── ── Cognitive bias
Zero-Sum Game
Zero-sum means total value is fixed — one player's gain is another's exact loss. Most real-world competition is NOT zero-sum: the pie can grow, shrink, or be split in many ways. Misdiagnosis sends strategy in the wrong direction from step one. The most consequential error is zero-sum bias: the tendency to perceive non-zero-sum situations as zero-sum.
Run Zero-Sum Game on a real problem
Bring something you're actually deciding — free, in the browser.
How it works
Run the Zero-Sum Diagnosis. Five gates; confirm or rule out at each one.
1. Define the contested resource. State precisely what is being competed over (market share, license, price spread, votes, contract). If you cannot name a concrete unit being divided, the zero-sum frame likely does not apply. 2. Test fixity. Can innovation/technology expand the total? Can cooperation create additional value? Can time change the total? If any answer is "yes," the situation is non-zero-sum in that dimension. 3. Check for zero-sum bias. Are you perceiving zero-sum because the resource is countable? Anchored on relative position over absolute gains? Ignoring comparative advantage? If expansion is feasible, you are in the wrong game. 4. If confirmed zero-sum: apply minimax. Enumerate strategies and worst-case payoffs; choose the strategy maximizing your minimum; consider mixed strategies to prevent exploitation. 5. If confirmed non-zero-sum: design for cooperative surplus. Quantify value neither party gets under pure competition. Specify the mechanism (contract, JV, standard, platform) to capture it. Stop-rule: if you cannot identify a concrete pie-growth mechanism, revert to zero-sum analysis. 6. State the time horizon. Many situations are zero-sum short-term and non-zero-sum long-term. State both frames explicitly.
When to use it
- someone asks 'are we fighting over a fixed pie?', 'should we cooperate or compete?', 'is the market growing or are we just stealing share?', uses phrases like 'winner-take-all', 'race to the bottom', or 'your gain is my loss', or needs to check whether a negotiation/market/policy situation is truly zero-sum before designing strategy
When not to use it
When the decision is routine and reversible, applying a formal method costs more than it returns.
Worked example
Where the AI Race Is Zero-Sum and Where It Isn't (2024–2026)
By 2024–2026 the dominant framing of the AI boom was a single "race" — one leaderboard, one winner, everyone else loses. That framing quietly bundles together resources with very different structures. Some inputs to AI are genuinely fixed in the near term and therefore zero-sum; the output — AI-driven productivity — is not. Treating the whole thing as one zero-sum contest is exactly the diagnosis error this skill is built to catch: it pushes firms toward pure capture (outbid rivals, hoard talent, block competitors) when…
Install this skill (free, MIT)
npx skills add deciqAI/knowledge-skillsUseful? Star the repo — stars help other builders find it.
Related mental models
We seek and weight evidence that confirms what we already believe. Hunt for what would prove you wrong.
Money and effort already spent are gone. Decide only on the marginal cost and benefit from here forward.
Faced with a non-trivial decision, ask three questions: How will I feel in 10 minutes?
When people estimate an unknown quantity they start from a reference point — an anchor — and adjust.