Polymarket is a prediction market: a venue where people trade on whether future events will happen. Instead of a bookmaker setting odds, users buy and sell shares in outcomes, and the price of a share is the crowd's estimate of the probability. This guide explains what you are actually buying when you trade, how prices are formed, who is on the other side of your trade, how a market is settled once the event is over, and what the legal picture looks like. It assumes no prior knowledge of prediction markets, but it does not talk down: if you already bet on sports, most of this will map onto things you know.
What Polymarket is
Every question on Polymarket is a market with a small number of possible outcomes. The simplest kind is binary: "Will X happen by date Y?" with a YES outcome and a NO outcome. Each outcome is represented by a share that pays exactly $1.00 if that outcome is the one that happens, and $0 otherwise. Prices are quoted in cents between 0¢ and 100¢, so a YES share trading at 64¢ costs $0.64 and returns $1.00 if the answer turns out to be yes: a profit of 36¢ per share.
That single design decision, one dollar per winning share, is what makes the price readable as a probability. A share that pays $1.00 with probability p is worth p dollars to a risk-neutral buyer, so if traders collectively think an outcome is 64% likely, the share settles around 64¢. There are no separate "odds" to interpret. The price is the odds.
Larger questions are grouped as an event: "Who will win the election?" is an event containing one binary market per candidate, each with its own YES and NO shares. Exactly one of them resolves YES, so in a well-traded group the YES prices add up to about 100¢, in the same way a bookmaker's outright market adds up to 100% plus margin.
Under the hood, Polymarket runs on the Polygon blockchain. Deposits are held in the stablecoin USDC, and shares are tokens in your wallet. You do not need to understand the blockchain to trade, but two consequences matter: your position is an asset you hold, not a bet slip the house holds for you, and money moves on-chain when you deposit, redeem and withdraw, which is where network costs show up.
What happens when you trade
A trade on Polymarket looks like this from your side:
- Deposit. You fund an account with USDC, either directly from a crypto wallet or through the card and bank on-ramps the site offers. The balance is your collateral.
- Pick a market and a side. Buy YES if you think the outcome will happen, NO if you think it will not. Buying NO at 36¢ is the same position as selling YES at 64¢; both pay $1.00 if the answer is no.
- Choose a price. A market order buys immediately at the best price currently offered. A limit order names your price and waits until someone takes the other side. Limit orders can sit unfilled indefinitely; market orders fill now but pay the spread.
- Hold, or sell early. Nothing forces you to wait for the event. Shares can be sold back into the market at any time, so a position bought at 40¢ can be sold at 55¢ next week for a profit that does not depend on the final answer. This is the largest practical difference from a sportsbook bet, which is locked until settlement.
- Redeem. Once the market resolves, winning shares are redeemed for $1.00 each, and losing shares are worth nothing. Redemption is a transaction you trigger; the money then sits in your balance until you withdraw it.
Behind step 3 is an order book, the same structure a stock exchange uses. Polymarket's book matches orders off-chain for speed and then settles matched trades on-chain, so the experience feels like a normal trading app but the outcome is a token transfer.
How the pricing works
There is no house setting the price. At any moment a market has a bid (the most anyone is willing to pay for a share) and an ask (the least anyone is willing to sell for). The headline price you see on a market card is usually the midpoint or the last trade, and neither of those is what you get. Buying takes the ask; selling takes the bid.
The gap between them is the spread, and it is the market's equivalent of a bookmaker's margin. On a liquid market about an election or a major sports event the spread can be a single cent; on a thin market about an obscure question it can be five or ten. A one-cent spread near 50¢ is far tighter than the standard -110 / -110 line at a sportsbook, which is one reason arbitrageurs watch prediction markets closely.
YES ask + NO ask − 100¢ = spread paid to buy both sides 64¢ + 38¢ = 102¢ → 2¢ spread
YES and NO are bound together by a mechanical rule: one YES share plus one NO share can always be split from, or merged back into, exactly $1.00 of collateral. That is why the two prices track each other so closely. If YES trades at 64¢ and NO at 34¢, anyone can pay $0.98 for a pair guaranteed to return $1.00, and they will keep doing so until the gap closes. On a healthy market YES plus NO sits at 100¢ plus the spread, never much below.
Prices move because orders move. When news arrives, holders lift their asks and buyers lift their bids, and the midpoint shifts. Nobody adjusts a line; the line is the sum of everyone's orders. This makes prediction market prices faster than most sportsbook lines on breaking news, and also more jittery on thin markets where a single large order can move the price several cents.
Two costs sit outside the quoted price. The first is slippage: a large order does not fill entirely at the best ask, it walks up the book and pays a worse average price as each level is exhausted. The second is fees. Polymarket's fee schedule has changed over time and is not the same for every market category, so this site does not state a rate; check the fee on the market you are trading before relying on any number.
The conversion from a cent price to decimal, American or fractional odds is a single division, and the Polymarket odds converter does it in both directions. The Polymarket prices as odds guide goes further into bids, asks, depth and fee-adjusted odds; this guide stays on what the price means.
What the odds tell you
A Polymarket price is a probability estimate produced by people risking money on it. That is worth more than a poll or a pundit's opinion, because a trader who thinks the price is wrong can profit by moving it, and traders who are consistently wrong run out of money. But it is an estimate, and reading it well means knowing what is inside the number.
It is the crowd's probability, including the spread. On a market at 64¢ / 38¢ the fair probability is not 64% but somewhere in the middle: roughly 64 ÷ (64 + 38), about 63%. On liquid markets the correction is a fraction of a point; on thin ones it can be several points, and the midpoint is the better single estimate.
It contains the value of time. A share at 90¢ on something nine months away pays 11% if it wins, and the market is also pricing the fact that the money is locked up until then. Long-dated markets tend to trade a little further from 0¢ and 100¢ than the underlying probability alone would justify, because holding a near-certain outcome for a year is a poor return on capital. Short-dated markets do not have this discount.
Longshots are overpriced, favourites underpriced. The same favourite-longshot bias seen at racetracks and sportsbooks appears on prediction markets: a share at 3¢ is usually worth less than 3%, because buyers of lottery tickets outnumber sellers who will tie up capital for a small gain. Treat prices under about 5¢ and over about 95¢ with more scepticism than prices in the middle.
Volume and liquidity are part of the reading. A price backed by millions of dollars of trading and a deep order book reflects many independent opinions. A price on a market with a few hundred dollars of volume reflects one or two people's guesses and can be hours stale. Every market shows its volume and its order book; look at both before treating the price as information.
A move tells you something happened, not necessarily what. A jump from 40¢ to 55¢ means buyers arrived. Sometimes that is news, sometimes it is one trader with a view, and sometimes it is a hedge unrelated to the question. Prices on Polymarket are widely quoted by journalists as "the market thinks", and they are a reasonable summary, but the size of the market behind a move is the difference between a signal and noise.
The price is a forecast, not a promise. A 90¢ share loses one time in ten if the market is well calibrated. Ten-percent events happen constantly. Nothing about a high price makes an outcome safe; it only makes the payout small.
Reading one market
A market asks whether a bill passes by year end. YES bid 41¢, ask 43¢, NO bid 56¢, ask 58¢, volume $2.4M, book deep on both sides. The midpoint is 42¢, so the market's probability is about 42%, and the spread is 1¢ (43 + 58 − 100). This is a price you can trust as a forecast. If instead the volume were $900 and the last trade was 41¢ four days ago with 50 shares on each side, the number would be a placeholder, not a forecast.
Who trades on Polymarket
The person on the other side of your trade is one of a handful of types, and knowing which one helps you judge whether the price is beatable.
Opinion traders. The largest group by headcount. They buy the side they believe in, often in small amounts, often on political or cultural questions they follow anyway. They supply most of the volume on headline markets and most of the mispricing on emotional ones, where partisan buyers push a favoured candidate's price above the polls.
Market makers. Traders and firms who post limit orders on both sides of a market, earning the spread and, on many markets, liquidity rewards that Polymarket pays for resting orders near the midpoint. They rarely have a strong view on the outcome; they profit from the difference between bid and ask and from being paid to provide depth. Their presence is what makes a market tradeable, and their absence is what makes a market thin.
Sharp forecasters. People who model the question seriously: election modellers, sports bettors with their own numbers, people who track court dockets or legislative calendars. They trade when the price differs from their estimate by more than the spread and fees, and they are the reason the prices on liquid markets are hard to beat. On a big political market you are usually trading against them, not against opinion traders.
Arbitrageurs. Traders who look for prices that are inconsistent with each other rather than wrong in absolute terms: a group of candidate markets whose YES prices sum to less than 100¢, a YES / NO pair that sums to less than 100¢, or a Polymarket price that disagrees with a sportsbook or with Kalshi on the same event. The group arbitrage calculator is built for the first case and the Polymarket vs sportsbook arbitrage guide for the last. Arbitrageurs keep related prices consistent, which is part of why the price is worth reading in the first place.
Hedgers. Anyone with real-world exposure to an outcome: a business that suffers if a tariff passes, a farmer exposed to a weather event, a token holder exposed to a regulatory decision. They buy insurance rather than express a view, and they are often willing to pay a price above fair probability for it. Hedging flow is one source of the favourite-longshot bias.
Insiders and the informed. Some traders know something the market does not. A staffer who knows an announcement is coming, a person close to a negotiation, a player in a sports market. Prediction markets differ from securities markets in that trading on non-public information is not, in general, illegal on the same terms as insider trading in stocks, though it may breach the platform's rules and other laws depending on the information and the jurisdiction. Practically, sudden one-sided volume shortly before a resolution is sometimes the informed at work, and it is a reason to be cautious about betting against a large move you cannot explain.
Bots. Automated traders that quote both sides, arbitrage between related markets, or react to data feeds faster than a person can. They are why a price on a liquid market updates within seconds of news, and why a stale limit order on a thin market gets picked off the moment it becomes wrong.
How settlement works
Settlement, which Polymarket calls resolution, is where prediction markets differ most from betting on a game with a scoreboard. A sportsbook grades a bet against a result everyone agrees on. A prediction market has to decide what happened, and the mechanism for deciding is written into each market.
The rules are the market. Every market has a resolution section describing what counts as YES, what counts as NO, which source will be consulted, and the date and time by which the question must be answered. These rules, not the headline question, decide the payout. "Will X be announced by March 31?" may resolve on an official press release, on a specific outlet's reporting, or on a UTC timestamp, and the difference decides borderline cases. Postponed events, disputed results, ambiguous announcements and sources that go silent are all handled by whatever the rules say, or by the resolution process if they say nothing. Read the rules before buying; the price already assumes everyone else has.
Who decides. Polymarket does not resolve markets itself. It uses UMA's optimistic oracle, a system in which anyone can propose an answer by posting a bond, after which there is a challenge window during which anyone else can dispute it by posting a bond of their own. If nobody disputes, the proposed answer stands and the market resolves. If someone does, the question goes to a vote of UMA token holders, who decide the outcome according to the rules; the losing side of the dispute forfeits its bond. This makes resolution a public process rather than a decision by the operator, at the cost of being slower and occasionally contentious.
What happens to your shares. When a market resolves, the winning outcome's shares become redeemable for $1.00 each and the losing outcome's shares for $0. Some markets can resolve to a split, for example 50 / 50 when the rules specify it for a tie or a cancellation, in which case each share pays the corresponding fraction. You redeem winning shares from the market page; until you do, they sit as tokens in your wallet.
Timing. Resolution usually happens within hours of the event, but it can take longer when the outcome is unclear, when the resolution source is slow, or when a dispute is raised. Markets on things like "by end of year" resolve NO at the deadline if the thing has not happened, which means a share you hold can go to zero on a date rather than on an event. Markets can also resolve early when the outcome becomes certain before the deadline.
Disputed resolutions. The system is not perfect. Markets have resolved in ways that surprised traders who read the question but not the rules, and a few high-profile disputes have gone to a vote where the voters' reading of the rules was the deciding factor. The practical lesson is the same one that applies to bookmaker terms: the text governs, and a price that looks like value against a sportsbook line is sometimes just a differently worded question.
Is Polymarket legal?
This is the question new users ask most, and the honest answer is that it depends on where you are and when you ask. Prediction markets sit between financial regulation and gambling law, and both have moved in the last few years. The account below describes the shape of the situation; treat every specific claim as something to check against your own regulator and the platform's own terms, because these facts change.
In the United States. Event contracts fall under the Commodity Futures Trading Commission (CFTC). In January 2022 Polymarket settled with the CFTC, paid a civil penalty and agreed to stop offering markets to US users, which is why the main site blocked US residents for the following years. Regulated alternatives such as Kalshi operate under CFTC oversight and have been the legal route for US residents. In 2025 Polymarket moved back toward the US market by acquiring a CFTC-licensed exchange and clearinghouse and launching a separate, regulated US product; check which product and which markets are available to you and under what terms, because the international site and the US entity are not the same thing.
Elsewhere. Outside the US there is no single answer. Some countries treat prediction markets as gambling and require a local licence that Polymarket does not hold; regulators in several jurisdictions have blocked the site or warned residents against it. Others have no specific rule, and the platform is accessible but the user is on their own. Polymarket's terms of service list the jurisdictions it excludes, and access is geo-restricted accordingly.
What this means for you. Three practical points follow. First, using a VPN to trade from a restricted jurisdiction breaches the terms of service, and an account found to be doing so can be frozen with funds in it. Second, winnings may be taxable where you live regardless of whether the platform is licensed there; the tax treatment of prediction market gains is unsettled in many countries and worth asking a professional about. Third, whether Polymarket is "gambling" is a legal question, not a mathematical one: the mechanics are those of a financial exchange, but the activity of buying a share in an election outcome is, for most people, a bet, and the responsible-play advice on this site applies.
Risks that are not in the price
- Resolution risk. The market resolves by its rules, not by your reading of the question. This is the risk that most often surprises bettors coming from sportsbooks.
- Liquidity risk. A position you can buy is not always a position you can sell at a fair price. Thin markets can move against you with no way out except holding to resolution.
- Counterparty and platform risk. Your funds are on a blockchain, held in a smart contract and a wallet. Smart contracts can have bugs, wallets can be compromised, and the platform can restrict your account. None of this is priced into a 64¢ share.
- Capital lockup. A long-dated share ties up money until resolution or until you find a buyer. Compare the return with the time before deciding a high price is "free money".
- Regulatory risk. Access can be withdrawn by a regulator or by the platform. Keep only what you are prepared to have locked for a while.
Frequently asked questions
Is Polymarket gambling or investing?
Mechanically it is an exchange: prices are set by the order book, positions are assets you hold and can sell, and the venue takes no side. Functionally, buying a share in an outcome is a bet with a defined payout. Regulators disagree on which label applies, and the disagreement is what the legal section above is about. For managing your money, treat it as betting: size positions as a bettor would, using the EV and Kelly calculator with your own probability rather than the market's.
Can I lose more than I put in?
No. The most a share can be worth is $1.00 and the least is $0. A YES share bought at 64¢ can lose at most 64¢. There is no leverage, no margin call and no liability beyond the price paid.
Can I cash out before the event happens?
Yes, by selling your shares. The price you get is the current bid, which may be higher or lower than what you paid. On a thin market the bid may be far below the displayed midpoint; check the order book before assuming a paper profit is real.
Why don't YES and NO add up to exactly 100¢?
Because you are looking at two asks (or two bids), and the gap between bid and ask on each side is the spread. On a liquid market the sum is 100¢ plus a cent or two. If it is ever below 100¢, buying both sides is a guaranteed profit and someone will take it within seconds.
How is a Polymarket price different from sportsbook odds?
They are the same information in different units. A 64¢ share is decimal odds of 1.5625 or about -178 in American odds; the odds converter switches between them. The differences are in the margin (a spread instead of vig), the ability to sell early, and the way the result is decided. The Polymarket prices as odds guide covers the conversion in detail.
What happens if a market is cancelled or the event does not occur?
Whatever the rules say. Many markets specify that NO wins if the event has not happened by the deadline; some specify a 50 / 50 resolution for a cancelled event; a few are worded so that a cancellation is itself the YES condition. The rules are on every market page, and they are the only reliable answer.
Who is on the other side of my trade?
Someone from the list above: a market maker earning the spread, an opinion trader who disagrees with you, a forecaster with a model, an arbitrageur keeping prices consistent, or occasionally someone who knows more than you do. On a liquid market about a major event, assume it is the forecaster or the market maker, and ask what you know that they do not.