Over/under betting explained: how totals work and what Polymarket's data says about the Under

Most Polymarket totals are soccer goals and esports games, so that is where this guide's numbers come from. A total is a bet on the combined score of a game, with no view on who wins. The market sets a number, 2.5 goals for a soccer match, 2.5 games for a best-of-three esports series, 8.5 runs for a baseball game, 47.5 points for an NFL game, and you take the Over if you expect more and the Under if you expect less.

A 3-0 soccer match goes Over 2.5 and Under 3.5. A 1-1 draw goes Under 2.5. The number is set so the two sides sit near a coin flip, which makes a total the bet where your view of the game matters more than your view of the teams: two strong defenses can produce an Under between good teams, and two weak goalkeepers an Over between poor ones.

At a sportsbook the total moves and the price on each side stays near -110; the line movement guide covers how a moving line works. On Polymarket it works the other way round.

A Polymarket game total is a market with two outcomes, Over and Under, and a line fixed in the market itself. Over 2.5 at 56¢ means the market gives the game a 56% chance of 3 or more goals, and buying it pays $1 a share if it does. The two prices sum to about $1, and the price is the probability, so there is nothing to convert.

A game rarely has just one line. Most list several total markets at different numbers, each with its own price, and a college football game listed the most in the chart above. Every line is a half point, so a total cannot push: of 82,182 settled total lines with a recorded line since April 2, 2026, every one carried a .5.

The cost of the bet is the spread between the bid and the ask, 1¢ at the median on Polymarket sports markets, plus the sports taker fee, which at 50¢ is 2.5% of the stake and falls as the price rises, on most sports; NFL and college football games carry 0.03 and NBA games no fee, as of October 21, 2026. A resting limit order pays no fee; the fees guide has the schedule.

Because the line does not move, the market's changing view of a game shows up in the price. Over 2.5 opening at 52¢ and trading at 58¢ by kickoff is the same event as a sportsbook moving its total from 2.5 to 3, expressed in probability instead of in goals, and every trade behind it is on a public ledger with a wallet attached.

Before kickoff, the large money on Polymarket totals leans Over, as the folklore says, but only just. Once the game starts it swings to the Under.

The folklore every sportsbook bettor learns has two halves. The public bets the Over, because rooting for points is more fun than rooting for a scoreless game. And the book knows it, so it shades the total up, which makes the Under the disciplined side. Both halves are claims about a bookmaker's customers, read from behind the counter, and they survive because they are plausible, not because anyone outside the book has measured them.

Polymarket has no book and every buy is public, so the halves can be measured separately. The over/under study measures both: which side the money takes, in the chart above, and whether the price already reflects it, in the sections below.

In soccer the Under hit 24.2% of matches at 1.5 goals, 47.2% at 2.5, 68.7% at 3.5 and 83.8% at 4.5. The chart above shows every line.

Soccer shows it most clearly, because its lines are the same numbers from match to match. Of 9,731 Polymarket matches with a settled 2.5-goal total and a kickoff between April 2 and September 13, 2026, 52.8% went over and 47.2% under, give or take a point.

Averaged over every line on every game, 52.6% of 83,465 settled lines went Under. That average describes which lines Polymarket lists as much as how games end: a soccer ladder that runs up to 5.5 goals averages to an Under lean because its high lines usually go Under. The lines on one game settle on the same score, so an average like this carries more noise than its count suggests.

Esports series went Under 2.5 games 56.4% of the time, ending in 2 games more often than 3. Counted by game, baseball's lines went Under 49.7% over 2,149 games, a coin flip. College football and the NBA leaned Over on too few games to rule out an even split, and MMA fights went past their round lines more often than not.

A rate like this tells you how often a line lands. Whether anyone is paying too much for it is the next question.

Neither side beat its price. Across 47,121 large buys on those totals, Under buyers paid 58.4¢ on average and won 58.2% of the time, 0.2 points short, and Over buyers paid 56.5¢ and won 55.6%, 0.9 points short. Both intervals run from about minus 4 to plus 3 and neither comes near excluding zero.

The 2¢ gap in what the sides cost comes from the lines buyers chose: an Under on a high line is an expensive bet that usually wins. There is no free Under on Polymarket and there is no Over tax. A trader who takes every Under at 58¢ is paying 58¢ for a 58% chance and losing the spread and the fee.

By sport the answer holds. The soccer Over is the closest thing to the folklore, minus 2.64 points and minus 8.1% on the dollar across 12,336 buys, and its interval runs from minus 8.2 to plus 2.7 because the buys cluster on a few big games. Baseball runs the other way, with Over buyers ahead and Under buyers behind, and the NBA's Over came in above zero. The sign of the Over's edge changed from month to month between April and July.

Over 3.5 is the line where buyers were wrong by a margin. They paid 51.2¢ and won 46.0% of the time, minus 5.1 points and minus 20.8% on the dollar across 2,223 buys, on an interval that runs from minus 16 to plus 5 and cannot tell a mispricing from a low-scoring summer. Under 3.5 buyers, at 66.1¢, won 65.3%.

The other lines leaned less. Over 2.5 buyers paid 54.4¢ and won 51.8%; Under 2.5 buyers paid 54.4¢ and won 56.2%. At 1.5 goals, a line the market lists for games it expects to be tight, Over buyers paid 73.7¢ and won 71.7%.

So in soccer totals, the question is the number rather than the side. The market priced the 2.5 line to within three points either way and the 4.5 and 5.5 Unders to within two. The one line where buyers were wrong by the widest margin was the one that asks for the most goals, and a trader who can say why a game reaches 4 goals has a bet the crowd could not price this summer.

Goals come in pairs, too: when both teams scored, a Polymarket match went over 2.5 goals 77.9% of the time across 6,903 settled matches, which the both teams to score guide works through.

The live prices were fair. In play, both sides finished within a point of their price; before kickoff, both came in up to a point short.

Those live prices sit well above the pre-kickoff 54¢ because by the time most live buys land the game has told the market something. An Under gets more likely every minute nobody scores; an Over gets more likely with every goal. The timing study found the same on every market type: the in-play price was kinder to the buyer than the pre-game one.

That is also why the large money swings to the Under once a game starts, as the folklore section above showed. A live Under is a bet that the game stays as it is, and every scoreless minute makes it more likely.

On game totals, wallets graded S, A or B put 60.0% of their buys on the Under. Wallets graded D or F put 49.8% on it. 0xinsider grades every Polymarket wallet from S to F on its settled record, and from June 2026 each large buy carries the grade its wallet held when it traded.

The graded cohort's Unders came in at +1.85 points and +5.31% on the dollar across 4,960 buys, with an interval that includes zero. Their Overs came in at -2.95 points. The poorly graded cohort lost on both sides, -3.43 points on Overs and -2.48 on Unders. These are the over/under study's numbers, split by side of the bet and re-measured on October 7, 2026 with each buy graded at the trade.

The fade study adds the other half: when the D and F wallets' money agreed on one side of a total before kickoff, that side trailed its price by 2.87 points across 351 one-sided markets, inside the margin of error. A crowd on the Over of one game is a reason to check the price.

The graded wallets' lean is a default, not a rule; its interval says so.

Write down the number of goals or points you expect before you read the price. Between 40¢ and 60¢, where most large buys land, both sides came in at their price: the market's number is good, and your edge has to come from a better one. The 60¢ to 80¢ band is the one to check twice: both sides came in two to three and a half points short there, a gap the study gives no interval, so check the price there before you buy.

Each game on a 0xinsider sports board, soccer, MLB, NBA, NFL, esports and the rest, lists its total beside the moneyline and the spread, with the Over and Under prices in cents.

Open the total and its market page lists the wallets holding each side with the grade each has earned. Graded wallets on the Under with the poorly graded wallets on the Over is the study's pattern; both cohorts split is a fair price. The large-trade feed shows each large buy with the buyer's grade, live with Pro and 24 hours later without Pro, and the sharp money guide walks through each page in order.

Then take the side whose price is below your probability, after the spread and the fee, and size it as the Kelly criterion guide describes. The market prices each line well; the edge, where there is one, is in knowing the game.