Markets vs. Sportsbooks
Sportsbook odds and prediction-market prices become comparable when expressed as the cost of the same return, including exchange fees.
The same basketball game can appear as Boston at -150 on a sportsbook and as Yes on Boston at 60¢ on a prediction market. In this hypothetical example, $60 committed returns $100 on a Boston win before fees, provided both positions use the same settlement terms. Converting the two quotes into dollars makes the prices directly comparable.
The choice also affects what happens before the game ends. If Boston takes the lead and you want to close the position, a sportsbook may offer cash out at a price it sets. Selling an exchange contract requires another participant willing to buy it. Those different arrangements influence the price available and how much of the position you can close, even when the final result would produce the same return.
What is the difference between prediction markets and sports betting?
Prediction markets and sports betting can give you the same gain or loss on a sports result. On Kalshi and Polymarket, you buy contracts you can offer for resale. With a sportsbook, you accept fixed odds for a position, and any cash out comes from the operator.
| What you are comparing | Kalshi or Polymarket | Sportsbook |
|---|---|---|
| Prices and the other side of a trade | Participants submit prices; another member takes the other side | The operator quotes odds and accepts the opposing position |
| Price format | A contract price in cents, such as 60¢ | Odds, such as the American line -150 |
| Trading costs | Fees and the gap between buying and selling quotes | A margin built into the odds |
| Early exit | A sale needs a willing buyer at an acceptable price | The operator may offer cash out at a price it sets |
| Multiple outcomes in one position | A combo has its own quote and settlement terms | A parlay has combined odds and rules for its component outcomes |
| Settlement | Contracts usually pay $1 or $0, with other values under some rules | Wins pay at the accepted odds; a push or void can return the original amount |
| Limits | Position rules and available counterparties constrain size | Operators can restrict the amounts an account may place |
| Reported activity | Volume can count contracts at face value every time they trade | Handle counts money committed to sportsbook positions |
A prediction-market contract and a sportsbook position can both depend on the result of the same basketball game.
How do sports prediction markets work?
You buy a Yes or No contract on a sports result at a price in cents. In the standard $1-or-$0 structure, a Yes contract pays $1 if its condition happens and $0 if it does not. During the game, changing scores and new information can move the price at which you can sell.
Suppose you buy Yes on Boston winning tonight. If Boston takes a large lead, buyers may offer more for the contract than you paid. You can sell to one of them while trading is open, or keep the position until it settles on the game's result. A late comeback by the opposing team can reduce the sale price or leave the contract worth $0 at settlement.
In this hypothetical game, a Yes contract on Boston bought for 60¢ can be sold for 72¢ at halftime, before the price falls to 38¢. Holding it until settlement returns $1 if Boston wins or $0 if Boston loses.
Sports event contracts can also concern a player's scoring threshold or a championship. A contract on Boston winning the championship can keep trading for weeks after tonight's game settles. An injury report can affect both contracts because it changes Boston's chances in both contests.
Who sets prices on prediction markets and sportsbooks?
A sportsbook such as DraftKings sets its odds with an intended margin. On Kalshi, participants submit buy and sell offers that the exchange matches. The list of those offers is an order book.
Participants include traders with a view on the result and market makers that regularly offer to buy and sell.
Suppose Boston is playing New York in a hypothetical basketball game. The sportsbook offers Boston at -150. A Yes contract on Boston winning costs 60¢ and pays $1 if Boston wins. Assume both use the same game result and ignore fees for the moment.
| Position on Boston winning | Cost | Total returned if Boston wins | Profit if Boston wins | Loss if Boston loses |
|---|---|---|---|---|
| Sportsbook position at -150 | $60 | $100 | $40 | $60 |
| 100 Yes contracts at 60¢ | $60 | $100 | $40 | $60 |
A sportsbook takes the opposing position when you accept its -150 odds on Boston. An exchange matches your 60¢ bid for Yes on Boston with another member's 60¢ offer, leaving that member on the other side.
On a betting exchange such as Betfair, customers trade at quoted odds. Taking a position for an outcome is called backing, and taking the opposing position is called laying.
Is Kalshi a sportsbook?
No. Kalshi lets you take positions on sports results through an exchange where members trade event contracts with one another. You can profit from the final result or from selling a contract after its price rises.
Kalshi's rules permit an affiliated firm to submit orders, with operational separation and restrictions on access to nonpublic information. That trading affiliate can be the member on the other side of your order.
Kalshi vs DraftKings
DraftKings Sportsbook sets American odds such as -150, while Kalshi participants quote contract prices such as 60¢. Those quotes offer the same return before fees. Kalshi's fees and available quantities affect the cost of the full order. To leave early, a Kalshi holder needs a buyer, while DraftKings sets its cash-out amount and decides when to offer it.
Kalshi says it does not limit customers for winning. DraftKings' Massachusetts rules allow it to refuse all or part of a proposed amount. Kalshi still has contract limits and constraints on available trading quantities.
For multiple outcomes, DraftKings sets a parlay's combined odds. A Kalshi combo receives its own quote from other participants, and the two products can return different amounts when a player does not participate.
DraftKings launched Predictions in 2025 as a separate event-contract service.
What contract price matches a sportsbook line?
Divide the amount paid at the sportsbook by the total returned if the position wins. The result is the cost per dollar returned, which you can compare with a $1-or-$0 contract's price before fees. At -150, $150 paid returns $250 on a win, so the equivalent contract price is 60¢.
With American odds, a negative line states the amount paid to earn $100 in profit, while a positive line states the profit on $100 paid. At +150, $100 paid returns $250 on a win, equivalent to a 40¢ contract.
Each bar divides a winning contract's $1 return into the purchase cost and the profit before fees. Paying 60¢ to earn 40¢ in profit corresponds to American odds of -150, while paying 40¢ to earn 60¢ corresponds to +150.
At -110, the conversion to a contract price is $110 divided by the $210 total return, or about 52.38¢. Compare that amount with the contract's purchase price plus fees, using positions that settle on the same result.
Does Kalshi have better odds than sportsbooks?
Kalshi can be cheaper on one outcome and more expensive on another. To compare sportsbook odds with contract prices, use the cost of obtaining the same potential return, including exchange fees. For a large order, calculate the cost of each portion available at a different price.
Suppose Boston is offered at -150, which costs $60 for a possible $100 return. Each exchange row below is a separate hypothetical offer for the same game result and the same $100 return if Boston wins.
| Route to a $100 return if Boston wins | Purchase cost | Additional fee in this example | Total cost |
|---|---|---|---|
| Sportsbook at -150 | $60 | None | $60 |
| 100 contracts at 58¢ | $58 | $1 | $59 |
| 100 contracts at 59¢ with the lower fee | $59 | $1 | $60 |
| 100 contracts at 59¢ with the higher fee | $59 | $2 | $61 |
| 20 contracts at 59¢ and 80 at 61¢ | $60.60 | $1 | $61.60 |
The 58¢ offer remains $1 cheaper after its fee. At 59¢, a $1 fee ties the sportsbook and a $2 fee makes the exchange $1 more expensive. In the final offer, 59¢ buys only 20 contracts, so the remaining 80 cost 61¢ each. Pricing all 100 contracts at 59¢ would understate that purchase cost by $1.60.
Only 20 Yes contracts on Boston are offered at 59¢ in this example, so a purchase of 100 also takes 80 at 61¢. The total is $60.60 before fees, which is $1.60 more than pricing the whole order at 59¢.
The ask is a seller's offer, and an immediate purchase takes the lowest available ask first. To sell immediately, you take a buyer's bid, starting with the highest. Each price is available only for the quantity quoted. Polymarket's displayed price can be a midpoint or a previous trade. A chart showing 58% might therefore accompany a 60¢ ask.
The positions need matching settlement terms to promise the same return. A lower price for a regulation-time win cannot be compared directly with a price that includes overtime.
Do prediction markets charge vig?
Vig, short for vigorish, and juice name the margin built into sportsbook odds. On an exchange, fees and the gap between buying and selling quotes can produce a similar cost. If sellers ask 60¢ for Yes and 42¢ for No, buying both costs $1.02 for a combined $1 settlement payment, before fees.
A 60¢ Yes contract and a 42¢ No contract cost $1.02 together and return a combined $1 at settlement. Holding the pair to settlement loses 2¢ before fees, whichever side wins.
When you leave an offer for someone else to accept, you are the maker. The trader who accepts it is the taker. Kalshi earns money by charging transaction fees, and Kalshi fees can also apply to makers in some markets when their resting orders trade. Polymarket's taker fees vary by market category and contract price.
Market makers can earn the bid-ask spread, the gap between the highest bid offered by a buyer and the lowest ask offered by a seller. With a 58¢ bid and 60¢ ask, buying and immediately selling at unchanged quotes loses 2¢ per contract before fees.
If you estimate a 64% chance of receiving $1 at settlement and pay a 60¢ ask plus a hypothetical 1¢ fee, the expected profit is 3¢ per contract.
How do overround and sportsbook hold differ?
Overround measures how far the break-even probabilities implied by all possible outcomes' odds add up beyond 100%. Suppose the sportsbook offers Boston at -150 and New York at +130. Their implied probabilities are 60% and about 43.48%, totaling 103.48%. The overround is about 3.48 percentage points.
Overround is calculated from prices. The sportsbook's hold is money retained after paying customers. DraftKings defines hold as settled handle minus payments to customers. Here, handle means the total amount customers committed to those settled positions. Hold percentage divides the retained amount by that total.
At the example's odds, $60 on Boston and about $43.48 on New York each produce a $100 return if successful. The sportsbook collects about $103.48 and retains $3.48 whichever team wins. Dividing the retained amount by the money collected gives a theoretical margin of about 3.36% for those balanced positions.
Actual amounts need not balance that way. With $60 on each team, the sportsbook collects $120 and owes $138 if New York wins, losing $18. A margin in the quoted odds therefore does not guarantee a profit on a particular game's accepted positions.
With Boston priced at -150 and New York at +130, collecting $60 on Boston and $43.48 on New York leaves the sportsbook about $3.48 whichever team wins. Collecting $60 on each team at the same odds gives it a $20 gain if Boston wins or an $18 loss if New York wins.
How do you calculate no-vig odds?
Convert every outcome's sportsbook odds to the probability needed to break even, then divide each probability by their total. This removes the overround proportionally. To express the result as decimal odds, divide 1 by each adjusted probability written as a decimal.
Suppose Boston is -150 and New York is +130. The -150 line implies 150 divided by 250, or 60%, for Boston. New York's +130 implies 100 divided by 230, or about 43.48%. Together these probabilities total about 103.48%.
Dividing each probability by that total gives about 57.98% for Boston and 42.02% for New York. Dividing 1 by those adjusted probabilities gives decimal odds of about 1.725 and 2.38, equivalent to American odds of -138 and +138. The adjusted probabilities now add to 100%.
The odds of -150 for Boston and +130 for New York imply 60% and about 43.48%, so their total extends past the 100% line. Dividing each probability by the 103.48% total removes the margin proportionally, leaving about 57.98% for Boston and 42.02% for New York.
For Boston, an exchange ask of 56¢ plus a hypothetical 1¢ fee costs 57¢ in total. Using 57.98% as the estimated chance of success gives about 0.98¢ of expected profit per $1 contract at that price. The sportsbook's unchanged -150 offer still needs 60% to break even.
When you devig this way, you assume the margin was distributed proportionally across outcomes. Clarke, Kovalchik, and Ingram's comparison found that this adjustment misses favorite-longshot bias, where low-probability outcomes are priced too expensively relative to favorites. An estimate derived from those prices can retain that bias even after its probabilities add to 100%.
How does Kalshi cash out compare with sportsbook cash out?
A Kalshi cash out is a sale to another participant, while a sportsbook cash out is an offer from the operator. The exchange sale can fill fully, partly, or not at all. Polymarket also supports early sales, with a buyer needed at an acceptable price.
Under DraftKings' Massachusetts house rules, DraftKings specifies the cash-out amount, can withdraw the offer, and can reject a request if the offer changes during a delay.
Suppose you bought 100 Boston contracts for $60 before a hypothetical basketball game. Boston takes the lead, and buyers now offer 72¢ for 40 contracts and 70¢ for another 60. Selling all 100 brings in $70.80, a gain of $10.80 before fees. The best bid, meaning the highest price a buyer offers, was available for only part of the position.
A limit sell order can specify 72¢ as the minimum acceptable price. In this example, that could sell 40 contracts and leave 60 exposed to the game's outcome. Canceling the unfilled order leaves those remaining contracts in the account. A sale also requires the market to remain open for trading, which can end before settlement.
From a position of 100 Yes contracts on Boston bought for $60, selling 40 at 72¢ and 60 at 70¢ brings in $70.80, a $10.80 gain before fees. Setting a 72¢ minimum sells only 40 for $28.80 at these bids, leaving the other 60 contracts in your account.
The choice to sell an event contract or hold it to settlement depends on the price available now and your estimate of the remaining chance of success. Selling a position and withdrawing money from an account are separate actions.
Can you parlay on Kalshi or Polymarket?
Yes. Kalshi Combos and Polymarket Combos combine selected outcomes into one position. Each outcome is called a leg.
How do Kalshi combos differ from sportsbook parlays?
Both need all legs to succeed for the full return. A sportsbook sets the combined odds, while other participants quote an exchange combo. The treatment of a leg that is canceled or unresolved can also change the amount returned.
- Quoting the combination. Kalshi requests prices from other participants through a request for quote, or RFQ, and each combo has its own order book. Polymarket also requests combo quotes from market makers.
- Settling the legs. DraftKings removes a voided or pushed leg from a standard parlay and recalculates the odds. A push is a result that returns the original amount on a standalone position. Kalshi multiplies its component settlement values. Under the rules for player nonparticipation in a combo, the affected leg can settle at its last fair price. A 70¢ leg combined with two $1 legs returns 70¢ per combo contract.
A sportsbook parlay can compound the margin in each leg's odds. For a hypothetical parlay priced by multiplying the decimal odds of two -110 outcomes, the total-return multiplier is about 3.645. If each outcome truly has a 50% chance and they are independent, both occur 25% of the time. The expected return is about 91.1¢ per dollar paid, an expected loss of 8.9¢. Taking just one of those -110 offers would have an expected loss of about 4.5¢ per dollar.
A single position at -110 with a true 50% chance of success loses about 4.5¢ per dollar on average. A parlay combining two independent outcomes on those terms returns about 3.645 times the amount paid when both succeed, which happens 25% of the time and leaves an expected loss of about 8.9¢ per dollar.
An exchange combo receives a quote for the whole position. That quote can account for connections between the outcomes and compensation for the participant taking the other side. For a $1-or-$0 combo, expected profit is the chance of all legs succeeding multiplied by $1, minus the quoted price and fees.
Why a combo differs from separate contracts
Buying one contract on Boston winning and another on a player reaching a scoring threshold lets either contract pay even if the other fails. The table compares this with one contract requiring both outcomes, showing settlement payments before subtracting the purchase costs.
| What happens | Two separate $1-or-$0 contracts return | One $1-or-$0 contract requiring both returns |
|---|---|---|
| Both outcomes happen | $2 | $1 |
| Only Boston wins | $1 | $0 |
| Only the player reaches the threshold | $1 | $0 |
| Neither happens | $0 | $0 |
If two outcomes are independent, their joint probability is the product of their individual probabilities. Chances of 60% and 50% give a 30% chance of both happening. Same-game outcomes can be connected. If the player's chance of reaching the threshold is 70% specifically when Boston wins, the joint chance is 60% multiplied by 70%, or 42%. Multiplying the two standalone probabilities would miss that relationship.
With a 60% chance of Boston winning and an independent 50% chance of the player reaching a scoring threshold, both happen 30% of the time. If the player instead reaches the threshold in 70% of Boston's wins, the chance of both rises to 42%, even though Boston's chance stays at 60%.
Can you sell a combo before the games end?
Polymarket allows combo sales before resolution, provided a buyer will take the position at the offered price and quantity.
A Kalshi combo has its own order book, so an early exit would need a buyer for that exact combination. Kalshi prohibits canceling or reversing a filled purchase.
Do prediction markets settle the same way as sportsbooks?
For an ordinary win or loss, a prediction-market contract settles at $1 or $0, while a sportsbook position pays at its accepted odds or loses the amount committed. Their written rules decide which result counts. Two positions on the same team or player can settle differently because their rules define a win differently:
- The period of play. Two-way moneylines let you choose either team to win, while three-way moneylines add a draw as a third outcome. DraftKings' general settlement rules include overtime and shootouts for two-way moneylines and exclude them for three-way moneylines, subject to sport-specific rules. A team can draw in regulation and win after extra time, producing different results for those two positions.
- The exact threshold. A contract on a player scoring at least 20 points succeeds at exactly 20. A contract on more than 20 does not. A sportsbook total of 20 may instead return the original amount when the player finishes on the line, under its push rules.
When a player scores exactly 20 points, Yes on at least 20 pays $1, Yes on more than 20 pays $0, and a sportsbook position on over 20 returns the original amount as a push. All three positions lose at 19 points and succeed at 21.
A disrupted game can also produce different returns. Under DraftKings' rules, a void returns the original amount. Polymarket's documented Unknown/50-50 resolution pays 50¢ per token, whatever its purchase price. A holder who paid 70¢ would lose 20¢ under that treatment, while someone who paid 20¢ would gain 30¢, before fees.
A sportsbook void returns the full 70¢ placed in this example, leaving no gain or loss. Under Polymarket's Unknown/50-50 resolution, each token settles at 50¢, so buying at 70¢ produces a 20¢ loss and buying at 20¢ produces a 30¢ gain before fees.
The rules for sports contracts on Polymarket and Kalshi also specify whose result counts, such as an official league record, and how postponements or player nonparticipation are handled.
Closing, resolution, and settlement can happen at different times. A finished game can leave a contract awaiting confirmation of the official result, with funds still tied up in the position after trading has ended.
Does Kalshi limit winners?
Kalshi says it does not limit customers for winning. Its exchange earns transaction fees while participants take opposing positions, so a successful trader's gain does not create the same direct loss for the exchange that it can for a sportsbook. Kalshi's position rules and the amount other participants are willing to trade still constrain position size.
At the Massachusetts Gaming Commission's September 2024 roundtable, BetMGM said restrictions reflected how customers placed their positions rather than results alone. DraftKings said its limits were tailored to customer behavior, and FanDuel discussed both individual and general limits.
At the commission's May 2024 meeting, Unabated co-founder Jack Andrews said Bally's risk provider identified customers who obtained favorable prices, then reduced the amounts they could place. He cited closing line value, which compares an entry price with a later pre-game price, and expected value, an estimate of a position's average gain or loss. Limits based on the prices a customer takes can affect an account before its long-run profit becomes clear.
Sportsbooks can review the prices a customer has taken and use that history to limit the amount the customer may place.
How much can you trade on Kalshi and Polymarket?
Kalshi's rules on position accountability and limits set limits for contracts and allow instructions to reduce large positions. Market-maker programs can have different treatment.
Polymarket says its order book has no trade-size limits by design. An order can still move the price or fail to find enough willing participants.
Polymarket's Last Look process lets eligible combo quote providers confirm or decline a selected quote after a final risk check. A request can go unfilled even without an exchange-imposed cap on its size.
Do sports betting strategies work on Kalshi and Polymarket?
You can use no-vig sportsbook odds as a probability estimate and measure whether a contract purchase beats the price just before the game. On Kalshi and Polymarket, those comparisons use cents per contract and must account for fees and the quantity available at each price.
How does closing line value work on prediction markets?
Closing line value, or CLV, compares your entry price with the price available just before an event starts. Buying a contract for less than its later pre-game price is the exchange equivalent of obtaining sportsbook odds that offer a larger return than the closing odds.
Suppose you buy Boston at 60¢ and the best available purchase price rises to 64¢ just before tipoff. The same contract then costs 4¢ more, which is favorable CLV before fees. A 99¢ quote after Boston has won includes the known game result, so it answers a different question from whether you obtained a favorable pre-game price.
Buying Yes on Boston for 60¢ beats the 64¢ purchase price just before tipoff by 4¢, which is favorable closing line value before fees. The later 99¢ quote includes Boston's known win and falls outside that pre-game comparison.
A consistent comparison uses the same point in the game's schedule and the same type of quote for each game. Your actual fill price and the venue's closing quote for a comparable quantity provide those inputs.
In Joseph Buchdahl's analysis of Pinnacle football odds, earlier prices that beat the closing odds were associated with higher subsequent returns across four seasons beginning in 2012/13.
Does sportsbook longshot bias also appear on Kalshi?
Low-priced contracts can be expensive relative to how often they succeed. In a working paper using Kalshi data from 2021 through April 2025, Bürgi, Deng, and Whelan found a favorite-longshot bias in which low-priced contracts succeeded less often than needed to break even after fees. Makers, who posted offers, earned more on average than takers, who accepted existing offers.
For a hypothetical 5¢ contract, a 4% forecast implies an expected loss before fees, while an 8% forecast implies an expected gain. Recording that probability before the result lets you assess forecast accuracy separately from the trade's gain or loss.
Can you arbitrage prediction markets and sportsbooks?
Yes, when opposing positions cover every outcome under their settlement rules and their combined cost is less than the amount returned in each case.
In a hypothetical Boston versus New York basketball game, $60 on Boston at -150 returns $100 if Boston wins. Buying 100 contracts on New York at 38¢ costs $38 and returns $100 if New York wins. Both positions include overtime, so a completed game has one winning side. If the exchange fee is $1, the pair costs $99 for a $100 return either way.
Changing the sportsbook side to a three-way moneyline introduces a third outcome, a tie at the end of regulation. A position on Boston then needs a regulation-time win. If New York's contract still includes overtime, a regulation tie followed by Boston winning in overtime leaves both positions losing.
A two-way sportsbook position on Boston and exchange contracts on New York each include overtime, so one pays whichever team wins. If the Boston position uses a three-way line that counts only regulation, a tie followed by Boston winning in overtime leaves both positions paying nothing.
Both sides also need to execute. DraftKings reserves the right to refuse all or part of the amount offered, and an exchange price can move before the second position fills. Someone holding only the first side remains exposed to the result. The same is true if one side is later voided under rules the other side does not share.
Why is prediction market volume different from sportsbook handle?
Sportsbook handle measures money committed to positions. Prediction-market trading volume can count the full face value of contracts each time they trade, including resales. A buyer's purchase cost can be much smaller than the volume attributed to that trade.
DraftKings counts money committed to settled sportsbook positions as handle. In its prediction-market analysis, Pew Research Center uses notional taker volume, counting each matched contract once at its $1 face value regardless of the purchase price.
Under that volume measure, 100 contracts bought at 20¢ cost the buyer $20 and contribute $100 of volume. Reselling those 100 contracts adds another $100 of volume. The first holder has exited, while ownership has passed to someone else.
The first buyer pays $20 for 100 contracts at 20¢, while volume records $100 because each contract is counted once at its $1 face value. Reselling those contracts adds another $100 of volume, taking the total to $200, ten times what the first buyer paid.
David Huffman, chief operating officer of the sports exchange Sporttrade, distinguishes purchases that add risk to a customer's account from sales that remove it.
A large volume number can reflect frequent trading, low-priced contracts, or both. It cannot be read directly as the amount customers deposited, the money they lost, or the number of people participating.
Prices and Probabilities
How contract prices relate to probabilities, expected returns, and trading costs.
Order Books
How bids, asks, and available quantities determine the price of a trade.
Event Contracts
What a contract promises and how buying, selling, and holding change a position.
Market Resolution
How written rules determine settlement when results or circumstances are disputed.
Markets vs. Polls
Election forecasts turn poll leads into winning chances that can be compared with prediction market prices to judge accuracy.
Prices and Probabilities
Polymarket odds and Kalshi percentages connect contract prices to probabilities, helping you compare the cost of a trade with your forecast.