← All posts · July 9, 2026 · 7 min read

How Polymarket resolves markets (and what happens when it goes wrong)

You bought Yes at $0.62 and the thing happened. Who decides you get paid?

On Polymarket the answer depends on the market. Event markets — elections, sports, "will X happen by Y" — are resolved by people posting money behind an answer, through a system called the UMA optimistic oracle. The crypto Up/Down markets are resolved by an automated price feed with no humans in the loop. Two machines, two very different failure modes, and it pays to understand both before your money rides on one.

Event markets: the optimistic oracle

"Optimistic" means the system assumes a proposed answer is honest unless someone pays to disagree. Per Polymarket's resolution docs, the flow works like this:

• Once the outcome is knowable, anyone can propose it — Yes or No — by posting a bond, typically $750.
• A 2-hour challenge window opens. Anyone who thinks the proposal is wrong can dispute it by posting an equal counter-bond.
• No dispute? The market resolves in about two hours and the proposer gets the bond back plus a reward.
• Disputed? The question can go to a second proposal round — and a second dispute escalates it to UMA's Data Verification Mechanism, where UMA token holders vote on the answer. That path takes roughly 4 to 6 days, including a debate period for evidence and about 48 hours of voting.
• Whoever the vote sides with gets their bond back plus half the loser's bond.

The design is simple game theory: proposing a wrong answer costs $750, so most markets resolve quietly and correctly. Most.

What resolution does to your shares

The moment a market resolves, trading stops. Winning shares become redeemable for $1.00 each; losing shares are worth nothing. Redemption is just claiming — the outcome is already locked.

Worth repeating for beginners: you never have to wait for any of this. A share bought at $0.62 can be sold at $0.90 whenever the market gets there. Resolution only matters for whatever you're still holding at the end — which is exactly why some traders sell at $0.90+ instead of holding through the machinery below.

When it goes wrong: the $7 million Ukraine market

In March 2025 a Polymarket market asked whether Ukraine would agree to a minerals deal with Trump before April. About $7 million was traded on it. On March 24–25 the odds surged from 9% to 100%, and the market resolved Yes — despite no official agreement existing.

How? A large UMA token holder threw their voting weight behind the Yes resolution during the dispute vote. Polymarket itself acknowledged "the market was resolved too soon, as no deal between Ukraine and the U.S. has been confirmed" — and then declined to refund anyone, on the grounds that a valid oracle vote isn't a "market failure." People holding No, who were factually right, got zero.

Polymarket promised clearer rules and process changes afterward, and disputes on this scale are rare. But the lesson stands: resolution risk is its own category of risk, separate from being right about the world. The rules text and the oracle decide, not the news.

Crypto markets skip the humans entirely

The BTC 5-minute and 15-minute Up/Down markets work on a completely different rail. There's no proposer, no bond, no vote. Settlement runs on automated BTC/USD data from Chainlink's high-frequency oracle infrastructure: the system compares the price at the start of the window to the price at the end, and the winning side is paid out within moments of the round closing.

If you're new to these, our 5-minute markets explainer covers the format. For resolution purposes the key point is what you gain and what you trade away. You gain certainty — no committee can vote your winning position into a loser, and there's nothing to dispute about a price print. Settlement risk in the human sense is gone.

What you inherit instead is the oracle print itself. The Chainlink number is the truth these markets settle on, even when it differs by a few dollars from the exchange chart you're staring at. And because a spot price can be nudged in the closing seconds, the print has become a target — researchers have documented traders doing exactly that, which we covered in is the 5-minute market manipulated.

A practical checklist

What I actually do before putting real size into a market, and what I'd suggest to anyone:

• Event markets: read the full rules text, identify the resolution source, and skip markets whose wording feels ambiguous. Ambiguity is where disputes live.
• Cheap "longshot" shares near a deadline: be careful. If the market needs official confirmation and the deadline is close, you're betting on paperwork speed, not the event.
• Crypto Up/Down: trust the print, not your chart. And don't hold marginal positions into the final seconds of a round.
• Everything: remember you can sell early. The last few cents before $1.00 are payment for carrying resolution risk — sometimes that's a fair trade, sometimes it isn't.

I've seen traders be dead right about the event and still lose the market. The mineral-deal crowd holding No is the canonical example — being right pays $1 only when the oracle agrees.

What this means if you copy trade

A wallet you copy carries its resolution risk into your account. If they hold everything to settlement, so do you — unless you set your own exits. That's why btc5min gives each copied wallet its own rules: exit when they exit, or run an auto take-profit ladder that sells a $0.93 favorite at $0.90 and lets someone else carry the last three cents through the machinery. Understanding what those prices mean makes that trade-off easy to reason about.

btc5min is an independent tool and is not affiliated with, endorsed by, or operated by Polymarket. Nothing here is financial advice. Prediction markets are risky, copying another trader does not guarantee profit, and you can lose the money you trade with.