Think Like
Everyone Else
No safe collusion. No shortcuts. Just you and a prompt. Pick the answer you think most others will pick.
Play NowThomas Schelling
In 1960, economist Thomas Schelling posed a deceptively simple question: if two strangers had to meet in New York City with no way to communicate, where would they go? Most people said Grand Central Station at noon. He called these natural meeting points focal points and showed that coordination without communication is not only possible, it is predictable. Schelling won the 2005 Nobel Prize in Economics for his work on conflict and cooperation through game-theory analysis. This game is a playable adaptation of that idea, built around a fixed seed set of canonical coordination prompts.
Focal Points
In 1960, economist Thomas Schelling discovered that people can coordinate without communicating by gravitating toward answers that feel naturally prominent: focal points.
People can often concert their intentions or expectations with others if each knows that the other is trying to do the same.
Thomas C. Schelling, The Strategy of Conflict (1960)
What if winning meant
agreeing with strangers?
The Schelling Game puts you in a room with other players. Everyone sees the same Schelling prompt. Everyone picks an answer alone. Those who converge on the most popular answer bucket split the pot. The rest lose their ante.
No Safe Collusion
You can talk, but you cannot prove what you will pick. Anyone who tries to coordinate by pre-revealing their choice opens themselves up to betrayal. The mechanism is designed to make focal-point reasoning the most reliable approach.
Commit-Reveal
Answers are cryptographically locked before anyone reveals. You cannot simply wait to see what others chose, then copy them.
Plurality Wins
The most popular answer wins the game. If two answers tie for first, both groups win. Everyone else loses their ante.
10 Games
Each match is 10 games of different prompts. Over many games, consistent coordination is rewarded more than any single lucky guess.
Commit, reveal, normalize, settle
Commit
Read the prompt, pick your answer, lock it in with a cryptographic hash. 60 seconds. No take-backs.
Reveal
When commit closes, the reveal window opens. You submit the exact answer and salt behind your commitment so the hash can be verified.
Normalize
Open-text answers pass through an AI normalization step so equivalent signals such as NYC and New York can land in the same bucket.
Settle
The plurality wins. An ante is collected from all players, then redistributed to those who converged. Any indivisible remainder is burned.
What will the
crowd choose?
100 is the only non-prime, the only round number, and the only three-digit number in the set. When you reason about what others will pick, that salience makes it the natural coordination target: the focal point.
What is a Schelling Point?
A Schelling point (or focal point) is a solution people naturally converge on without communicating. It works because some answers feel more obvious, more prominent, or more "default" than others.
Schelling showed experimentally that people succeed at these coordination tasks far more often than chance would predict. The key insight: shared context and cultural salience create natural convergence points.
Why does this mechanism work?
The Schelling Game turns focal-point theory into a playable incentive structure. Four properties make it robust:
Plurality Wins
The answer picked by the largest group wins. Ties split the pot. The Nash equilibrium is to pick the most obvious choice, not to be clever or contrarian.
Blind Commitment
Commit-reveal cryptography means you lock your answer before seeing anyone else's. No copying, no last-second switches.
Truth as Equilibrium
When the most obvious answer is always the best strategy, "pick what's salient" becomes the dominant play.
Beyond the Game
This is exactly why the mechanism is useful in the real world: it produces consensus without requiring trust or a central authority.
Where Schelling games are used
Decentralized Price Feeds
Example: a derivatives protocol needs the ETH/USD price at exactly 4:00 PM UTC to settle expiring contracts. Independent reporters submit the price they observed, and the coordinated result becomes the onchain settlement value, as in UMA's oracle and DVM.
Outcome Resolution
Example: a market asks whether a merger closed before the deadline. Reporters review the filings, submit yes or no independently, and the majority answer resolves the market without a central judge. This is the same reporting and dispute pattern used by Augur.
Decentralized Courts
Example: a freelancer and client dispute whether a milestone was actually delivered. Jurors review the contract, messages, and submitted work in isolation, then vote; jurors who align with the majority are rewarded. Kleros Escrow applies this model to real dispute resolution.
From theory to cryptographic mechanism
In 2014, Vitalik Buterin proposed SchellingCoin: the first design for using Schelling-point incentives on a blockchain. The idea was simple: if you reward participants for reporting values that match the majority, honest reporting becomes the dominant strategy. This concept underpins oracle designs, curation markets, and governance systems across Ethereum where decentralized truth is needed without a trusted authority.
Further Reading
- Thomas Schelling — The Strategy of Conflict 1960
- Mehta, Starmer & Sugden — "The Nature of Salience: An Experimental Investigation" 1994
- Roger Myerson — "Learning from Schelling's Strategy of Conflict" 2009
- Vitalik Buterin — SchellingCoin 2014
- Paul Sztorc — "Truthcoin" whitepaper 2014
- Augur — Schelling games for prediction markets 2015
- P+ε Attack — a known bribery vulnerability in Schelling-point mechanisms 2015
- Kleros Whitepaper — commit-reveal voting for decentralized dispute resolution 2019
- UMA Optimistic Oracle — Schelling-point dispute resolution in production DeFi 2020
- AI Alignment Forum — "Schelling game evaluations for AI control" 2024
- Aharon et al. — "Tacit Coordination of Large Language Models" 2026
Ready to find the focal point?
Any crowd from 3 to 21 players. 10 games. One prompt: can you think like everyone else?
Play Now