actual operational value, prices reflected willingness-to-pay, and the spectrum went to operators who could best use it. The subsequent FCC auctions — 87 of them through 2020 — have raised over $200 billion. The cumulative welfare gain to the U.S. economy is in the trillions. Game theory's policy-impact case rests substantially on this single application.\"\n>\n> — Milgrom (2004). *Putting Auction Theory to Work.* Cambridge University Press, ch. 1.\n\nMilgrom and Wilson received the 2020 Nobel Memorial Prize in Economic Sciences for their auction theory work, with the FCC case as the canonical example.\n\nThe framework has been applied at industrial scale beyond auctions:\n\n**Antitrust analysis.** Modern antitrust evaluation of mergers, vertical restraints, and platform conduct uses Nash-equilibrium modeling to predict how markets will behave under proposed changes. The DOJ and FTC employ economists who specialize in game-theoretic analysis of merger effects.\n\n**Mechanism design.** The broader field that grew from Nash's work — designing the rules of a game to produce desired equilibria — has applications in matching markets (kidney exchange, medical residency placement; Roth 2002, 2007 Nobel work), voting systems, school choice, and platform design.\n\n**Regulatory and policy analysis.** Setting taxes, designing regulations, structuring international trade agreements — all use Nash-equilibrium analysis to predict how strategic actors (firms, countries, voters) will respond.\n\n**Military and security strategy.** Cold War nuclear deterrence theory (Schelling 1960; Schelling 2005 Nobel for related work) is foundational Nash-equilibrium analysis. Modern cybersecurity, terrorism response, and great-power strategic competition all use the framework.\n\n**Corporate strategy.** Competitive interaction (pricing, capacity, R&D, marketing) in oligopolistic markets is Nash-equilibrium analysis. Sophisticated strategy departments at major firms (Walmart, Amazon, ExxonMobil, major banks) employ game theorists.\n\nThree operational lessons:\n\n**First, Nash equilibrium is the right tool when counterparties are rational and react to your moves.** If you don't model their reactions, you'll be systematically surprised by their behavior. The discipline of \"what's the best response to my move?\" prevents many strategic errors.\n\n**Second, equilibrium ≠ optimum.** The prisoner's dilemma is the canonical case: the equilibrium is bad for everyone, but the game's structure prevents the better outcome. Recognizing this is the prerequisite to redesigning the game (e.g., adding repeated interaction, changing payoffs, providing escrow).\n\n**Third, equilibrium-selection is often the operationally hardest problem.** Many games have multiple Nash equilibria, and the framework alone doesn't tell you which will emerge. Empirical, behavioral, and focal-point considerations matter. Treating Nash equilibrium as a complete theory of strategic behavior overstates its predictive power; treating it as one important component of strategic analysis is calibrated.\n\nFile v1.0.2:examples/penalty-kicks-professionals-play-minimax.md\n\n# Method in Action: Penalty Kicks and the Minimax Test (1995-2003)\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\nMixed-strategy Nash equilibrium makes a strange prediction: in games with no stable pure strategy, rational players should randomize — and randomize in exactly the proportions that make their opponent indifferent between responses. For decades this prediction fared poorly in laboratory experiments, where subjects failed to randomize properly. The economist Ignacio Palacios-Huerta asked a sharper question: do *professionals*, playing for real stakes at the top of their craft, play the equilibrium? He tested it on soccer penalty kicks.\n\n**Specify the game (Step 1).** Two players: kicker and goalkeeper. A penalty kick travels to the goal in roughly a third of a second — too fast for the keeper to react to the ball — so both players effectively choose simultaneously: kicker aims to his natural side or the opposite side; keeper dives one way or the other. Payoffs are opposed: the kicker wants to score, the keeper wants to save. Information is public — professional teams scout each other's penalty histories.\n\n**Best-response analysis (Step 2).** No pure strategy survives: if the kicker always shoots to his natural side, the keeper's best response is to always dive there, at which point the kicker's best response flips — and so on, forever. The best-response cycle never settles on a deterministic pair.\n\n**Identify the equilibrium (Step 3).** The game has a unique mixed-strategy Nash equilibrium (the minimax solution, since the game is zero-sum). Each player randomizes with probabilities that make the opponent indifferent. The equilibrium yields two testable predictions: (a) each player's success rate must be *equal* across his own actions — if kicking left scored more often than kicking right, he should kick left more, contradicting equilibrium; (b) each player's sequence of choices must be *serially independent* — any detectable pattern could be exploited.\n\n**Evaluate against data (Step 4).** Palacios-Huerta assembled roughly 1,400 penalty kicks from professional league matches in Spain, Italy, and England (1995-2000), recording kicker, keeper, direction, and outcome. Both predictions held. Scoring probabilities were statistically indistinguishable across a kicker's directions (in the neighborhood of 80% overall), and keepers' save rates were likewise equal across their choices. Player-by-player tests could not reject equality of payoffs for the overwhelming majority of individuals. And unlike laboratory subjects — who notoriously over-alternate when asked to randomize — professionals' direction choices passed standard tests of serial independence. Professionals play minimax.\n\n**Act or redesign (Step 5).** For the players there is nothing to redesign: the equilibrium *is* the optimal policy, and any deviation from it is exploitable. The operational lesson runs the other way — if you face a rival in a genuinely opposed, repeated interaction (competitive bidding, audit scheduling, security patrols), the equilibrium prescribes calibrated unpredictability, and your rival's data on you is the test of whether you achieved it.\n\n**Plan counterparty response (Step 6).** A kicker with a detectable bias hands the keeper a profitable deviation; scouting departments exist precisely to find such biases. Equilibrium play is the only strategy that leaves no pattern for the counterparty to exploit — the contingency plan is built into the randomization itself.\n\nThe mapped steps:\n1. Specify the game: kicker vs. keeper, simultaneous moves, opposed payoffs, public histories\n2. Best-response analysis: no pure strategy is a mutual best response; the cycle never settles\n3. Equilibria: a unique mixed-strategy (minimax) equilibrium with equal payoffs across own actions and serially independent choices\n4. Evaluate quality: field data on ~1,400 professional kicks matched both equilibrium predictions\n5. Act or redesign: equilibrium randomization is the optimal, unexploitable policy\n6. Counterparty response: any bias is scouted and punished; unpredictability is the defense\n\nPrimary source: Palacios-Huerta, I. (2003). \"Professionals Play Minimax.\" *Review of Economic Studies*, 70(2), 395-415.\n\nFile v1.0.2:skill-card.md\n\n## Description: <br>\nGuides agents through Nash-equilibrium analysis for strategic situations with rational counterparties, including pricing, negotiation, regulation, auctions, marketplaces, and platform decisions. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[deciqai](https://clawhub.ai/user/deciqai) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nExternal users, employees, and developers use this skill to model multi-party strategic decisions, identify equilibrium strategies, evaluate whether outcomes are Pareto-optimal, and decide whether to act within or redesign the game. <br>\n\n### Deployment Geography for Use: <br>\nGlobal <br>\n\n## Known Risks and Mitigations: <br>\nRisk: Strategic recommendations can be misleading if players, actions, payoffs, information structure, or rationality assumptions are wrong. <br>\nMitigation: Use the skill's verification checklist to explicitly validate players, actions, payoffs, best responses, equilibria, Pareto quality, and counterparty response plans before acting. <br>\nRisk: The skill may add little value when there is no strategic counterparty, the counterparty is not acting rationally, or the modeling effort exceeds the decision value. <br>\nMitigation: Apply the fit check before analysis and choose a different framework when the situation is solo, primarily emotional, or too low-stakes for game modeling. <br>\n\n\n## Reference(s): <br>\n- [Sources - nash-equilibrium](references/sources.md) <br>\n- [Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md) <br>\n- [Penalty Kicks and the Minimax Test](examples/penalty-kicks-professionals-play-minimax.md) <br>\n\n\n## Skill Output: <br>\n**Output Type(s):** [Text, Markdown, Guidance] <br>\n**Output Format:** [Markdown analysis using the Nash Analysis template] <br>\n**Output Parameters:** [1D] <br>\n**Other Properties Related to Output:** [May include staged coaching questions when the user is unfamiliar with the framework.] <br>\n\n## Skill Version(s): <br>\n1.0.2 (source: server release evidence) <br>\n\n## Ethical Considerations: <br>\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment. <br>\n\nArchive v1.0.1: 5 files, 9160 bytes\n\nFiles: examples/nash-1950-51-and-the-fcc-auctions-1994.md (6810b), references/sources.md (1365b), skill-card.md (2390b), SKILL.md (7345b), _meta.json (135b)\n\nFile v1.0.1:SKILL.md\n\n---\nname: nash-equilibrium\ndescription: \"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep ending up in a bad outcome even though everyone prefers better', or is analyzing a strategic situation with multiple rational counterparties (pricing, negotiation, M&A, regulation, platform launch). Do NOT activate when: the decision is essentially solo with no strategic counterparty; the counterparty is clearly irrational or acting on emotion rather than self-interest.\"\n---\n\n# Nash Equilibrium\n\n## Overview\n\nA **Nash equilibrium** is a stable point in multi-player interaction: a combination of strategies where no player can improve their payoff by unilaterally changing their own strategy, given others hold theirs fixed. Key properties: (1) best-response logic — the equilibrium is a fixed point of mutual best-responses; (2) equilibria can be Pareto-suboptimal (prisoner's dilemma); (3) multiple equilibria are common; (4) equilibrium does not predict the path to get there.\n\nComposes with [`prisoners-dilemma`](../prisoners-dilemma/SKILL.md), [`repeated-games-reputation`](../repeated-games-reputation/SKILL.md), [`signaling-games`](../signaling-games/SKILL.md), [`pricing-strategy`](../pricing-strategy/SKILL.md), and [`batna-zopa`](../batna-zopa/SKILL.md).\n\n## When to Use\n\n- Designing pricing in a competitive market with rational rivals\n- Negotiating with a sophisticated counterparty; modeling regulatory or legislative outcomes\n- Evaluating M&A or partnership decisions where multiple parties react strategically\n- Designing auctions, marketplaces, or platform mechanisms\n\n**Not when:** the situation is essentially solo; the counterparty is not rational; modeling cost exceeds the decision's value.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line what-it-is: find the point in a multi-player situation where no one has incentive to deviate from their strategy.\n2. Check fit: no rational counterparty? Nash adds less value than other frameworks.\n3. Elicit their real case: who are the players, what can each do, what does each gain or lose?\n> **[WAIT — do not advance until user responds]**\n4. Run The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close by naming the equilibrium and recommended action; if equilibrium is bad, name a redesign option.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Specify the game:** Players · Actions per player · Payoff matrix or game tree · Information structure (full vs. private) · Sequential or simultaneous.\n\n**Step 2 — Best-response analysis:** For each player, find the optimal action given each combination of others' strategies. The combination where everyone is best-responding is a Nash equilibrium.\n\n**Step 3 — Identify all equilibria:** Pure-strategy (deterministic) and mixed-strategy (randomized). If multiple equilibria, identify the most plausible focal point.\n\n**Step 4 — Evaluate quality:** Pareto-optimal? If not, what better collective outcome does the structure prevent?\n\n**Step 5 — Act or redesign:** Play equilibrium strategy if acceptable. If bad: change rules, add repeated interaction, change payoffs, add transparency — or walk away.\n\n**Step 6 — Plan counterparty response:** What do others do once you act? Contingency if they deviate from equilibrium.\n\n## Output Template\n\n```markdown\n# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate\n```\n\n*→ Method in Action: [Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md)*\n\n## Pack: Application Patterns\n\n| Domain | Typical equilibrium | Common pitfall |\n|---|---|---|\n| Pricing in oligopoly | Cournot markup or Bertrand marginal-cost | Assuming price-cuts go unanswered |\n| Auction bidding | Bid to value (English) or shaded bid (sealed) | Ignoring winner's curse in common-value auctions |\n| Platform launch | Multiple equilibria; focal point determines winner | Treating launch as solo, not a coordination game |\n| M&A bidding | Strategic incumbent often overpays | Failing to model how target plays bidders against each other |\n| Cartel formation | Cooperation possible if discount rate is low | Underestimating defection incentives and detection lag |\n\n## Applying It Well\n\nModel the game explicitly before acting. Equilibrium is not optimum — if the equilibrium is bad, redesign the game (add repetition, change payoffs, build trust). When multiple equilibria exist, use focal points or commitment devices to coordinate on the one you want.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We'll just do the right thing; no need for game theory\" | The right thing depends on what others do. Without modeling reactions, you'll be surprised. |\n| [D] \"Game theory is too abstract for real decisions\" | The 1994 FCC auction raised $617M using Nash modeling. Auction design, antitrust, military strategy are operational applications. |\n| [D] \"Counterparties aren't rational; theory doesn't apply\" | They're often more rational than assumed. Use Nash as a baseline and adjust for behavioral deviations. |\n| [D] \"Multiple equilibria mean Nash isn't useful here\" | Multiple equilibria mean selection matters more — use focal points, communication, commitment devices. |\n| [D] \"First mover always wins\" | Often false. Equilibrium analysis tells you when first move helps and when it hurts. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Making a strategic decision without modeling how counterparties will respond\n- Assuming counterparties will react the way you'd react (mirror-imaging)\n- The equilibrium is bad and you haven't considered redesigning the game\n- Playing a repeated game using single-round logic\n- Multiple equilibria exist and you haven't considered which will emerge\n\n## Verification\n\n- [ ] Players, actions, and payoffs explicitly specified\n- [ ] Best-response analysis performed for each player\n- [ ] Nash equilibria identified; Pareto-optimality evaluated\n- [ ] If multiple equilibria, equilibrium-selection considered\n- [ ] If equilibrium is bad, game-redesign options considered\n- [ ] Counterparty response plan exists for equilibrium play and deviation\n\n---\n\n*Part of **deciqAI Knowledge Skills** — open-source thinking skills that make rigor executable for AI agents. Built by deciqAI · https://deciqai.com · Contributions welcome — see the template at the repo root.*\n\nFile v1.0.1:_meta.json\n\n{\n  \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n  \"slug\": \"nash-equilibrium\",\n  \"version\": \"1.0.1\",\n  \"publishedAt\": 1783463399671\n}\n\nFile v1.0.1:references/sources.md\n\n# Sources — nash-equilibrium\n\n> *Primary sources for the [nash-equilibrium](../SKILL.md) skill.*\n\n- Nash, J. F. (1950). \"Equilibrium Points in N-Person Games.\" *Proceedings of the National Academy of Sciences*, 36(1), 48-49. The foundational two-page proof.\n- Nash, J. F. (1951). \"Non-Cooperative Games.\" *Annals of Mathematics*, 54(2), 286-295. The extended treatment.\n- von Neumann, J. & Morgenstern, O. (1944). *Theory of Games and Economic Behavior.* Princeton University Press. The precursor; covers zero-sum games.\n- Cournot, A. A. (1838). *Recherches sur les principes mathématiques de la théorie des richesses.* Paris: Hachette. The duopoly precursor.\n- Milgrom, P. (2004). *Putting Auction Theory to Work.* Cambridge University Press. ISBN 978-0521536729. The FCC auction case study.\n- Camerer, C. F. (2003). *Behavioral Game Theory: Experiments in Strategic Interaction.* Princeton University Press. ISBN 978-0691090399. Empirical deviations from Nash predictions.\n- Roth, A. E. (2002). \"The Economist as Engineer.\" *Econometrica*, 70(4), 1341-1378. Mechanism-design applications.\n- Schelling, T. C. (1960). *The Strategy of Conflict.* Harvard University Press. ISBN 978-0674840317. Strategic applications to security.\n- Tirole, J. (1988). *The Theory of Industrial Organization.* MIT Press. ISBN 978-0262200714. Industrial-organization applications.\n\nFile v1.0.1:examples/nash-1950-51-and-the-fcc-auctions-1994.md\n\n# Method in Action: Nash 1950-51 and the FCC Auctions 1994\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\n**John Forbes Nash Jr.** (1928-2015) wrote his doctoral dissertation at Princeton in 1950, at age 21. The thesis was 28 pages. From it came the 1950 PNAS paper (two pages) and the 1951 *Annals of Mathematics* paper (10 pages). For these two papers — establishing the equilibrium concept that bears his name — Nash received the 1994 Nobel Memorial Prize in Economic Sciences (shared with Reinhard Selten and John Harsanyi for the broader development of non-cooperative game theory).\n\nThe theorem's mathematical proof used the Kakutani fixed-point theorem (Nash 1951). The proof's essence: in the space of mixed strategies, the best-response correspondence is upper-hemicontinuous and the strategy space is compact and convex; therefore by Kakutani, there exists a fixed point — i.e., a strategy combination where every player is best-responding. That fixed point is the equilibrium.\n\nWhat made Nash's contribution revolutionary was not the equilibrium concept itself (which had been informally articulated by Cournot in 1838 for the duopoly case) but the *generality* — the proof that *every* finite n-person game has at least one equilibrium. Before Nash, game theory could analyze only special cases. After Nash, game theory could analyze any finite strategic interaction. This is why Nash's two short papers are considered among the most important in 20th-century mathematics and economics.\n\nThe framework took decades to translate into operational policy. The breakthrough moment was the **1994 FCC spectrum auction**.\n\nThe U.S. Federal Communications Commission had historically allocated radio spectrum via \"comparative hearings\" (essentially bureaucratic beauty contests) and lotteries. Both methods had clear flaws: hearings were corruptible and slow; lotteries gave spectrum to entities who promptly resold it for windfall profits to actual operators. By the early 1990s, the value of spectrum was understood to be enormous and growing (especially for cellular telephony), and pressure was rising for a better mechanism.\n\nPaul Milgrom and Robert Wilson at Stanford were retained as consultants to design the new auction. Their core insight: spectrum auctions are complex strategic games with multiple bidders who hold private valuations for combinations of licenses. The auction rules must be designed so that bidders' Nash-equilibrium strategies — what each bidder rationally does given what every other bidder rationally does — produce efficient allocations (the spectrum ends up with the operators who value it most) and reasonable revenue.\n\nMilgrom and Wilson designed the **simultaneous multiple-round (SMR) auction**: licenses are auctioned simultaneously, in multiple rounds with rising bids, until no bidder is willing to raise any bid. This rule produces a Nash equilibrium in which bidders coordinate on combinations of licenses they want, while preventing bidders from being \"exposed\" (winning licenses individually but not the package they wanted). The design explicitly modeled bidder Nash strategies and chose rules to push the equilibrium toward efficient outcomes.\n\nFrom Milgrom's retrospective:\n\n> \"The SMR design rests on game-theoretic reasoning at every step. The rule structure determines what bidder strategies are available; the strategies determine the Nash equilibria; the equilibria determine the auction outcome. The FCC's 1994 auction raised $617 million for 99 licenses, against pre-auction estimates of $400 million for the prior beauty-contest method. The equilibrium logic worked: bidders coordinated through the rounds on packages of licenses that reflected their actual operational value, prices reflected willingness-to-pay, and the spectrum went to operators who could best use it. The subsequent FCC auctions — 87 of them through 2020 — have raised over $200 billion. The cumulative welfare gain to the U.S. economy is in the trillions. Game theory's policy-impact case rests substantially on this single application.\"\n>\n> — Milgrom (2004). *Putting Auction Theory to Work.* Cambridge University Press, ch. 1.\n\nMilgrom and Wilson received the 2020 Nobel Memorial Prize in Economic Sciences for their auction theory work, with the FCC case as the canonical example.\n\nThe framework has been applied at industrial scale beyond auctions:\n\n**Antitrust analysis.** Modern antitrust evaluation of mergers, vertical restraints, and platform conduct uses Nash-equilibrium modeling to predict how markets will behave under proposed changes. The DOJ and FTC employ economists who specialize in game-theoretic analysis of merger effects.\n\n**Mechanism design.** The broader field that grew from Nash's work — designing the rules of a game to produce desired equilibria — has applications in matching markets (kidney exchange, medical residency placement; Roth 2002, 2007 Nobel work), voting systems, school choice, and platform design.\n\n**Regulatory and policy analysis.** Setting taxes, designing regulations, structuring international trade agreements — all use Nash-equilibrium analysis to predict how strategic actors (firms, countries, voters) will respond.\n\n**Military and security strategy.** Cold War nuclear deterrence theory (Schelling 1960; Schelling 2005 Nobel for related work) is foundational Nash-equilibrium analysis. Modern cybersecurity, terrorism response, and great-power strategic competition all use the framework.\n\n**Corporate strategy.** Competitive interaction (pricing, capacity, R&D, marketing) in oligopolistic markets is Nash-equilibrium analysis. Sophisticated strategy departments at major firms (Walmart, Amazon, ExxonMobil, major banks) employ game theorists.\n\nThree operational lessons:\n\n**First, Nash equilibrium is the right tool when counterparties are rational and react to your moves.** If you don't model their reactions, you'll be systematically surprised by their behavior. The discipline of \"what's the best response to my move?\" prevents many strategic errors.\n\n**Second, equilibrium ≠ optimum.** The prisoner's dilemma is the canonical case: the equilibrium is bad for everyone, but the game's structure prevents the better outcome. Recognizing this is the prerequisite to redesigning the game (e.g., adding repeated interaction, changing payoffs, providing escrow).\n\n**Third, equilibrium-selection is often the operationally hardest problem.** Many games have multiple Nash equilibria, and the framework alone doesn't tell you which will emerge. Empirical, behavioral, and focal-point considerations matter. Treating Nash equilibrium as a complete theory of strategic behavior overstates its predictive power; treating it as one important component of strategic analysis is calibrated.\n\nFile v1.0.1:skill-card.md\n\n## Description: <br>\nNash Equilibrium helps agents analyze strategic decisions involving rational counterparties by specifying players, actions, payoffs, best responses, equilibria, and redesign options. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[deciqai](https://clawhub.ai/user/deciqai) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nDevelopers, operators, and decision makers use this skill to model competitive pricing, negotiation, M&A, regulation, auction, marketplace, and platform-design scenarios where multiple rational parties react to each other. It guides the agent to identify Nash equilibria, assess whether outcomes are Pareto-optimal, and recommend equilibrium play or game redesign. <br>\n\n### Deployment Geography for Use: <br>\nGlobal <br>\n\n## Known Risks and Mitigations: <br>\nRisk: The security review flags use in ClawHub operational environments as requiring authorization and scoped third-party service access. <br>\nMitigation: Install or use only where the operator is authorized, keep tokens scoped, and review any referenced service connections before use. <br>\nRisk: The security guidance calls out production, moderation, or destructive maintenance actions as requiring care. <br>\nMitigation: Require explicit confirmation before production, moderation, or destructive maintenance actions. <br>\n\n\n## Reference(s): <br>\n- [Nash Equilibrium Skill on ClawHub](https://clawhub.ai/deciqai/skills/nash-equilibrium) <br>\n- [Primary sources for Nash Equilibrium](references/sources.md) <br>\n- [Method in Action: Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md) <br>\n- [deciqAI](https://deciqai.com) <br>\n\n\n## Skill Output: <br>\n**Output Type(s):** [text, markdown, guidance] <br>\n**Output Format:** [Markdown analysis with structured decision sections] <br>\n**Output Parameters:** [1D] <br>\n**Other Properties Related to Output:** [May pause for user input in coach mode before completing the full analysis.] <br>\n\n## Skill Version(s): <br>\n1.0.1 (source: server release metadata) <br>\n\n## Ethical Considerations: <br>\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment. <br>\n\nArchive v1.0.0: 5 files, 9167 bytes\n\nFiles: examples/nash-1950-51-and-the-fcc-auctions-1994.md (6810b), references/sources.md (1365b), skill-card.md (2395b), SKILL.md (7345b), _meta.json (135b)\n\nFile v1.0.0:SKILL.md\n\n---\nname: nash-equilibrium\ndescription: \"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep ending up in a bad outcome even though everyone prefers better', or is analyzing a strategic situation with multiple rational counterparties (pricing, negotiation, M&A, regulation, platform launch). Do NOT activate when: the decision is essentially solo with no strategic counterparty; the counterparty is clearly irrational or acting on emotion rather than self-interest.\"\n---\n\n# Nash Equilibrium\n\n## Overview\n\nA **Nash equilibrium** is a stable point in multi-player interaction: a combination of strategies where no player can improve their payoff by unilaterally changing their own strategy, given others hold theirs fixed. Key properties: (1) best-response logic — the equilibrium is a fixed point of mutual best-responses; (2) equilibria can be Pareto-suboptimal (prisoner's dilemma); (3) multiple equilibria are common; (4) equilibrium does not predict the path to get there.\n\nComposes with [`prisoners-dilemma`](../prisoners-dilemma/SKILL.md), [`repeated-games-reputation`](../repeated-games-reputation/SKILL.md), [`signaling-games`](../signaling-games/SKILL.md), [`pricing-strategy`](../pricing-strategy/SKILL.md), and [`batna-zopa`](../batna-zopa/SKILL.md).\n\n## When to Use\n\n- Designing pricing in a competitive market with rational rivals\n- Negotiating with a sophisticated counterparty; modeling regulatory or legislative outcomes\n- Evaluating M&A or partnership decisions where multiple parties react strategically\n- Designing auctions, marketplaces, or platform mechanisms\n\n**Not when:** the situation is essentially solo; the counterparty is not rational; modeling cost exceeds the decision's value.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line what-it-is: find the point in a multi-player situation where no one has incentive to deviate from their strategy.\n2. Check fit: no rational counterparty? Nash adds less value than other frameworks.\n3. Elicit their real case: who are the players, what can each do, what does each gain or lose?\n> **[WAIT — do not advance until user responds]**\n4. Run The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close by naming the equilibrium and recommended action; if equilibrium is bad, name a redesign option.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Specify the game:** Players · Actions per player · Payoff matrix or game tree · Information structure (full vs. private) · Sequential or simultaneous.\n\n**Step 2 — Best-response analysis:** For each player, find the optimal action given each combination of others' strategies. The combination where everyone is best-responding is a Nash equilibrium.\n\n**Step 3 — Identify all equilibria:** Pure-strategy (deterministic) and mixed-strategy (randomized). If multiple equilibria, identify the most plausible focal point.\n\n**Step 4 — Evaluate quality:** Pareto-optimal? If not, what better collective outcome does the structure prevent?\n\n**Step 5 — Act or redesign:** Play equilibrium strategy if acceptable. If bad: change rules, add repeated interaction, change payoffs, add transparency — or walk away.\n\n**Step 6 — Plan counterparty response:** What do others do once you act? Contingency if they deviate from equilibrium.\n\n## Output Template\n\n```markdown\n# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate\n```\n\n*→ Method in Action: [Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md)*\n\n## Pack: Application Patterns\n\n| Domain | Typical equilibrium | Common pitfall |\n|---|---|---|\n| Pricing in oligopoly | Cournot markup or Bertrand marginal-cost | Assuming price-cuts go unanswered |\n| Auction bidding | Bid to value (English) or shaded bid (sealed) | Ignoring winner's curse in common-value auctions |\n| Platform launch | Multiple equilibria; focal point determines winner | Treating launch as solo, not a coordination game |\n| M&A bidding | Strategic incumbent often overpays | Failing to model how target plays bidders against each other |\n| Cartel formation | Cooperation possible if discount rate is low | Underestimating defection incentives and detection lag |\n\n## Applying It Well\n\nModel the game explicitly before acting. Equilibrium is not optimum — if the equilibrium is bad, redesign the game (add repetition, change payoffs, build trust). When multiple equilibria exist, use focal points or commitment devices to coordinate on the one you want.\n\n*→ Primary sources: [references/sources.md](references/sources.md)*\n\n## Common Rationalizations\n\n**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**\n\n| Fake move | Reality |\n|---|---|\n| [D] \"We'll just do the right thing; no need for game theory\" | The right thing depends on what others do. Without modeling reactions, you'll be surprised. |\n| [D] \"Game theory is too abstract for real decisions\" | The 1994 FCC auction raised $617M using Nash modeling. Auction design, antitrust, military strategy are operational applications. |\n| [D] \"Counterparties aren't rational; theory doesn't apply\" | They're often more rational than assumed. Use Nash as a baseline and adjust for behavioral deviations. |\n| [D] \"Multiple equilibria mean Nash isn't useful here\" | Multiple equilibria mean selection matters more — use focal points, communication, commitment devices. |\n| [D] \"First mover always wins\" | Often false. Equilibrium analysis tells you when first move helps and when it hurts. |\n| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |\n\n## Red Flags\n\n- Making a strategic decision without modeling how counterparties will respond\n- Assuming counterparties will react the way you'd react (mirror-imaging)\n- The equilibrium is bad and you haven't considered redesigning the game\n- Playing a repeated game using single-round logic\n- Multiple equilibria exist and you haven't considered which will emerge\n\n## Verification\n\n- [ ] Players, actions, and payoffs explicitly specified\n- [ ] Best-response analysis performed for each player\n- [ ] Nash equilibria identified; Pareto-optimality evaluated\n- [ ] If multiple equilibria, equilibrium-selection considered\n- [ ] If equilibrium is bad, game-redesign options considered\n- [ ] Counterparty response plan exists for equilibrium play and deviation\n\n---\n\n*Part of **deciqAI Knowledge Skills** — open-source thinking skills that make rigor executable for AI agents. Built by deciqAI · https://deciqai.com · Contributions welcome — see the template at the repo root.*\n\nFile v1.0.0:_meta.json\n\n{\n  \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n  \"slug\": \"nash-equilibrium\",\n  \"version\": \"1.0.0\",\n  \"publishedAt\": 1782818280313\n}\n\nFile v1.0.0:references/sources.md\n\n# Sources — nash-equilibrium\n\n> *Primary sources for the [nash-equilibrium](../SKILL.md) skill.*\n\n- Nash, J. F. (1950). \"Equilibrium Points in N-Person Games.\" *Proceedings of the National Academy of Sciences*, 36(1), 48-49. The foundational two-page proof.\n- Nash, J. F. (1951). \"Non-Cooperative Games.\" *Annals of Mathematics*, 54(2), 286-295. The extended treatment.\n- von Neumann, J. & Morgenstern, O. (1944). *Theory of Games and Economic Behavior.* Princeton University Press. The precursor; covers zero-sum games.\n- Cournot, A. A. (1838). *Recherches sur les principes mathématiques de la théorie des richesses.* Paris: Hachette. The duopoly precursor.\n- Milgrom, P. (2004). *Putting Auction Theory to Work.* Cambridge University Press. ISBN 978-0521536729. The FCC auction case study.\n- Camerer, C. F. (2003). *Behavioral Game Theory: Experiments in Strategic Interaction.* Princeton University Press. ISBN 978-0691090399. Empirical deviations from Nash predictions.\n- Roth, A. E. (2002). \"The Economist as Engineer.\" *Econometrica*, 70(4), 1341-1378. Mechanism-design applications.\n- Schelling, T. C. (1960). *The Strategy of Conflict.* Harvard University Press. ISBN 978-0674840317. Strategic applications to security.\n- Tirole, J. (1988). *The Theory of Industrial Organization.* MIT Press. ISBN 978-0262200714. Industrial-organization applications.\n\nFile v1.0.0:examples/nash-1950-51-and-the-fcc-auctions-1994.md\n\n# Method in Action: Nash 1950-51 and the FCC Auctions 1994\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\n**John Forbes Nash Jr.** (1928-2015) wrote his doctoral dissertation at Princeton in 1950, at age 21. The thesis was 28 pages. From it came the 1950 PNAS paper (two pages) and the 1951 *Annals of Mathematics* paper (10 pages). For these two papers — establishing the equilibrium concept that bears his name — Nash received the 1994 Nobel Memorial Prize in Economic Sciences (shared with Reinhard Selten and John Harsanyi for the broader development of non-cooperative game theory).\n\nThe theorem's mathematical proof used the Kakutani fixed-point theorem (Nash 1951). The proof's essence: in the space of mixed strategies, the best-response correspondence is upper-hemicontinuous and the strategy space is compact and convex; therefore by Kakutani, there exists a fixed point — i.e., a strategy combination where every player is best-responding. That fixed point is the equilibrium.\n\nWhat made Nash's contribution revolutionary was not the equilibrium concept itself (which had been informally articulated by Cournot in 1838 for the duopoly case) but the *generality* — the proof that *every* finite n-person game has at least one equilibrium. Before Nash, game theory could analyze only special cases. After Nash, game theory could analyze any finite strategic interaction. This is why Nash's two short papers are considered among the most important in 20th-century mathematics and economics.\n\nThe framework took decades to translate into operational policy. The breakthrough moment was the **1994 FCC spectrum auction**.\n\nThe U.S. Federal Communications Commission had historically allocated radio spectrum via \"comparative hearings\" (essentially bureaucratic beauty contests) and lotteries. Both methods had clear flaws: hearings were corruptible and slow; lotteries gave spectrum to entities who promptly resold it for windfall profits to actual operators. By the early 1990s, the value of spectrum was understood to be enormous and growing (especially for cellular telephony), and pressure was rising for a better mechanism.\n\nPaul Milgrom and Robert Wilson at Stanford were retained as consultants to design the new auction. Their core insight: spectrum auctions are complex strategic games with multiple bidders who hold private valuations for combinations of licenses. The auction rules must be designed so that bidders' Nash-equilibrium strategies — what each bidder rationally does given what every other bidder rationally does — produce efficient allocations (the spectrum ends up with the operators who value it most) and reasonable revenue.\n\nMilgrom and Wilson designed the **simultaneous multiple-round (SMR) auction**: licenses are auctioned simultaneously, in multiple rounds with rising bids, until no bidder is willing to raise any bid. This rule produces a Nash equilibrium in which bidders coordinate on combinations of licenses they want, while preventing bidders from being \"exposed\" (winning licenses individually but not the package they wanted). The design explicitly modeled bidder Nash strategies and chose rules to push the equilibrium toward efficient outcomes.\n\nFrom Milgrom's retrospective:\n\n> \"The SMR design rests on game-theoretic reasoning at every step. The rule structure determines what bidder strategies are available; the strategies determine the Nash equilibria; the equilibria determine the auction outcome. The FCC's 1994 auction raised $617 million for 99 licenses, against pre-auction estimates of $400 million for the prior beauty-contest method. The equilibrium logic worked: bidders coordinated through the rounds on packages of licenses that reflected their actual operational value, prices reflected willingness-to-pay, and the spectrum went to operators who could best use it. The subsequent FCC auctions — 87 of them through 2020 — have raised over $200 billion. The cumulative welfare gain to the U.S. economy is in the trillions. Game theory's policy-impact case rests substantially on this single application.\"\n>\n> — Milgrom (2004). *Putting Auction Theory to Work.* Cambridge University Press, ch. 1.\n\nMilgrom and Wilson received the 2020 Nobel Memorial Prize in Economic Sciences for their auction theory work, with the FCC case as the canonical example.\n\nThe framework has been applied at industrial scale beyond auctions:\n\n**Antitrust analysis.** Modern antitrust evaluation of mergers, vertical restraints, and platform conduct uses Nash-equilibrium modeling to predict how markets will behave under proposed changes. The DOJ and FTC employ economists who specialize in game-theoretic analysis of merger effects.\n\n**Mechanism design.** The broader field that grew from Nash's work — designing the rules of a game to produce desired equilibria — has applications in matching markets (kidney exchange, medical residency placement; Roth 2002, 2007 Nobel work), voting systems, school choice, and platform design.\n\n**Regulatory and policy analysis.** Setting taxes, designing regulations, structuring international trade agreements — all use Nash-equilibrium analysis to predict how strategic actors (firms, countries, voters) will respond.\n\n**Military and security strategy.** Cold War nuclear deterrence theory (Schelling 1960; Schelling 2005 Nobel for related work) is foundational Nash-equilibrium analysis. Modern cybersecurity, terrorism response, and great-power strategic competition all use the framework.\n\n**Corporate strategy.** Competitive interaction (pricing, capacity, R&D, marketing) in oligopolistic markets is Nash-equilibrium analysis. Sophisticated strategy departments at major firms (Walmart, Amazon, ExxonMobil, major banks) employ game theorists.\n\nThree operational lessons:\n\n**First, Nash equilibrium is the right tool when counterparties are rational and react to your moves.** If you don't model their reactions, you'll be systematically surprised by their behavior. The discipline of \"what's the best response to my move?\" prevents many strategic errors.\n\n**Second, equilibrium ≠ optimum.** The prisoner's dilemma is the canonical case: the equilibrium is bad for everyone, but the game's structure prevents the better outcome. Recognizing this is the prerequisite to redesigning the game (e.g., adding repeated interaction, changing payoffs, providing escrow).\n\n**Third, equilibrium-selection is often the operationally hardest problem.** Many games have multiple Nash equilibria, and the framework alone doesn't tell you which will emerge. Empirical, behavioral, and focal-point considerations matter. Treating Nash equilibrium as a complete theory of strategic behavior overstates its predictive power; treating it as one important component of strategic analysis is calibrated.\n\nFile v1.0.0:skill-card.md\n\n## Description: <br>\nGuides agents through Nash equilibrium analysis for strategic situations with multiple rational counterparties, including pricing, negotiation, regulation, platform, auction, marketplace, and mechanism-design decisions. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[deciqai](https://clawhub.ai/user/deciqai) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nDevelopers, analysts, and operators use this skill to model strategic decisions where counterparties respond rationally. It helps specify players, actions, payoffs, information structure, equilibria, Pareto quality, redesign options, and counterparty response plans. <br>\n\n### Deployment Geography for Use: <br>\nGlobal <br>\n\n## Known Risks and Mitigations: <br>\nRisk: Strategic analysis can be misleading when players, payoffs, or assumptions are wrong or incomplete. <br>\nMitigation: Verify assumptions, payoff estimates, and sources independently before relying on the analysis for high-stakes pricing, M&A, regulatory, auction, or market decisions. <br>\nRisk: The skill provides analytical guidance and is not a substitute for legal, financial, market, or domain expert review. <br>\nMitigation: Use expert review for consequential decisions and treat the generated Nash analysis as one decision input, not final advice. <br>\n\n\n## Reference(s): <br>\n- [ClawHub Skill Page](https://clawhub.ai/deciqai/skills/nash-equilibrium) <br>\n- [deciqAI Publisher Profile](https://clawhub.ai/user/deciqai) <br>\n- [deciqAI Homepage](https://deciqai.com) <br>\n- [Primary Sources](references/sources.md) <br>\n- [Method in Action: Nash 1950-51 and the FCC Auctions 1994](examples/nash-1950-51-and-the-fcc-auctions-1994.md) <br>\n\n\n## Skill Output: <br>\n**Output Type(s):** [text, markdown, guidance] <br>\n**Output Format:** [Markdown analysis with structured sections] <br>\n**Output Parameters:** [1D] <br>\n**Other Properties Related to Output:** [May pause for user input in coach mode before completing the analysis.] <br>\n\n## Skill Version(s): <br>\n1.0.0 (source: server release evidence) <br>\n\n## Ethical Considerations: <br>\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment. <br>","readmeExcerpt":"Skill: Nash Equilibrium Owner: deciqai Summary: Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep e... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-07-16T18:08:07.042Z | user Description tail link + agents machine-readable metadata line (deciqai.com/s/nash-equilibrium.json) v1.0.4 | 2026-07-09T11:19:05.292Z ","codeSnippets":[],"executableExamples":[{"language":"markdown","snippet":"# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate"},{"language":"markdown","snippet":"# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate"},{"language":"markdown","snippet":"# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate"},{"language":"markdown","snippet":"# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate"},{"language":"markdown","snippet":"# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate"},{"language":"markdown","snippet":"# Nash Analysis: <situation>\nPlayers / Actions / Payoffs / Info structure / Sequential or simultaneous\nBest-response analysis (per player) →\nEquilibria found: <strategy combination + payoffs>\nPareto-optimal? (Y/N) | If N: better outcome + why game doesn't produce it\nDecision: my action | Game redesign considered\nCounterparty response plan: expected action | contingency if they deviate"}],"parameters":null,"dependencies":[],"permissions":[],"extractedFiles":[{"path":"SKILL.md","content":"---\nname: nash-equilibrium\ndescription: \"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep ending up in a bad outcome even though everyone prefers better', or is analyzing a strategic situation with multiple rational counterparties (pricing, negotiation, M&A, regulation, platform launch). Do NOT activate when: the decision is essentially solo with no strategic counterparty; the counterparty is clearly irrational or acting on emotion rather than self-interest. More: deciqai.com/c/nash-equilibrium\"\n---\n\n# Nash Equilibrium\n\n## Overview\n\nA **Nash equilibrium** is a stable point in multi-player interaction: a combination of strategies where no player can improve their payoff by unilaterally changing their own strategy, given others hold theirs fixed. Key properties: (1) best-response logic — the equilibrium is a fixed point of mutual best-responses; (2) equilibria can be Pareto-suboptimal (prisoner's dilemma); (3) multiple equilibria are common; (4) equilibrium does not predict the path to get there.\n\nComposes with `prisoners-dilemma`, `repeated-games-reputation`, `signaling-games`, `pricing-strategy`, and `batna-zopa`.\n\n## When to Use\n\n- Designing pricing in a competitive market with rational rivals\n- Negotiating with a sophisticated counterparty; modeling regulatory or legislative outcomes\n- Evaluating M&A or partnership decisions where multiple parties react strategically\n- Designing auctions, marketplaces, or platform mechanisms\n- Assessing an investment or capacity arms race where rivals match each other (e.g., AI capex spending, AI-native competition, matching AI adoption to keep position)\n\n**Not when:** the situation is essentially solo; the counterparty is not rational; modeling cost exceeds the decision's value.\n\n## Coaching Novices (Adaptive Front Door)\n\n- **Engine mode:** user has a concrete case → run The Process directly.\n- **Coach mode:** user is unfamiliar → guide step by step.\n\nIn Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.\n\n1. One-line what-it-is: find the point in a multi-player situation where no one has incentive to deviate from their strategy.\n2. Check fit: no rational counterparty? Nash adds less value than other frameworks.\n3. Elicit their real case: who are the players, what can each do, what does each gain or lose?\n> **[WAIT — do not advance until user responds]**\n4. Run The Process one step at a time with their input.\n> **[WAIT — do not advance until user responds]**\n5. Close by naming the equilibrium and recommended action; if equilibrium is bad, name a redesign option.\n> **[WAIT — do not advance until user responds]**\n\n## The Process\n\n**Step 1 — Specify the game:** Players · Actions per player · Payoff matrix or game tree · Information structure (full vs. private) · Sequential or simultaneous.\n\n**Step 2 — Best-response analysis:** For each player, fin"},{"path":"_meta.json","content":"{\n  \"ownerId\": \"kn754b8sk22s8c6gjxt02bftbn88q7ye\",\n  \"slug\": \"nash-equilibrium\",\n  \"version\": \"1.0.5\",\n  \"publishedAt\": 1784225287042\n}"},{"path":"references/sources.md","content":"# Sources — nash-equilibrium\n\n> *Primary sources for the [nash-equilibrium](../SKILL.md) skill.*\n\n- Nash, J. F. (1950). \"Equilibrium Points in N-Person Games.\" *Proceedings of the National Academy of Sciences*, 36(1), 48-49. The foundational two-page proof.\n- Nash, J. F. (1951). \"Non-Cooperative Games.\" *Annals of Mathematics*, 54(2), 286-295. The extended treatment.\n- von Neumann, J. & Morgenstern, O. (1944). *Theory of Games and Economic Behavior.* Princeton University Press. The precursor; covers zero-sum games.\n- Cournot, A. A. (1838). *Recherches sur les principes mathématiques de la théorie des richesses.* Paris: Hachette. The duopoly precursor.\n- Milgrom, P. (2004). *Putting Auction Theory to Work.* Cambridge University Press. ISBN 978-0521536729. The FCC auction case study.\n- Camerer, C. F. (2003). *Behavioral Game Theory: Experiments in Strategic Interaction.* Princeton University Press. ISBN 978-0691090399. Empirical deviations from Nash predictions.\n- Palacios-Huerta, I. (2003). \"Professionals Play Minimax.\" *Review of Economic Studies*, 70(2), 395-415. Field test of mixed-strategy equilibrium on professional penalty kicks.\n- Roth, A. E. (2002). \"The Economist as Engineer.\" *Econometrica*, 70(4), 1341-1378. Mechanism-design applications.\n- Schelling, T. C. (1960). *The Strategy of Conflict.* Harvard University Press. ISBN 978-0674840317. Strategic applications to security.\n- Tirole, J. (1988). *The Theory of Industrial Organization.* MIT Press. ISBN 978-0262200714. Industrial-organization applications.\n- Microsoft, Alphabet, Amazon, and Meta (2024–2026). Quarterly earnings calls and investor disclosures on AI-related capital expenditure guidance. Contemporary example of investment competition sustaining a high-capex equilibrium among rational rivals.\n- Investor commentary and reporting on the hyperscaler AI-capex build-out (2024–2026), e.g., quarterly earnings coverage in the financial press. Documents the \"risk of under-investing exceeds risk of over-investing\" framing that signals best-response spending."},{"path":"examples/ai-capex-race-among-hyperscalers-2024-2026.md","content":"# Method in Action: The AI-Capex Race Among Hyperscalers (2024–2026)\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\nAcross 2024 and 2025 (and into 2026), the largest U.S. cloud-and-platform companies — Microsoft, Alphabet (Google), Amazon, and Meta — raised their capital expenditure to record levels, the bulk of it directed at AI data centers, accelerators, and power. On successive earnings calls, each firm's leadership publicly framed the risk of *under*-investing in AI capacity as outweighing the risk of *over*-investing. That is a telltale sign of a Nash equilibrium: not four firms independently arriving at the same plan, but four firms each best-responding to what the others are doing. This example runs the case through the skill's six-step Process.\n\n**Step 1 — Specify the game.** *Players:* the major hyperscalers (Microsoft, Alphabet, Amazon, Meta), plus adjacent capacity buyers. *Actions per player:* along a spectrum from \"spend aggressively on AI compute\" to \"hold capex flat and harvest cash.\" *Payoffs:* market position in AI-driven cloud and consumer products, which depends on relative capacity, not just absolute spend. *Information:* largely public — capex guidance is disclosed on quarterly earnings calls, so each player observes rivals' commitments with a short lag. *Timing:* effectively simultaneous and repeated, quarter after quarter, since all players revise guidance on a similar cadence.\n\n**Step 2 — Best-response analysis.** Consider any single firm's choice given the others are spending heavily. If it also spends, it holds its position and shares in whatever AI demand materializes. If it unilaterally pulls back, it frees cash and lifts near-term margins — but it risks ceding compute capacity, model quality, and enterprise-cloud share to rivals who kept building, in a market where capacity is a binding constraint and lead times for chips and power are long. Given rivals spending, the best response is to keep spending. The same logic holds for each player. The mutual-best-response fixed point is \"everyone spends heavily.\"\n\n**Step 3 — Identify all equilibria.** The high-capex profile is a pure-strategy Nash equilibrium: no firm can improve its position by unilaterally cutting while the others build. A low-capex \"everyone restrains\" profile would be more profitable collectively in the short run, but it is *not* an equilibrium — from it, any single firm gains by defecting and out-building the restrained rivals to capture position, so it unravels. This is the structural signature of a prisoner's-dilemma-type game: the cooperative outcome is not self-enforcing. The most plausible focal point is therefore the high-spend equilibrium, which is what the public capex trajectory through this period reflects.\n\n**Step 4 — Evaluate quality.** The equilibrium is not Pareto-optimal for the firms as a group: collectively they would earn higher near-term free cash flow spending less, if all could credibly commit to restraint. The game st"},{"path":"examples/nash-1950-51-and-the-fcc-auctions-1994.md","content":"# Method in Action: Nash 1950-51 and the FCC Auctions 1994\n\n> *Example for the [nash-equilibrium](../SKILL.md) skill.*\n\n**John Forbes Nash Jr.** (1928-2015) wrote his doctoral dissertation at Princeton in 1950, at age 21. The thesis was 28 pages. From it came the 1950 PNAS paper (two pages) and the 1951 *Annals of Mathematics* paper (10 pages). For these two papers — establishing the equilibrium concept that bears his name — Nash received the 1994 Nobel Memorial Prize in Economic Sciences (shared with Reinhard Selten and John Harsanyi for the broader development of non-cooperative game theory).\n\nThe theorem's mathematical proof used the Kakutani fixed-point theorem (Nash 1951). The proof's essence: in the space of mixed strategies, the best-response correspondence is upper-hemicontinuous and the strategy space is compact and convex; therefore by Kakutani, there exists a fixed point — i.e., a strategy combination where every player is best-responding. That fixed point is the equilibrium.\n\nWhat made Nash's contribution revolutionary was not the equilibrium concept itself (which had been informally articulated by Cournot in 1838 for the duopoly case) but the *generality* — the proof that *every* finite n-person game has at least one equilibrium. Before Nash, game theory could analyze only special cases. After Nash, game theory could analyze any finite strategic interaction. This is why Nash's two short papers are considered among the most important in 20th-century mathematics and economics.\n\nThe framework took decades to translate into operational policy. The breakthrough moment was the **1994 FCC spectrum auction**.\n\nThe U.S. Federal Communications Commission had historically allocated radio spectrum via \"comparative hearings\" (essentially bureaucratic beauty contests) and lotteries. Both methods had clear flaws: hearings were corruptible and slow; lotteries gave spectrum to entities who promptly resold it for windfall profits to actual operators. By the early 1990s, the value of spectrum was understood to be enormous and growing (especially for cellular telephony), and pressure was rising for a better mechanism.\n\nPaul Milgrom and Robert Wilson at Stanford were retained as consultants to design the new auction. Their core insight: spectrum auctions are complex strategic games with multiple bidders who hold private valuations for combinations of licenses. The auction rules must be designed so that bidders' Nash-equilibrium strategies — what each bidder rationally does given what every other bidder rationally does — produce efficient allocations (the spectrum ends up with the operators who value it most) and reasonable revenue.\n\nMilgrom and Wilson designed the **simultaneous multiple-round (SMR) auction**: licenses are auctioned simultaneously, in multiple rounds with rising bids, until no bidder is willing to raise any bid. This rule produces a Nash equilibrium in which bidders coordinate on combinations of licenses they want, while preventing bidde"}],"languages":[],"docsSourceLabel":"CLAWHUB","editorialOverview":"Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep e... Skill: Nash Equilibrium Owner: deciqai Summary: Activate when: user asks 'what will they do if we do X', 'how will competitors react to our pricing', 'how do I design this auction or mechanism', 'we keep e... Tags: latest:1.0.5 Version history: v1.0.5 | 2026-07-16T18:08:07.042Z | user Description tail link + agents machine-readable metadata line (deciqai.com/s/nash-equilibrium.json) v1.0.4 | 2026-07-09T11:19:05.292Z","editorialQuality":{"score":100,"threshold":65,"status":"ready","wordCount":2034,"uniquenessScore":50,"reasons":[]}},"media":{"evidence":{"source":"no-media","verified":false,"confidence":"low","updatedAt":"2026-10-11T09:18:05.017Z","emptyReason":"No screenshots, media assets, or demo links are available."},"primaryImageUrl":null,"mediaAssetCount":0,"assets":[],"demoUrl":null},"ownerResources":{"evidence":{"source":"unclaimed","verified":false,"confidence":"low","updatedAt":"2026-10-11T09:18:05.017Z","emptyReason":"This page has not been claimed by the agent owner."},"hasCustomPage":false,"customPageUpdatedAt":null,"customLinks":[],"structuredLinks":{"docsUrl":null,"demoUrl":null,"supportUrl":null,"pricingUrl":null,"statusUrl":null},"customPage":null},"relatedAgents":{"evidence":{"source":"protocol-neighbors","verified":false,"confidence":"medium","updatedAt":"2026-10-11T11:25:43.489Z","emptyReason":null},"items":[{"id":"8ebccd8e-3863-4187-8355-c3f14e1f9edf","entityType":"agent","canonicalPath":"/agent/iofficeai-aionui","slug":"iofficeai-aionui","name":"AionUi","description":"Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!","url":"https://github.com/iOfficeAI/AionUi","homepage":"https://www.aionui.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-10-09T19:11:12.944Z","createdAt":"2026-02-25T03:38:16.584Z","downloads":null},{"id":"b917f68a-ebff-438e-84f8-3f4b2494c0bc","entityType":"agent","canonicalPath":"/agent/activepieces-activepieces","slug":"activepieces-activepieces","name":"activepieces","description":"AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents","url":"https://github.com/activepieces/activepieces","homepage":"https://www.activepieces.com","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-15T02:22:12.426Z","createdAt":"2026-02-25T03:38:12.412Z","downloads":null},{"id":"5cb26759-3a39-483f-94cf-276a98c13bb8","entityType":"agent","canonicalPath":"/agent/cherryhq-cherry-studio","slug":"cherryhq-cherry-studio","name":"cherry-studio","description":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs","url":"https://github.com/CherryHQ/cherry-studio","homepage":"https://cherry-ai.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-11T14:38:40.986Z","createdAt":"2026-02-25T03:38:19.379Z","downloads":null},{"id":"6f6582d0-5d76-4f0f-b81d-86520247950b","entityType":"agent","canonicalPath":"/agent/copilotkit-copilotkit","slug":"copilotkit-copilotkit","name":"CopilotKit","description":"The Frontend for Agents & Generative UI. React + Angular","url":"https://github.com/CopilotKit/CopilotKit","homepage":"https://docs.copilotkit.ai","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-03-25T09:50:57.846Z","createdAt":"2026-02-25T03:39:14.617Z","downloads":null}],"links":{"hub":"/agent","source":"/agent/source/clawhub","protocols":[{"label":"OpenClaw","href":"/agent/protocol/openclew"}]}}}