The Resolution Source Problem: Why How a Market Settles Matters More Than What It Asks
The question in the title isn't the question you're actually betting on. The resolution source is.
The Question Is Not the Question
Every prediction market has two layers. There's the surface question — *Will Trump win the popular vote?* *Will Fed cut rates in September?* *Will Taylor Swift attend the Super Bowl?* — and then there's the actual contract language: the resolution source, the deadline, and the tiebreakers.
Most traders bet on the surface question. Sharps bet on the contract language. That gap is where a huge amount of edge lives, and where a huge amount of money gets vaporized by people who thought they were right.
The Three Layers of Resolution Risk
Every market resolution depends on three things: what source decides, when it decides, and what happens if the source is ambiguous or missing.
A market resolving on "official announcement from the Federal Reserve" is not the same as one resolving on "news reports of a Fed decision." A market closing at 11:59 PM ET is not the same as one closing at market close. And a market with no clear tiebreaker is a lawsuit waiting to happen — except there's no lawsuit, just a UMA vote or a Kalshi ruling you can't appeal.
Read the rules. Then read them again. Then check if the rules have been edited since the market opened.
The Polymarket UMA Wrinkle
Polymarket resolves via UMA's optimistic oracle — token holders vote on the outcome. Most of the time this works fine because most markets are unambiguous. But when a market gets weird, UMA voters interpret the rules, and their interpretation is final.
This has produced some famously ugly resolutions. Markets that seemed to clearly resolve one way have flipped based on technicalities. Markets with ambiguous language have gone to the side with more voting power, not the side with the better argument.
Before you take a big position on a Polymarket contract, ask yourself: if this resolves ambiguously, which side will UMA favor? The answer is usually "the side that's easier to defend on a strict reading of the rules," not "the side that matches the spirit of the question."
The Kalshi Version
Kalshi resolves internally, which means their compliance team is your oracle. They're generally careful and predictable, but they have their own quirks. They lean on specific named sources (BLS reports, official government data, specific news outlets) and they interpret their rules literally.
If a market says "according to the BLS employment report released on X date," and the BLS delays the report, Kalshi doesn't guess. The market either extends or voids depending on the specific rule language. Know which one before you put money down.
The Trade
Here's the actionable part. When you find a market where the surface question and the resolution language diverge, you have edge — because most traders are pricing the surface question.
Example: a market asks "Will Country X invade Country Y by December 31?" but the resolution requires *official U.S. State Department designation as an invasion*. That's a much higher bar than the surface question implies. If the market is priced at 40¢ based on "will there be conflict," but the actual resolution requires a specific diplomatic declaration, the true probability might be 15-20¢. That's a fadeable spread.
The reverse also works. Sometimes the resolution language is *looser* than the surface question suggests, and the No side is mispriced because everyone's imagining a stricter standard.
The Habit to Build
Before every trade over a size that would ruin your week, do this: open the rules, read them out loud, and articulate the exact scenario that would flip the resolution against you. If you can't name that scenario, you don't understand the market well enough to be sized where you are.
The question in the title is marketing. The rules are the contract. Trade the contract.
Our Calls From This Date
These are the exact picks our AI flagged on August 31.
Will the Democratic Party candidate win the 2026 NY-14 House election by 65% or more?