How to tell if a Polymarket price is fair

Usually, yes. In 8 of 11 sports, large Polymarket buyers from April to September 2026 won within a point of the price they paid, so a sports price starts out as a good estimate of the chance. The 3 that missed by more, cricket, MMA and golf, each rest on fewer than 350 markets, where one card or one tournament moves the whole sport.

A price is fair when it matches the chance the outcome happens. You can't read that chance directly, so you judge the price by how it was made: what a buy would cost you, how deep the book is, whether the outcomes add up to 100%, how the price moved, and who holds each side. After the game, closing line value grades the price you got.

The exception is the cheap end: large buys under 10¢ won less often than their price said, covered in the favorite-longshot bias guide. Everywhere else, each check below looks for a reason the price in front of you is the exception.

0xinsider covers Polymarket only. It doesn't compare prices across venues or publish a fair price of its own. It shows both sides of every game with their prices, which side the profitable wallets hold, and the closing line value of every Pick of the Day on a public record.

Add up the underdogs' prices on the 14 NFL games of October 11 and 12, 2026, and you get 5.05: the market expected about 5 upsets. If 5 favorites lose, nothing went wrong.

That is what reading a price as a probability buys you. Every Polymarket share pays $1 if its outcome happens, so a price in cents reads as a chance in percent, the implied probability defined in the glossary. Even the weekend's biggest favorite, the Jaguars at 77.5¢, loses a game like that more than 1 time in 5.

Yes and No on the same market are two sides of one dollar. Polymarket's documentation describes a first trade as a buy of Yes at 60¢ matched with a buy of No at 40¢: together they make $1, which becomes one Yes share and one No share. So a fair price on one side fixes the other. The Eagles' 22.5¢ and the Jaguars' 77.5¢ came from separate books and added to exactly 100.

On October 11, 2026, 30% of Polymarket's sports moneyline books were more than 10¢ wide, and on those Polymarket doesn't show you the midpoint. It shows the last traded price, which can sit a long way from what a buy costs now.

Neither number is what you pay. A buy fills at the ask, a sale fills at the bid, and a taker pays a fee on top. A 60¢ midpoint with a 58¢ bid and a 62¢ ask costs 62¢ to buy, and the outcome has to be better than 62% for the buy to be fair to you.

Most of the wide books were tennis and table tennis, 339 of the 408; in soccer and the NFL the median book was 1¢ wide. The order book guide reads bids, asks and the spread, and the fees guide has the fee formula. Whatever the sport, judge the ask, not the headline.

A $10,000 buy couldn't be filled at all on 729 of the 1,348 sports moneyline books 0xinsider read on October 11, 2026. Where it could, the median fill cost 4.3¢ more than the quoted ask, and in hockey 13.6¢. A $100 buy barely noticed: 0.5¢.

The best ask holds a set number of shares. An order larger than that fills at the next price up, and the next, so your average price is worse than the quote. That gap is slippage, and the glossary entry on market depth explains how to read it before you size an order.

Depth also tells you whose price it is. A price on a book with resting orders at every level is set by many traders. A price on a thin book can be one participant's quote, and one order can move it. Trust a thin price less, whichever way it points.

Buying both sides of an NFL game on October 11, 2026 cost 1¢ more than the $1 it pays. A tennis match cost 13¢ more, and the table tennis and pickleball books were close to empty.

Each outcome has its own order book, so the prices to buy every outcome can add up to more than 100%. The excess is the overround: what it costs to take every side at once. In a two-way market it's the spread, because buying No at its ask is the same trade as selling Yes at its bid. So the check matters most on games with three or more outcomes, like soccer's home, draw and away, where the median game added up to 102¢.

To read each price with the margin taken out, divide it by the total. The odds calculator does that for 2 to 64 outcomes and states the no-vig fair price in cents. It converts a line from another venue the same way, so the comparison is yours to make, spread and fee included. The no-vig number removes the margin in proportion. It adds no information about who wins.

Large buys placed 1 to 7 days before kickoff lost 2.19 points to the price, on an interval from minus 4.01 to minus 0.43 that leaves out zero. From a day out to the final whistle, no window missed by more than a point. An early price is the one to doubt, and the line movement guide has the detail.

Then ask what moved it. A price that moved because a lineup or an injury became known reflects the news. A price that moved because one large order walked through a thin book can come back once the book refills. Before you trust a move, look at what traded behind it: the large-trade feed shows each large fill with the buyer's grade beside it. A move with one fill behind it is one trader's view.

Before kickoff, the wallet grades told you nothing about a sports price: S, A and B wallets came in 0.85 points behind D and F wallets, on an interval from minus 3.11 to plus 1.41 that includes zero. In play they were 2.45 points ahead, on an interval from plus 0.50 to plus 4.40 that doesn't.

Every Polymarket trade has a public wallet behind it, and 0xinsider grades wallets on settled P&L. Each game on a sports board, from the sports hub, shows both sides with their prices and which side the profitable wallets hold. With Pro the split is live. Without Pro it shows on a delay, and a game with no reading that old shows the money total and wallet count with the split locked.

With Pro, a game where the profitable wallets lean against the price also shows the gap in points between their share of the money and the probability the price implies. It measures positioning. It is not a win probability or an expected return, and the sharp money guide shows how to read the bars against the price.

So weigh the split by the clock. During a game, the graded side is worth a second look. Before it, the data gives their side no weight over the price.

0xinsider's own Pick of the Day beat the closing price on 115 of 436 measured picks, and on average paid 0.3 points more than the close. By the measure this section recommends, the picks have not beaten the market's last price.

Closing line value compares your entry with the last price before kickoff. If the market moved toward your side after you bought, you got the better number, whatever the result. One game says little about a price; a few hundred closing prices say much more. The closing line value guide has the formula and the fee adjustment.

The record is checkable. Each pick is released before its kickoff at a frozen price that is never rewritten, and where the capture works the record stores the last valid price before kickoff beside it. The figures sit in the Approximate market price movement panel on the Pick of the Day stats page, which calls them a reference, not an executed fill.

Run the same test on yourself. Record your entry and the close on every bet. If your average entry sits above the close, as these picks' did, the price was never your edge.