Guide

How prediction markets resolve and settle

A prediction market is a contract that pays out based on a real-world outcome, so the moment that decides who gets paid is the resolution. This page explains how that works: the settlement math where each winning share pays $1 and each losing share pays $0, where the "truth" of an outcome comes from, and how the two largest venues differ. Kalshi resolves through its own rules and named source agencies under U.S. derivatives regulation, while Polymarket resolves on-chain through the UMA optimistic oracle with a public dispute window. We also cover what happens when an outcome is ambiguous or contested, why those cases have drawn scrutiny, and why the single most useful habit a reader can build is checking a specific market's rules before trusting its price. None of this is financial advice; market prices are crowd-implied probabilities that can be, and sometimes are, wrong.

The settlement mechanism: $1 for a winning share, $0 for a loser

Most prediction-market contracts are binary. A question is framed as a yes/no proposition, and a share in the correct outcome settles at $1 while a share in the wrong outcome settles at $0. That fixed payout is what lets the price read as a probability: a contract trading around $0.63 implies the crowd is putting roughly a 63% chance on that outcome happening, because $0.63 is what a $1 payoff is worth if it pays off about 63% of the time.

Payout is not the same as profit. The $1 settlement is fixed by the contract; the profit depends on the price paid to get in. Buying a share at $0.62 that later settles at $1 returns a $0.38 gain per share before fees, while the same share settles at $0 and is a total loss if the other outcome occurs. Some markets are multi-outcome (for example, several candidates in one race), but they are typically built from the same logic: exactly one outcome resolves to $1 and the rest to $0.

The price moving toward $1 or $0 before an event is decided does not mean the question has resolved. Settlement happens only when the resolution criteria are met and the outcome is confirmed against the designated source. Until then, a price near $0.95 reflects high crowd confidence, not a finished result, and confidence can still be misplaced.

Where the answer comes from: resolution sources and rules

Every market resolves against a defined source of truth, and the quality of that definition matters more than the headline question. Well-written contracts name exactly what will be consulted: an official result (an election authority certifying a winner, a sports league's final score), a designated source agency (a government statistics office, a regulator, a court record), or a specific named data feed (an economic data release, a published price index). The rules also specify the exact date, time, and condition that count, and what happens in edge cases such as a postponement, a tie, or a cancelled event.

This is why two markets that look like they ask the same thing can settle differently. The truth conditions live in the rules, not the title. A market that resolves on "officially announced by [a named body] by [a named date]" can pay out differently from one that resolves on "reported by major news outlets," even when both are nominally about the same event. Wording around timing, sourcing, and what qualifies as the triggering event is where most genuine disputes begin.

For a reader, the practical takeaway is that the resolution source and rule are the contract. Crowdtells reads markets as a signal of what people think will happen and what matters, but the number only means what the rules say it means. Before treating a price as informative, it is worth knowing which source decides it.

How Kalshi resolves: exchange rules and source agencies under CFTC oversight

Kalshi is a CFTC-regulated derivatives venue. It operates as a Designated Contract Market (DCM), a category of exchange overseen by the U.S. Commodity Futures Trading Commission, which has regulated U.S. derivatives markets since 1974. Within that framework, Kalshi lists event contracts and self-certifies their terms with the regulator, meaning the contract specifications, including how each one resolves, are filed rather than improvised after the fact.

Resolution on Kalshi is centralized. Each contract's terms reference the source agency or underlying data whose value at the stated expiration determines the outcome, and Kalshi's market operations team confirms settlement once the criteria are met. Traders can flag a market they believe is ready to settle, but that flag is a prompt, not a binding instruction; the exchange controls the final determination. Kalshi's rulebook also provides for a market outcome review process the exchange can invoke before settling in cases that need closer examination.

The trade-off is straightforward. A single accountable operator under regulatory supervision can resolve quickly and apply consistent, pre-filed rules, but it also concentrates the resolution decision in one party. That has not made Kalshi immune from controversy; specific settlements have drawn user complaints when the contract wording and the messy real world did not line up cleanly, which is the same ambiguity problem every venue faces.

How Polymarket resolves: the UMA optimistic oracle and the dispute window

Polymarket settles on-chain and outsources the "what actually happened" question to the UMA optimistic oracle, a decentralized, permissionless system. The word optimistic captures the design: once a market's event window closes, anyone can propose the outcome by posting a bond (commonly around $750), and that proposed answer is assumed correct unless someone challenges it. There is a short challenge or liveness window, typically about two hours, during which any participant can dispute the proposal by posting a matching bond.

If no one disputes within the window, the market resolves to the proposed outcome, the proposer recovers the bond plus a small reward, and winning shares become redeemable for $1 each while losing shares go to $0. If the proposal is disputed, it escalates to UMA's Data Verification Mechanism (DVM), where UMA token holders vote on the correct outcome over a period that runs into days. The side the vote agrees with recovers its bond and is rewarded from the loser's forfeited bond, which is the economic incentive meant to keep honest answers cheap to defend and bad answers expensive to push.

The strength of this model is that it is open and adversarial: anyone can contest a wrong answer rather than relying on one operator. The weakness is that final say in a contested case rests with token-holder voting rather than a named official source, and the outcome turns heavily on how the market's rules were written. Specific bond sizes, window lengths, and timings can change, so the current parameters should be read from the live market and UMA's documentation rather than memorized.

When resolution is contested, and why the rules page is the thing to read

Most markets resolve uneventfully. The hard cases are the ones where reality does not map cleanly onto the contract: an event partly happens, an official source is slow or silent, a key word in the rules is read two ways, or the triggering condition is arguable. In those situations the resolution mechanism, centralized review on Kalshi, or proposal-and-dispute on Polymarket, becomes the whole story, and the eventual payout can hinge on a single clause.

Contested resolutions have drawn real scrutiny. High-value markets on geopolitical and financial questions have entered formal disputes, a notable case in 2026 centered on how a market about a corporate Bitcoin sale should resolve, and the broader pattern of disputed and unclear outcomes has attracted attention from commentators and at least one U.S. senator. The recurring theme is not fraud so much as ambiguity: when wording leaves room, participants with opposite positions can each believe they are right. These specifics will keep changing, so treat any particular case as illustrative rather than definitive and check a live source.

The defensive habit is simple and applies to every venue. Before you trust a price, read the market's own resolution rules: what the exact question is, which source decides it, by what date and time, and how postponements, ties, or cancellations are handled. A price is only as trustworthy as the rule behind it. Crowdtells treats these markets as a money-weighted probability estimate of what may happen, useful as a signal, not a guarantee, and nothing here is financial advice.