How prediction markets resolve

Every prediction market ends the same way: someone has to decide what actually happened, and winning shares have to be paid. That step — resolution — is where prediction markets earn or lose their credibility, and it is worth understanding before you trade or lean on a market’s price.

The written rules govern, not the headline

A market’s question (“Will X win?”) is shorthand. What actually settles it is the resolution text: a written rule stating precisely what counts, by when, and according to which source. On Polymarket, whose markets Knoww tracks, this text plus any named resolution source — an official announcement, a government data release, a sports governing body — is the contract. When a result feels ambiguous, the rules text decides, which is why experienced traders read it before the order book. Knoww shows each market’s rules and resolution source on its event page.

“Closed” and “resolved” are different states

When a market’s end conditions are met, trading closes — prices stop moving and Knoww labels the event with its final odds. But closing is not settlement. Between close and payout sits a verification step, so a market can be closed for hours or days while its outcome is confirmed. Only after that does it become resolved, when shares actually redeem: $1 for the winning side, $0 for the losing side.

How outcomes are verified: the optimistic oracle

Polymarket markets settle on-chain, so the outcome has to be reported to the blockchain by something trustworthy. Most markets use UMA’s optimistic oracle, which works like a challenge system rather than a referee:

• Proposal. After the outcome is knowable, anyone can propose a result, posting a monetary bond behind it.
• Challenge window. The proposal then sits open for a dispute period. If nobody objects, it is accepted — the “optimistic” part — and the market resolves.
• Dispute. If someone posts a counter-bond, the question escalates to a vote of UMA token holders, who settle it per the market’s rules. The wrong side loses its bond, which is what makes false proposals and frivolous disputes expensive.

In practice, clear outcomes resolve quickly and quietly; disputes are the exception and add days when they happen. The system’s integrity rests on incentives — being wrong costs money at every step — rather than on trusting a single referee.

Edge cases to expect

• Early resolution. Many markets can settle before their end date once the outcome is locked in — a candidate clinches, a bill passes, a match ends.
• Deadline passes, nothing happens. “By date” markets resolve No when the date arrives without the event. A price drifting toward zero as a deadline approaches is this mechanic at work, not fresh news.
• Ambiguity. Occasionally reality outruns the rules text — a postponed event, a renamed metric, a disputed announcement. Resolution then turns on close reading of the rules, and prices can swing hard while traders argue about interpretation.

What this looks like on Knoww

Knoww labels events by trading state: live events show current odds, and closed events show the final odds at the time trading ended. Because settlement can lag the close, a closed market’s last price is the market’s final estimate, not always the official outcome — the definitive result is the resolution itself. Knoww displays this data but plays no part in deciding outcomes: resolution happens entirely on Polymarket and its oracle.

For the fundamentals behind all of this, start with what is a prediction market, learn the numbers in how to read prediction-market odds, or watch live markets heading toward resolution on the markets feed and category pages like geopolitics.