“You can be right about the headline and still lose.”
Who this guide is for
You traded a prediction market. The headline said you were right. The market resolved against you. You want to understand why — and prevent it from happening again. This guide explains oracle sources, resolution criteria, and gives you a 5-step checklist to audit any contract before you buy.
“Candidate X Won the Election”
This is what news coverage reports. It reflects the called result, the popular vote, or the electoral math — depending on the outlet.
⚠️ This is NOT what the market measures.
“Official certification by [DATE] per [SOURCE]”
The market's resolution criteria specifies a named data source and a precise trigger condition. The contract settles on that — not on the news cycle.
✓ Read this BEFORE you trade. It's in the Rules section.
A prediction market resolves based on its written resolution rules — not the headline, not common sense, not the obvious outcome. The resolution criteria is the contract. Read it first.
The oracle is the named authority that answers “did it happen?” It translates a real-world event into a contract settlement. Every prediction market contract relies on one.
The oracle concept, simply:
You and another trader disagree about whether an event happened. The oracle is the third party — pre-agreed in the contract — whose answer is final. No matter what you believe, no matter what the news says, the oracle's data determines the payout.
Different named sources
Platform A uses the AP call. Platform B uses the official state certification. Same election — different oracle — different resolution date.
Different trigger thresholds
Platform A resolves YES if GDP exceeds 2.5%. Platform B uses 3.0%. Same GDP print — different thresholds — different outcomes.
Different resolution windows
Platform A resolves on December 31. Platform B resolves when the event is officially confirmed. Same event — different timing — one resolves, one doesn't.
| Platform | Oracle / Who decides | Where to find rules | Dispute window |
|---|---|---|---|
| Kalshi | Kalshi internal markets team | Kalshi Market Rules (help.kalshi.com) | Contact Kalshi support; reference resolution rules if violated |
| Polymarket | UMA Optimistic Oracle (decentralized) | Polymarket Resolution Docs (docs.polymarket.com) | 2-hour challenge window on oracle.uma.xyz |
| IBKR / ForecastEx | ForecastEx internal settlement (CFTC-regulated rulebook) | ForecastEx Rulebook (data.forecastex.com) | Contact ForecastEx via IBKR support; reference rulebook |
Platform resolution details verified against primary sources. Subject to change — always check each platform's current help documentation before trading.
Run through this before buying any contract. It takes 2–3 minutes and can save you from a resolution that surprises no one except you.
Every regulated prediction market contract has a written rules document. Do not rely on the headline or market title — the contract text is the legal definition of what resolves YES or NO.
Look for a 'Rules,' 'Resolution Criteria,' or 'Rules Summary' tab/section on the market detail page. On Kalshi, it's in the Rules summary panel. On Polymarket, it's under the order book.
No rules document is accessible before you buy. Avoid any market where resolution criteria aren't publicly disclosed pre-trade.
The oracle is the named authority whose data decides YES or NO. This is not always 'the news.' It may be a specific government database, a named index, an official announcement channel, or a decentralized oracle contract.
Look for a 'Resolution source,' 'Data source,' or 'Oracle' field in the rules. It should name a specific source — not just 'official sources' generically.
Vague resolution sources like 'credible news' or 'official sources' without a named primary source give the platform discretion to resolve however they interpret events.
The market title summarizes the question. The rules define the precise trigger. A contract titled 'Will inflation exceed 3%?' may use CPI, PCE, or a specific sub-index — each would give a different answer for the same real-world inflation rate.
Read the YES and NO conditions in the rules document. Note: thresholds (above/below a specific number), timeframes (by end of month vs. end of quarter), and data vintage (first release vs. revised).
If you can imagine a scenario where 'you were right' but the market resolves NO due to a technicality, that technicality is in the rules. Read it before trading.
Markets have a resolution window. If the outcome is clear before the expiry date, the market may still not resolve until that date. Conversely, some markets resolve immediately when the trigger event occurs, which may happen faster than you expect.
Look for 'expiry date,' 'end date,' 'resolution date,' or 'close date' in the market details. Compare it to when the underlying data is typically released.
If you're trading a market that resolves based on a monthly government report but the report comes out after the market's expiry date, the market may resolve on incomplete data.
Well-written contracts define what happens when the outcome is genuinely ambiguous — the event didn't clearly happen or not happen. These clauses tell you who has discretion and under what conditions.
Look for language like 'in the event of ambiguity,' 'if the event is not definitively determined,' 'the exchange reserves the right,' or 'resolution committee will decide.'
Unlimited discretion clauses ('at the exchange's sole discretion') mean you have no recourse if you disagree with the resolution. Understand the dispute process before you trade a large position.
These are the patterns that catch traders who read the headline but skipped the rules document.
Act quickly — dispute windows are short and often non-negotiable once they close.
Step 1: Read the resolution announcement and the source data
Find the official resolution notice from the platform. Identify which specific source data they cited. Take a screenshot.
Step 2: Compare to the written rule text
Open the market's rules document (Rules summary / Resolution Criteria tab). Compare the stated source and trigger condition to what the platform used to resolve.
Step 3: Determine if the rule text was violated
There's a difference between 'I disagree with the interpretation' and 'the platform used a different source than the rules specified.' Only the latter is a legitimate dispute. If the rules gave the platform discretion, exercise your judgment before trading — not after.
Step 4: File a dispute via the platform's official channel
Use the platform's official dispute or support process within the dispute window. On Polymarket: submit a challenge on oracle.uma.xyz. On Kalshi: contact support with your documented evidence. Act fast — windows close.
Note: Prediction market disputes are rare but time-sensitive. Once a dispute window closes, most platforms treat the resolution as final regardless of circumstances. Document everything immediately.
Kalshi vs. Polymarket
How these platforms resolve the same event differently
Are Prediction Markets Legitimate?
Resolution integrity and platform accountability
Can Prediction Markets Change Their Rules?
Criteria locked vs. interpretation clarifiable — and what recourse you have
Why Prediction Markets Lose Trust
Resolution failures and what erodes user confidence
Odds vs. Outcomes: Model vs. Market
What each is actually measuring — and how to read the gap
Editorial note: Resolution process details verified against primary sources: Kalshi help documentation (help.kalshi.com), Polymarket developer docs (docs.polymarket.com), and ForecastEx rulebook (data.forecastex.com). Platform resolution processes may change; always verify current rules on the platform before trading. Dispute window durations verified from platform support articles as of June 2026.
