Resolution is the mechanism by which a prediction-market contract is settled at the conclusion of the event it references. The pricing of every contract during its life is anchored to the expectation of resolution, so the mechanics by which the resolution decision is made are central to how the markets work. Polymarket and Kalshi use fundamentally different resolution architectures. This article describes both systems in detail, addresses the common edge cases, and explains the practical implications for traders and copy-traders.
The general structure of a prediction-market contract
Every prediction-market contract has a written rule that specifies the question to be resolved, the resolution source, and the resolution date. A typical example might read as follows: the contract resolves YES if the Bureau of Labor Statistics reports nonfarm payrolls of 180,000 or higher for the reference month, as published in the initial release; otherwise NO. Resolution date is the first Friday of the following month. The rule is the contract; the market price during the life of the contract is an estimate of the probability that the rule will resolve YES on the resolution date.
The clarity of the rule is the single most important factor in whether a market resolves cleanly. Rules that depend on subjective judgments, like whether a public figure has formally conceded a contest, are more likely to produce contested resolutions than rules that depend on specific numeric thresholds published by named sources. Traders who read the rule carefully before entering a position avoid most of the resolution risk that catches less careful participants.
Polymarket: UMA-based resolution
Polymarket uses an external oracle called UMA, short for Universal Market Access, to resolve its contracts. UMA is an independent project that provides oracle services to a range of decentralized protocols, of which Polymarket is one of the largest users.
The resolution process on Polymarket follows a defined sequence. When a market reaches its resolution date, a UMA proposer submits a proposed answer, typically YES, NO, or invalid. The proposed answer is accompanied by a bond that the proposer forfeits if the proposal is successfully disputed. There is then a dispute window, typically two hours but configurable per market, during which any participant can post a counter-bond to challenge the proposal. If no dispute is filed, the proposal is accepted and the market resolves accordingly.
If a dispute is filed, the question goes to a vote by UMA token holders. The voters review the evidence on both sides and submit a binding answer within a window that is typically forty-eight to seventy-two hours. The side of the dispute that loses the vote forfeits its bond, which discourages frivolous disputes and creates a positive incentive for accurate proposals.
In practice, the vast majority of Polymarket markets resolve without dispute because the underlying answer is unambiguous. The dispute mechanism becomes relevant for a narrow class of contested markets where the rule wording and the realized event diverge in important ways. The most prominent disputed markets in Polymarket's history have involved questions of timing (did an event occur before midnight UTC on the reference date), interpretation (does a partial event satisfy the rule), or source availability (the named source did not publish on time).
Kalshi: source-based resolution
Kalshi uses a fundamentally different resolution model. Each contract has a specified external source, named in the rules document at the time the contract is listed, and the contract resolves automatically to the value published by that source on the resolution date. There is no proposer, no dispute window, and no community vote.
The source for a given Kalshi contract is typically a government agency, an industry body, or another third-party publication. A contract on the December rate decision resolves to the Federal Open Market Committee announcement; a contract on the September CPI release resolves to the Bureau of Labor Statistics publication; a contract on a sports outcome resolves to the league's official scoring authority. The choice of source is binding for the life of the contract.
Disputes do occur on Kalshi but they go through the platform's internal compliance process rather than a public vote. The dispute resolution is overseen by the same regulatory framework that applies to designated contract markets generally; the platform must demonstrate to the CFTC that its resolution decisions are consistent with the rules of the listed contract. The result is a more predictable resolution process at the cost of less in-flight transparency than the UMA model provides.
Invalid resolutions
Both platforms have a mechanism for resolving a contract as invalid when the underlying event does not occur in a way that matches the rule. The most common reasons for an invalid resolution are that the named source did not publish, that the underlying event was canceled or postponed past the resolution window, or that the rule wording does not cleanly apply to the realized facts.
On Polymarket, an invalid resolution is one of the three options available to the UMA proposer. If a market is resolved as invalid, the contracts on both sides are refunded at a price determined by the market mechanics. In most cases the refund approximates the entry cost, but traders who entered at very different prices may see different refund treatments depending on the mechanics of the specific market.
On Kalshi, invalid resolution is handled through the platform's compliance process. The contracts are typically refunded at the last traded price prior to the invalidating event, although the platform retains discretion to apply a different refund treatment if the circumstances warrant it.
Invalid resolution sounds superficially like a free option, but the market embeds the small probability of invalid resolution into the price during the contract's life. Traders cannot reliably profit by intentionally seeking out markets likely to resolve invalid; the prices already reflect the risk.
How Rivo uses resolution data
Rivo records every whale-sized trade across both platforms as it occurs. When a market resolves, the recorded trades on that market are scored against the resolution outcome. A buy at $0.34 on a contract that resolves YES is recorded as a paper gain of $0.66 per contract; the same buy on a contract that resolves NO is recorded as a paper loss of $0.34 per contract.
The scoring uses the entry price and assumes the position was held to resolution. Some whales exit earlier, either to take profits or to cut losses; Rivo does not attempt to reconstruct those interim decisions. The paper PnL figure is therefore an upper bound on the realized return for a position taken at the entry price, which is the consistent way to rank trades across platforms and across traders.
Sell trades are excluded from the win-rate calculations because a sell is an exit rather than a directional opening position. Including sells in the win-rate statistics would conflate exits with new bets and contaminate the signal that a copy-trader is trying to extract from the data.
The role of resolution timing
A market that resolves in two hours and a market that resolves in two months are structurally different instruments even when the implied probability is identical. The short-dated market behaves more like a traditional bet, with the outcome largely determined by a single event near resolution. The long-dated market behaves more like a position, with the outcome determined by an accumulation of information and narrative across the holding period.
Whales tend to perform differently on short-dated and long-dated markets. Short-dated markets reward speed and information edge; long-dated markets reward conviction and the willingness to hold through adverse moves. A wallet that is profitable on short-dated sports markets may not be profitable on long-dated political markets, and vice versa. The category-fit analysis that Rivo uses elsewhere on the site is partly an attempt to capture this difference in trader skill across resolution durations.
For copy-traders, the practical implication is that copying a long-dated position requires committing to hold through what may be a multi-month period of price movement that diverges from the eventual resolution. Traders who cannot tolerate that exposure should concentrate on short-dated markets where the time between entry and resolution is short enough to manage emotionally.
Frequently asked questions
What happens if a Polymarket market is disputed?
The question goes to a vote by UMA token holders, who submit a binding answer within a window of typically forty-eight to seventy-two hours. The side of the dispute that loses the vote forfeits the bond it posted to initiate the dispute. Contracts in a disputed market are not tradeable until resolution is finalized.
How often are Polymarket markets disputed?
A small minority of markets see formal disputes; the majority resolve cleanly. The disputed cases are concentrated in markets with ambiguous rule wording or with events that occurred close to the resolution boundary. Traders who avoid contracts with subjective resolution criteria can largely avoid dispute risk.
Can Kalshi change a contract's resolution source?
No. The resolution source is fixed at the time the contract is listed and cannot be changed without invalidating the contract. If the named source becomes unavailable, the platform's compliance process determines whether the contract resolves to an alternative figure or is treated as invalid.
What does it mean for a market to resolve "invalid"?
The underlying event did not occur in a way that maps cleanly to the contract's rule, and the platform's resolution process has determined that neither YES nor NO is the correct answer. Contracts on both sides are refunded according to the platform's invalid-resolution policy, which approximates the entry cost in most cases.
How is paper PnL different from realized PnL?
Paper PnL assumes the position was held from entry to resolution at the entry price. Realized PnL is the actual profit or loss after accounting for any interim exits, partial closes, or position adjustments. Rivo publishes paper PnL because it is the consistent measure that can be calculated from the trade tape alone. The relationship between paper PnL and the scoring methodology is described above. For more background, see are Polymarket whales actually profitable.