Polymarket Sports Betting 2026: Costs and Limits
Every guide to Polymarket sports betting says the fees are almost nothing, and not one of them prices a real market. We pulled the order books for the Colts at Texans game on both Polymarket and Kalshi at 21:06 UTC on 19 November 2026, three hours before kickoff: the headline moneyline quoted 15 to 16 cents on both venues, a one cent spread, while the median spread across the same game's other 304 markets was 16 cents and 264 of them had not traded all day. This guide prints the all-in cost of entering at every price band, against the 4.55 percent of stake a minus 110 sportsbook line expects to keep, and what changes when Polymarket's fee and rebate rules reset on 20 November 2026. Every figure comes from the two venues' public order books and Polymarket's own live fee page.

Key takeaways
- Polymarket and Kalshi quoted the Colts moneyline identically at 21:06 UTC on 19 November 2026: 15 cents bid, 16 cents asked. Two independent books, agreeing to the cent.
- A one cent spread is only cheap where the price is high. On a 50 cent market it costs 1.0 percent of your stake; on the 16 cent Colts it costs 3.1 percent, because the same cent is a bigger slice of a smaller number.
- The sports taker fee is shares times 0.05 times price times one minus price. Against what you stake that simplifies to 5 percent times one minus price: 2.5 percent at 50 cents, 4.2 percent at 16 cents, 0.5 percent at 90 cents.
- All in, entering at 50 cents costs about 3.5 percent of stake against the 4.55 percent a minus 110 line takes: cheaper by a quarter, not by the almost nothing the guides claim. At 16 cents it is 7.3 percent.
- Depth is the real limit. The moneyline held 943,630 dollars of liquidity, but 191 of the game's 305 markets held under 10,000 dollars, 264 had no 24 hour volume, and only 16 quoted a one cent spread.
On this page
- How Polymarket sports betting actually works
- What the Colts at Texans game really cost
- The overround: 3.1 percent across 32 teams
- The fee that sits on top of the spread
- All in: what Polymarket sports betting costs
- What changes on 20 November 2026
- 305 markets, 16 of them tradeable
- Where a sportsbook still wins
- How to price any market yourself in five minutes
- Frequently asked questions
- Sources and when we last checked
This is a plain list rather than jump links, because guide pages on this site render section text as paragraphs. If you read two sections, read the one on the fee and the one on depth: between them they are why a dozen markets per game are a bargain and the other 290 are not.
How Polymarket sports betting actually works
- Buy a share at its ask, hold to resolution, collect one dollar if it wins.
- Sell at the bid at any time before resolution to take a profit or cut a loss.
- Price in cents is the implied probability in percent. 16 cents is 16 percent.
- Minimum order on the markets we priced was 5 shares, on a one cent tick.
- Makers, meaning anyone whose resting limit order gets filled, pay no trading fee at all.
Polymarket is an exchange, not a bookmaker. No line is set for you and no house takes the other side. Every contract pays one dollar if its outcome happens and nothing if it does not, and the price is whatever another trader will sell for. A Colts share at 16 cents is a claim on one dollar costing 16 cents, which is the same statement as a 16 percent implied probability.
That changes three things. Your cost is a spread rather than a margin baked into the odds, you can sell before the game ends because somebody else is bidding, and nobody can limit you for winning. What it does not change is that you still pay to get in: the exchange charges a taker fee and the book charges you a spread, and on a thin market the spread is far the larger.
What the Colts at Texans game really cost
- Polymarket, Colts: 15 cents bid, 16 asked. Liquidity 943,630 dollars, 24 hour volume 200,474 dollars.
- Kalshi, Indianapolis wins, ticker KXNFLGAME-26NOV19INDHOU-IND: 15 cents yes bid, 16 yes ask, 84 to 85 on the no side.
- Both venues run a one cent tick, so one cent is the minimum displayable spread.
- Two sided cost on either venue: 101 cents for a one dollar payout, a 1 percent overround.
- Cost of crossing it: 3.1 percent of stake on the 15.5 cent Colts, 0.6 percent on the 84.5 cent Texans.
We priced Indianapolis at Houston, scheduled for 8:15 PM ET on 19 November 2026. Polymarket's Gamma API publishes a best bid and ask per outcome, Kalshi's public markets endpoint a yes bid and ask. Both are prices an order would meet, not the last prints most comparisons quote and nobody can fill.
The two venues were identical. Polymarket showed the Colts at 15 bid and 16 asked, last trade 16. Kalshi's Indianapolis market showed a 15 cent yes bid and a 16 cent yes ask, its no side at 84 to 85. Both run a one cent tick, so one cent is the tightest spread either venue can display. Adding both sides, Colts at the 16 cent ask plus Texans at 85 cents, the mirror of the Colts bid, comes to 101 cents for a certain one dollar payout: a 1 percent overround, and the same on Kalshi.
One cent of spread is not one percent of cost, though, and this is the part every guide skips. You pay the spread on the price you buy, not on the dollar you might win. Half a cent against a 15.5 cent mid is 3.1 percent of your stake; the identical half cent against the Texans at 84.5 cents is 0.6 percent. The favourite side of this market was five times cheaper to enter than the underdog side, on the same screen, at the same moment.
The overround: 3.1 percent across 32 teams
- Pro Football: 2027 Champion, pulled 21:06 UTC on 19 November 2026: 74.4 million dollars volume, 7.33 million dollars liquidity, 32 team outcomes.
- Sum of best asks: 103.10 cents, an overround of 3.10 percent.
- Sum of best bids: 97.60 cents. Sum of midpoints: 100.35 cents.
- Tightest quotes: San Francisco 14.7 to 14.8 cents, Jacksonville 12.4 to 12.5, Buffalo 14 to 15.
- Seven of the 32 teams had no bid on the book, so a position in them cannot be sold before resolution.
A two way market is the easy case. What a long futures market costs is the harder question, and it is where a sportsbook's margin is widest and where nobody publishes a measurement.
So we added up Polymarket's whole Super Bowl book. The 2027 Champion market carried 74.4 million dollars of lifetime volume and 7.33 million dollars of live liquidity across 32 team outcomes. Buying every team at its best ask would cost 103.10 cents for a guaranteed dollar, a 3.10 percent overround across 32 outcomes; selling every team at its best bid would raise 97.60 cents. The midpoints sum to 100.35 cents, about as close to a fair book as a penny tick allows.
Hold that next to a two way minus 110 line at 4.76 percent: the futures market with sixteen times as many outcomes carries the smaller total margin, which is what an exchange looks like when market makers compete with each other rather than quote against you. The margin is spread unevenly, though. The 49ers at 14.7 to 14.8 cents have a tenth of a cent spread, while seven teams sat with no bid and a 0.1 cent ask, a market offering a lottery ticket nobody will buy back.
The fee that sits on top of the spread
- Formula: shares times 0.05 times price times one minus price, for all sports markets.
- As a share of what you stake: 5 percent times one minus price.
- 5 cents: 0.24 cents per share, 4.75 percent of stake.
- 16 cents: 0.67 cents per share, 4.20 percent of stake. On 100 shares, a 16 dollar outlay and a 67 cent fee.
- 50 cents: 1.25 cents per share, 2.50 percent of stake. The worst case in cents, the middle case in percent.
- 85 cents: 0.65 cents per share, 0.78 percent of stake.
- 95 cents: 0.24 cents per share, 0.25 percent of stake.
- Makers pay zero. The fee reaches only the side that takes an existing offer.
Polymarket's fee page gives the formula as shares times feeRate times price times one minus price, with a Sports rate of 0.05 and no fee for makers. The shape matters more than the constant: running on price times one minus price, the fee per share peaks at 1.25 cents on a 50 cent contract and falls away towards both ends.
In cents that looks friendly to longshots. Against what you actually stake it is the opposite. Divide the fee by the price and it collapses to 5 percent times one minus price, so the fee is 2.5 percent of stake at 50 cents, 4.2 percent at 16 cents and 4.75 percent at 5 cents. The cheaper the contract, the larger the bite.
There is one way around it entirely, and it is the most useful line on this page: rest a limit order instead of taking the offer. Post a bid at 15 cents rather than paying the 16 cent ask and you skip both the spread and the fee, at the cost of not knowing whether you will be filled.
All in: what Polymarket sports betting costs
- Minus 110 both ways: 52.381 percent implied per side, 104.762 percent total, 4.76 percent overround, 4.55 percent of stake expected cost.
- Polymarket at 50 cents: 1.00 percent spread plus 2.50 percent fee, so 3.50 percent of stake.
- At 25 cents: 2.00 plus 3.75, so 5.75 percent.
- At 16 cents: 3.12 plus 4.20, so 7.33 percent.
- At 85 cents: 0.59 plus 0.78, so 1.37 percent.
- At 95 cents: 0.53 plus 0.25, so 0.78 percent.
- All six assume you pay the ask on a one cent spread. Rest a limit order and both costs go to zero.
Now put both costs together against the thing readers are comparing, which is their sportsbook. A minus 110 line on both sides implies 52.381 percent per team, a book summing to 104.762 percent, an overround of 4.76 percent and an expected cost of 4.55 percent of every dollar staked. That is arithmetic, not an estimate: a 110 dollar stake returning 210 dollars on a true coin flip loses 5 dollars on average.
Polymarket's all-in entry cost is the half spread plus the fee, both measured against stake. At 50 cents that is 1.0 plus 2.5, so 3.5 percent. The exchange is cheaper than a minus 110 line near even money by about a quarter of the cost. It is not free, and it is not the order of magnitude the comparison guides imply.
Where the picture inverts is the longshot: 7.3 percent of stake at 16 cents and 14.75 percent at 5. We are not claiming your sportsbook is cheaper there, because books price longshots with margins far wider than a minus 110 line and we did not measure one. The claim is narrower and nobody else makes it: Polymarket's own cost of entry roughly doubles between a coin flip and a 16 cent underdog, and quadruples by 5 cents. Favourites are the bargain, at 1.37 percent on an 85 cent contract.
What changes on 20 November 2026
- Effective Friday 20 November 2026 at 23:59 UTC, on Polymarket DeFi at Polymarket.com.
- Sports taker coefficient: 0.05, all sports including NFL and college football. Unchanged.
- Taker rebates: nothing under 200,000 dollars, Gold 10 percent to 1 million, Platinum 25 percent to 4 million, Diamond 40 percent to 10 million, Obsidian 50 percent to 75 million, Onyx 60 percent above.
- Maker rebates: nothing under 200,000 dollars, then 3, 7, 10 and 15 percent at 1 million, 4 million and 10 million.
- Bronze and Silver taker tiers, category multipliers and level-up bonuses are all removed.
- Weighted volume equals shares times price times one minus price, much smaller than dollar volume.
- Geopolitics markets stay fee free. Combos keep a 0.05 coefficient whatever their legs are, and earn no maker rebate.
Anything written about Polymarket sports fees before this week is about to be out of date, and the timing is tight: the new fee and rebate rules take effect on Friday 20 November 2026 at 23:59 UTC, the day after we priced these markets.
The taker coefficient is not the change. Polymarket's change page restates it as 0.05 for all sports, naming NFL and college football, which is what its current fee page already shows. What reaches ordinary accounts is the rebates: both the taker and the maker rebate become volume tiers with a 200,000 dollar floor, and below that floor the rebate is zero. The Bronze and Silver taker tiers go, along with the category multipliers and the one-time level-up bonuses. For a recreational bettor that is a straightforward rise in the effective cost of taking liquidity, because the partial refund that used to offset part of the fee is gone.
Makers are not charged, so nothing you pay goes up there. What disappears below 200,000 dollars of 30 day maker weighted volume is the 15 percent rebate on taker fees your fills generate. Weighted volume is not dollar volume: it is shares times price times one minus price per fee-paying fill, so that threshold takes far more than 200,000 dollars of trading to reach.
305 markets, 16 of them tradeable
- 305 open markets on one game. Liquidity 6.93 million dollars, 24 hour volume 530,222 dollars.
- Median market liquidity: 5,652 dollars. Markets at or above 100,000 dollars: 12.
- Markets under 10,000 dollars of liquidity: 191 of 305.
- Markets with zero 24 hour volume: 264 of 305. Under 100 dollars: 282.
- Median spread: 16 cents. Markets quoting a one cent spread: 16 of 305.
- Check the bid before you plan an exit. A market with no bid can be bought and not sold.
One NFL game on Polymarket is not one market. The Colts at Texans event carried 305 open markets: the moneyline, forty-odd spread lines, totals at every half point, first half versions of most, and a long tail of props. The headline figures look enormous, at 6.93 million dollars of liquidity and 530,222 dollars of 24 hour volume.
Both numbers are concentrated in a dozen markets. The median market held 5,652 dollars of liquidity, 191 of the 305 held under 10,000 dollars, and only 12 held 100,000 or more. On volume the concentration is starker: 264 of the 305 had traded nothing at all in 24 hours and 282 had traded under 100 dollars.
The spreads say the same thing in the number that costs you money. The median spread across the event was 16 cents, not one; only 16 markets quoted a one cent spread, and the widest sat at a full dollar, a nominal offer with no bid underneath. A 16 cent spread on a 50 cent contract is not a price, it is a placeholder. Read the menu as a dozen real markets and 290 listings, and check the quote on anything outside that dozen.
Where a sportsbook still wins
- Props and in play: a sportsbook quotes them, Polymarket mostly lists them without a quote.
- Promotions and boosts: no exchange equivalent exists.
- Familiar American odds and one balance, against prices in cents and a wallet.
- Against that: no limits for winning, an exit at any time, and 3.5 percent all-in at even money versus 4.55 percent.
- Access depends on your jurisdiction on both sides. Check that before you fund anything.
The cost case for the exchange is strong on the markets that have a book, and it would be dishonest to stop there. A sportsbook prices player props with real two way quotes, in play markets that update by the second, and same game parlays. Polymarket lists many of those and quotes almost none of them, so if your betting is props and live markets the exchange is not yet a substitute.
Promotions are the other gap. A bonus or a boost can be worth more than the entire margin this guide has measured, and an exchange has no equivalent because there is no house to fund it. The third difference is access: which venue you can legally fund depends on where you are, and our legality guide keeps the state by state and country by country position.
How to price any market yourself in five minutes
- Open the market and read its best bid and best ask, not its last traded price.
- Add the two sides: the ask on one outcome plus one minus the bid on the other. Anything over 100 cents is your two way cost.
- Work out the half spread as a percentage of the price, not of a dollar. This is the step that changes the answer on longshots.
- Add 5 percent times one minus price for the sports taker fee, or zero if you intend to rest a limit order.
- Check the bid side has size. An exit you cannot make is a position you must hold to resolution.
- Compare the total against 4.55 percent of stake, which is what a standard minus 110 line expects to keep.
Nothing above needed a key or an account, so you can repeat it on any game. The order of operations matters more than the arithmetic: read the quote first and the headline second, because a market with 940,000 dollars of liquidity and a one cent spread looks identical to the other 300 listings until you check the bid.
Frequently asked questions
Can you bet on sports on Polymarket?
Yes, and sports is now one of the busiest parts of the platform. A single NFL game carried 305 open markets when we checked on 19 November 2026, covering the moneyline, spreads, totals, first half lines and a long tail of props. The qualification that matters is that roughly a dozen per game carry a real two way quote and the rest have little or no liquidity. Which markets you can legally trade depends on your jurisdiction.
Does Polymarket charge fees on sports markets?
Yes. The published formula is shares times feeRate times price times one minus price, with a Sports rate of 0.05, charged only to the side taking an existing offer. In cents it peaks at 1.25 cents per share at 50 cents. Against what you stake it is 5 percent times one minus price, so 2.5 percent at 50 cents and 4.2 percent at 16. Resting limit orders pay nothing.
Is Polymarket cheaper than a sportsbook?
Near even money, yes, by about a quarter. Entering at 50 cents costs roughly 3.5 percent of your stake once you add the half spread to the fee, against the 4.55 percent of stake a minus 110 line expects to keep. The gap narrows and then reverses as the price falls: at 16 cents the all-in cost is 7.3 percent and at 5 cents it is 14.75 percent. It is cheapest of all on favourites, at 1.37 percent on an 85 cent contract.
Can you cash out of a Polymarket sports bet early?
You can sell to another trader at any time before the market resolves, with no haircut applied by a house. Check there is a bid first. Seven of the 32 teams in the Super Bowl market had no bid at all when we looked, and 264 of one game's 305 markets had not traded in 24 hours, so an exit is only as good as the other side of the book.
How much can you actually bet on a Polymarket NFL game?
On the headline market, a lot: the Colts moneyline showed 943,630 dollars of liquidity and had traded 200,474 dollars in 24 hours. On everything else, far less than the event-level number suggests. Total liquidity on the game was 6.93 million dollars, but the median market held 5,652 dollars and 191 of the 305 held under 10,000. Size your order against the individual market's book, never the event total.
Did Polymarket's sports fees change in November 2026?
The rebates did, on 20 November 2026 at 23:59 UTC. The sports taker coefficient stays at 0.05 for all sports including NFL and college football. What changed is that taker and maker rebates became volume tiers paying nothing below 200,000 dollars of 30 day weighted volume, and the Bronze and Silver tiers, the category multipliers and the level-up bonuses went. For a recreational account that is a rise in the effective cost of taking liquidity.
Do Polymarket and Kalshi price NFL games differently?
Not on the game we measured. At 21:06 UTC on 19 November 2026 both quoted the Colts at 15 bid and 16 asked, with Kalshi's no side at 84 to 85, and both run a one cent tick. Identical quotes on two independent books is what a liquid market should produce, and it also means there was no cross-venue gap to trade. Our arbitrage guide measures what happens when the two diverge.
Sources and when we last checked
- Polymarket Gamma API, events and markets endpoints, pulled 21:06 UTC on 19 November 2026: best bid and ask, liquidity, 24 hour volume, the 5 share order minimum, the one cent tick and the Colts at Texans resolution text. No key required.
- Kalshi public markets API, ticker KXNFLGAME-26NOV19INDHOU-IND, pulled the same evening: yes bid, yes ask, no bid, no ask and a linear one cent price structure.
- Polymarket fee documentation at docs.polymarket.com, read 19 November 2026: the shares times feeRate times price times one minus price formula, the 0.05 Sports coefficient, zero maker fees, and the fee and rebate changes effective 20 November 2026 at 23:59 UTC with their full taker and maker tier tables.
- New Jersey Division of Gaming Enforcement press release, as of 19 October 2026: sports wagering gross revenue of 76,983,243 dollars for September 2026, down 14.3 percent year on year, and 748,354,348 dollars year to date, down 6.3 percent.
Two limits on what this measures. It is one timestamp on one Thursday night game plus one futures market, so it describes the shape of the cost rather than a seasonal average. A Sunday slate, or the same game near kickoff, will show tighter quotes on more markets.
The minus 110 comparison is deliberately the standard case: exact arithmetic for a two way line at even money, the benchmark for the markets where the exchange is cheapest. We did not price a sportsbook's longshot or futures margins, so the longshot figures describe Polymarket's own cost curve and are not a claim that your book is better value there.
One source conflict is worth naming rather than hiding. Several prominent guides give a 0.03 taker coefficient for NFL and college football. Polymarket's current fee page shows a single Sports rate of 0.05 with no league carve-out, and its November 2026 change page restates 0.05 for all sports, naming both. We have used 0.05. Price the fee from the live page rather than from any guide, including this one. Nothing here is financial advice. The line to carry away: the spread and the fee are both percentages of the price you pay, so the cheaper the ticket, the dearer the trade.