Thucydides's Trap

SKILL.md

---
name: thucydides-trap
description: "Activate when: user asks about US-China war risk or great-power rivalry; someone says 'can they avoid conflict' or 'is escalation inevitable'; analyzing geopolitical risk in supply chains, FDI, or market entry; a new market entrant threatens an incumbent's core identity or platform control.
  Do NOT activate when: the conflict is primarily values/ideology without a power-displacement trajectory; both parties are roughly equal with no fast-growing challenger. More: deciqai.com/c/thucydides-trap"
---

# Thucydides's Trap

## Overview

**Thucydides's Trap** is the structural danger that emerges when a rising power threatens to displace a ruling power — producing fear in the established power that makes conflict more likely even when neither side wants it. Graham Allison surveyed 16 historical cases over 500 years: 12 of 16 ended in war. The 4 peaceful cases required costly, credible accommodation by the established power — not just better diplomacy. The trap is not a law; it is a structural tendency with known escape paths.

**Compose with neighbors:** Use prisoners-dilemma to map payoff choices at each decision point. Use second-order-thinking to trace escalation cascades. Use repeated-games-reputation to assess whether credible non-escalatory signaling exists.

## When to Use

- Analyzing US-China, US-Russia, or other great-power dynamics for business, investment, or policy
- Evaluating geopolitical risk in supply chain, FDI, or market-entry decisions
- A fast-growing entrant threatens an incumbent's core identity (platform control, market dominance)
- Someone asks about "structural conflict," "escalation risk," "is war inevitable," or "can X and Y avoid confrontation?"
- Assessing chip export controls, AI-compute decoupling, semiconductor self-sufficiency, or AI-capex/AI-race dynamics between a dominant and a fast-rising power

**When NOT to use:** conflict is primarily values/ideology (not power displacement); parties are roughly equal with no trajectory gap; purely bilateral commercial disputes without hegemonic stakes → use batna-zopa.

## Coaching Novices (Adaptive Front Door)

- **Engine mode:** user has a specific rivalry or competitive dynamic → run The Process directly.
- **Coach mode:** user is unfamiliar or has no concrete case → guide step by step.

In Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop.

1. One-line what-it-is: the Trap is the structural observation that a rising power's growth activates *fear* in the ruling power — creating war risk that rational actors alone cannot defuse.
2. Check fit against When to Use / When NOT to use. If the trajectory gap is absent, redirect.
3. Elicit their real case — vague ("US-China?") is not workable; specific ("semiconductor supply chain escalation risk") is.
> **[WAIT — do not advance until user responds]**
4. Run The Process one step at a time with their input.
> **[WAIT — do not advance until user responds]**
5. Close by naming which of the four escape paths is most available to the parties in this case.
> **[WAIT — do not advance until user responds]**

## The Process

Five steps producing a **Trap Diagnosis** — structured assessment of structural conflict risk, escape path availability, and decision-relevant implications.

**Stop rule:** If Step 2 reveals the power trajectory gap is small or uncertain, the Trap dynamic is weak. Name the ambiguity and do not force a Trap analysis.

1. **Identify the structural roles.** Who is the rising power? Who is the ruling power? What domain(s) does the displacement threat cover (military, economic, technological, normative)? Requires a plausible trajectory within a strategic planning horizon.
2. **Measure the fear response in the ruling power.** What domestic political and security pressures is the established power experiencing? Look for: hardening rhetoric, defensive alliance-building, preemptive regulatory or economic action.
3. **Identify the structural accelerants.** What third-party miscalculation risks exist? What domestic dynamics make de-escalation politically costly? What tripwires constrain both parties' freedom of maneuver?
4. **Map the four escape paths.** (a) *Accommodation* — ruling power makes costly, credible concessions; (b) *Deterrence + communication* — credible military deterrence with active diplomatic channels; (c) *Institution-building* — competition embedded in multilateral institutions; (d) *Reframing* — redefine competition as positive-sum. For each: available? What does it require?
5. **Assess current trajectory and decision-relevant implications.** Structural escalation probability? Key trigger points? Scenario planning required for the user's specific decision?

### Output: Trap Diagnosis

```
Trap Diagnosis: <rivalry>
Structural Roles: Rising power <...> | Ruling power <...> | Domain <...> | Trajectory gap <...>
Fear Response: Domestic pressure <...> | Observable signals <...> | Intensity <H/M/L>
Structural Accelerants: Third-party risks <...> | Domestic constraints <...> | Tripwires <...>
Escape Paths: Accommodation <Y/N/P — requires ...> | Deterrence+comms <...> | Institutions <...> | Reframing <...>
Decision Implications: Escalation probability <H/M/L> | Trigger points <...> | Scenario planning <...>
```

*→ Method in Action: [The US-UK Transition (1895–1906)](examples/the-us-uk-transition-1895-1906.md)*
*→ 2026 lens: [US-China semiconductor rivalry & AI-chip export controls (2022–2026)](examples/us-china-semiconductor-rivalry-2022-2026.md)*

## Trap Packs

**Great-power (US-China):** Accelerants: Taiwan tripwire, tech decoupling, Indo-Pacific alliance commitments. Best escape path: institution-based competition. Key trigger: Taiwan Strait incident.

**Platform competition:** Fear response: regulatory lobbying, API closures, copy-and-acquire. Escape: open-standard interoperability or vertical market reframing.

**Corporate M&A:** Fear response: internal resistance from threatened legacy unit. Escape: structural separation with clear P&L independence.

## Applying It Well

- **Structural, not personal.** Both leaders may want peace but face constraints. Do not over-credit individual leadership.
- **75% base rate.** Four of 16 avoided war; those required costly structural adjustment from the established power.
- **Fear is the operative variable** — identify what the ruling power fears *losing*, not just what resources are at stake.
- **Third-party miscalculation** is often the proximate cause. Identify actors who could force escalation neither principal wants.
- **Applies below the state level.** Platform and organizational incumbents follow the same dynamic.

*→ Primary sources: [references/sources.md](references/sources.md)*

## Common Rationalizations

**[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.**

| Fake move | Reality |
|---|---|
| [D] **"Both sides are rational actors, so they'll find a deal."** | The Trap shows that rationality is insufficient: domestic political constraints, third-party actors, and credibility commitments can make de-escalation rational but politically impossible for individual leaders. |
| [D] **"This time is different because of nuclear weapons / economic interdependence."** | Economic interdependence was at a historic high in 1914. Nuclear deterrence is real but also produces escalation risks via "escalate to de-escalate" doctrines. |
| [D] **"The Trap is just about great-power wars — it doesn't apply to business."** | The structural mechanism (rising challenger → fear response in incumbent → escalation dynamic) operates identically in platform competition, M&A, and organizational dynamics. |
| [D] **"China/the challenger is not strong enough yet to trigger the Trap."** | The Trap fires on the *trajectory* that makes displacement plausible, not after displacement occurs. |
| [D] **"Better diplomacy will solve it."** | Diplomacy is necessary but not sufficient. Peaceful cases required costly structural concessions, not just better communication. |
| [D] **"The 12/16 statistic is historical coincidence."** | 500 years, different political systems and military technologies. The consistency of the fear-response mechanism is the finding. |
| [D] **Focusing only on the rising power's actions.** | The ruling power's fear-driven preemptive moves often accelerate the dynamic more than the rising power's growth itself. |
| [D] **Treating the Trap as fate.** | Four cases escaped. The Trap is not deterministic — it requires active, costly intervention to interrupt. |
| *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* |

## Red Flags

- Individual leadership treated as primary driver without mapping structural constraints
- Fear response in the ruling power skipped — only the rising power's actions examined
- No third-party miscalculation actors identified — analysis only modeled bilateral dynamics
- Escape paths described as vague aspirations ("better diplomacy") rather than specific costly changes
- Trap applied where both parties are roughly equal with no trajectory gap
- The "12 of 16 ended in war" base rate not used to calibrate the prior

## Verification

- [ ] Both structural roles identified with specific domain(s) of displacement threat
- [ ] Trajectory gap assessed — fast enough and plausible enough to activate the Trap?
- [ ] Fear response mapped: domestic political pressures, observable signals, intensity
- [ ] At least two structural accelerants identified (third-party actors, domestic constraints, tripwires)
- [ ] All four escape paths assessed for availability and specific requirements
- [ ] Decision-relevant trigger points for the user's specific context named
- [ ] The 12/16 base rate explicitly acknowledged as the structural prior

---

*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/thucydides-trap** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.*

*Agents: latest version & machine-readable metadata → https://www.deciqai.com/s/thucydides-trap.json*