Learn it → Apply it → Save as Skill
The big win only happens if everyone commits at the same time. Two hunters can bring down a stag together and eat well, or each can catch a hare alone and eat a little. The hard part isn't temptation to betray each other — it's trusting that your partner won't quietly settle for the smaller, safer prize.
The Dr. Wallace's Favorite Mental Models is designed to turn useful ways of thinking into reusable tools. Each model gives you a fast visual overview, deeper interactive learning, practical application exercises and an AI-ready Markdown skill you can save for future use.
Instead of merely reading about mental models, you can gradually build your own AI skills library: a collection of portable reasoning skills that can help ChatGPT, Claude, Grok and other AI assistants apply the mental models you value.
Unlike most "cooperation" puzzles, there's no reward for secretly betraying a partner who stays committed. The real risk is coordination failure: if you doubt your partner will commit, playing it safe yourself becomes rational too — and the group ends up with the smaller prize even though everyone preferred the bigger one.
| Partner: Stag | Partner: Hare | |
|---|---|---|
| You: Stag | Big win, both | You get nothing |
| You: Hare | Small win, partner gets nothing | Small win, both |
Important: two outcomes are stable (both Stag, or both Hare). The dilemma is which one you'll actually land on — not whether to exploit each other.
The model is most useful whenever a goal is only reachable if multiple people commit together, and no one gains by betraying a partner who stays committed.
The Stag Hunt is a short filter you run before assuming a cooperation problem is really about temptation to betray.
Check whether anyone actually gains by defecting on a committed partner. If not, this is a Stag Hunt, not a Prisoner's Dilemma.
Look for a visible, costly sign that the other side is genuinely committing, not just saying so.
Once trust is reasonable, commit to the bigger goal rather than quietly hedging with the safe option.
Re-check commitment regularly. A Stag Hunt can slide to the smaller equilibrium if any party starts hedging.
Open a use case to see what the model changes in practice.
If both founders commit full-time, the pivot has a real shot. If either quietly keeps a safety-net side project "just in case," the pivot likely fails anyway, and the half-committed founder's hedge becomes the real outcome. The fix is an explicit, visible commitment from both sides, not a verbal promise alone.
If every team commits full resources, the launch can hit its stretch date. If one team hedges by under-staffing "in case it slips," that hedge becomes self-fulfilling and the group lands on the smaller, safer outcome everyone could have avoided.
If every major emitter commits to real cuts, everyone benefits. If each fears the others will under-deliver, everyone has an incentive to under-commit too — not because cheating pays, but because matching a partner who might not show up feels safer. This is a textbook stag hunt dynamic cited in international-law scholarship on cooperative agreements.
A deep, dedicated partnership (shared tooling, shared roadmap) beats a transactional one-off arrangement for both sides — but only if both commit. Either side settling for the arm's-length deal "to be safe" locks in the worse outcome for both.
Full integration (shared systems, shared headcount decisions) usually outperforms a loose holding-company structure, but only works if both legacy teams commit. Either side protecting its own turf "just in case the merger doesn't stick" reliably produces the shallower, weaker outcome.
A joint rewrite can produce a much better foundation, but only if both maintainers show up for the multi-week effort. If trust is shaky, each maintainer is individually better off patching around the old code alone — a smaller, safer outcome that locks in long-term technical debt.
This is an illustrative example designed to teach the method.
Expand for the origin of the idea and how it differs from more famous cooperation games.
The parable traces to Jean-Jacques Rousseau's 1754 Discourse on the Origin and Basis of Inequality Among Men. Rousseau described hunters waiting together for a stag who are tempted to abandon the group hunt the moment a hare passes by, since a hare can be caught alone and eaten immediately, while the stag requires everyone to stay at their post. The precise mathematical payoff structure used in modern game theory is a later formalization and differs somewhat from Rousseau's original telling, but the core tension — a better joint outcome that requires simultaneous commitment — is his.
The Prisoner's Dilemma has exactly one stable outcome: both parties defect, even though mutual cooperation would leave both better off, because each side can always improve its own result by betraying a cooperating partner. The Stag Hunt has two stable outcomes: both hunt stag (the best outcome for everyone) or both hunt hare (safer, but worse for everyone). Critically, if your partner commits to the stag, you do not improve your own result by secretly hunting hare instead — you do worse. The problem is coordinating on which stable outcome to land on, not resisting a temptation to cheat.
Game theorists call the stag-stag outcome "payoff dominant" (it's better for everyone if reached) and the hare-hare outcome "risk dominant" (it's safer to default to if you're unsure what your partner will do, since hunting stag alone while your partner hunts hare leaves you with nothing). Groups with low trust or no way to signal commitment tend to drift toward the risk-dominant, hare-hare outcome even when everyone would prefer the better one.
Philosopher David Hume described structurally similar problems: two people rowing a boat, or neighbors who must each drain part of a shared meadow for either of their efforts to pay off. Both require simultaneous effort from more than one party with no benefit to defecting on a committed partner. Researchers studying cooperation have also used the stag hunt structure to analyze coordination problems in fields from biology to international governance.
If there's a real incentive to secretly betray a committed partner for personal gain, that is a Prisoner's Dilemma, not a Stag Hunt. Treating a genuine betrayal risk as a mere trust problem will leave you exposed.
Naive, unconditional cooperation with no visible sign that the other side is also committing is not the same as trust — it is a gamble. A real signal (cost, publicity, a binding step) is what actually shifts the odds.
Not every joint effort has two stable equilibria. Check the actual incentives before assuming hesitation is the only barrier.
When a group lands on the smaller outcome because trust never solidified, the right diagnosis is usually a signaling failure, not bad faith.
Asking a partner to go first with no reciprocal signal from you is asking them to carry all the risk. Commitment usually has to be built in matched steps.
A single missed commitment can slide a group from the stag-stag equilibrium back toward hare-hare. Ongoing signals matter, not just the first one.
a party genuinely benefits from exploiting a cooperating partner. If betrayal is individually profitable even when the other side commits, you need a different model and different safeguards (binding contracts, reputation, repeated play).
the situation is a one-shot, adversarial negotiation over how to split a fixed pie rather than a joint effort to create a bigger one together.
one party can simply require the outcome (a manager mandating a process) rather than needing voluntary mutual commitment. Modeling it as a trust problem when it is actually an authority problem wastes effort.
Play several rounds against a simulated partner whose willingness to commit depends on what happened last round. Watch how trust builds, or collapses, over time.
Round 1. Your simulated partner is watching how the last round went. Choose your move.
This browser tool structures your thinking. It does not send your information anywhere or perform AI analysis.
Save as Skill converts the model into portable Markdown instructions that teach an AI when and how to apply the Stag Hunt, what questions to ask, what signals to look for, what mistakes to avoid and how to structure its answer.
Save the skills you find useful and organize them into your own reusable library. The Markdown is intentionally model-neutral, so it can be used wherever an AI system accepts uploaded files, project knowledge, custom instructions or reusable text context. Platform capabilities differ, so this is a portable skill file rather than a claim of one-click installation.
Because the risk isn't betrayal, it's miscoordination. Even with no temptation to cheat, you can still end up worse off simply by guessing wrong about what your partner will do. Trust and signaling exist to reduce that uncertainty, not to prevent exploitation.
Yes. Adding a real incentive to defect on a committed partner (for example, a side deal that rewards whoever defects first) turns a Stag Hunt into a Prisoner's Dilemma. Conversely, adding real punishment for defection (enforceable contracts, reputational cost) can turn what looks like a Prisoner's Dilemma back into something closer to a Stag Hunt, because it removes the payoff for betraying a committed partner.
No. If trust genuinely cannot be established, or the resources for the big outcome don't exist, hunting hare is the correct, rational choice. The model's lesson isn't "always go for the big win." It's "diagnose honestly whether the barrier is real resources, real betrayal risk, or just unresolved trust and signaling."