Prediction Market Arbitrage: How It Works Across Kalshi, Polymarket, And The Sportsbooks

The same real-world question is now priced in two different languages. A sportsbook quotes a World Cup final or an NBA game in American odds, +130 or -150. A prediction market like Kalshi or Polymarket quotes the identical outcome in cents, where a contract at 60 cents means the market is pricing it near 60 percent. Those two prices are set by different crowds, on different platforms, under different rules, and they do not always agree. Prediction market arbitrage is the practice of covering both sides of the same event across two venues when their prices disagree enough that the two legs together cost less than the dollar they pay out.

That gap is the whole story here, so let us be honest about its size up front: it is real, it is narrower than the "all-green calendar" screenshots make it look, and the places you can legally access it are a moving target. What this actually rewards is not a magic scanner. The edge goes to whoever can translate between the two price languages faster and more accurately than everyone else. And translating a sportsbook line into the same unit a prediction market uses turns out to be a single, specific calculation. We will get to that number.

In Summary

• Prediction Market Arbitrage exploits a price disagreement on the same event across two venues: a prediction market (Kalshi, Polymarket) priced in cents, and either a sportsbook priced in American odds or a second prediction market.
• Prediction markets are CFTC-regulated event contracts, not sports betting. That legal distinction matters, and it is covered below.
• The mechanism: if you can cover both outcomes for a combined cost under $1.00 (before fees), then once both legs are placed at those prices the position is structured to return $1.00 whichever side resolves. The gross gap is the difference.
• The gap is fragile. Trading fees, spreads, slippage between your two legs, sportsbook limits, and collateral tied up on both platforms at once all eat into it. None of this is a promise of profit, and this guide never states a result you should expect.
• The hard part is comparison, not execution. To spot the gap, read both venues in the same unit, the cost right now to return $1.00 on each outcome; our Arbitrage tool surfaces those crossed prices across every major book, and no-vig fair-odds pricing then tells you whether the gap is a real edge.
• Availability Is State-Specific And Changing. As of July 2026, both platforms are blocked in several states, and Minnesota makes participation a felony effective August 1, 2026. Check each platform's own eligibility rules before you do anything.

If you are brand new to the concept of locking both sides of a market, start with our sportsbook arbitrage explainer. This guide assumes you already grasp that idea and focuses on what changes when one of the venues is a prediction market.

Two Price Languages For One Event

A prediction market is an exchange. You are not betting against a house that sets a line and bakes in a margin. You are buying and selling contracts against other traders, and each contract pays $1.00 if the event happens and $0.00 if it does not. The price, quoted in cents, is the market's live estimate of the probability. A contract trading at 42 cents is the crowd saying "about 42 percent," and if the event resolves in your favor, that 42-cent contract pays a dollar.

A sportsbook works the other way. It posts a number, it takes both sides, and it builds in a hold, the vig, so the two sides do not add up to a clean 100 percent. That hold is how the book makes money, and it is why the same event can carry a different implied probability at a book than it does on an exchange.

Here is why that difference creates an opening. The sportsbook's price reflects a margin plus wherever the recreational public has pushed the line. The prediction market's price reflects a separate pool of traders, often reacting to news on its own timeline. When those two crowds land in different places, the same outcome ends up cheaper on one venue than the other. Buy the cheap side of the question on one platform, cover the other side on the second, and if your combined cost comes in under a dollar, you have built a position that returns a dollar either way. That is the crossed market the whole strategy hunts for.

The Mechanism: A Worked Example

Let us walk one all the way through. These numbers are illustrative, chosen to show the math cleanly rather than to quote any live market.

Say a two-way matchup, the home team versus the road team, is trading like this:

Venue	Outcome you back	Price	Cost to collect $1.00
Kalshi (Exchange)	"Home team wins" YES contract	53 cents	$0.53
A Major Sportsbook	Road team moneyline	+130	$0.435

Read the most important row first: the sportsbook has the road team at +130, which is a way of saying the book's price implies the road team wins roughly 43 percent of the time. A +130 stake of about 43.5 cents returns a full dollar if the road team comes in. Meanwhile the exchange will sell you "Home team wins" for 53 cents, and that contract pays a dollar if the home team comes in.

Add the two costs: 53 cents plus 43.5 cents is 96.5 cents. Once both legs are placed at these prices, that 96.5 cents is structured to return a full $1.00 whichever team wins. The gross gap is about 3.5 cents on the dollar, close to 3.5 percent. On the surface that looks too good to be true. Often it is.

That surface is misleading, and the next section is the reason.

Why The Gap Is Smaller Than It Looks

That 3.5-cent gross number is the ceiling, not the take-home. Four frictions stand between it and reality, and any one of them can shrink the edge to nothing or push it negative.

The one thing to remember: a gross gap is a starting point, not a payout. Price the fees, the slippage, and the doubled collateral before you decide any prediction market arbitrage is worth the capital.

Trading fees. Kalshi charges a per-trade fee that is largest on near-coin-flip contracts and shrinks toward the extremes, so a leg priced at 53 cents sits close to the peak of that curve. Whatever the exchange keeps comes straight out of your 3.5 cents before you have done anything else.

Slippage between the two legs. You cannot place both sides at the exact same instant. You buy the exchange contract, then turn to the sportsbook, and in those seconds the number can move. If the sportsbook shortens the road team from +130 to +118 before you get down, the leg that was going to cost 43.5 cents now costs more, and your gap narrows or vanishes. Sharp bettors call the price that walks away from you slippage, and on a fast-moving market it is the single biggest killer of a thin arb.

Sportsbook limits. Books do not enjoy paying arbitragers. Consistent two-sided, gap-hunting action is exactly the pattern their risk systems flag, and the operational endpoint is reduced limits or, in the worst case, a closed account. Getting limited is a compliment, as the saying goes, but it also caps how long any sportsbook leg of this stays available to you.

Collateral on both platforms at once. This is the friction specific to cross-venue arbitrage, and it is the one newcomers miss. Your exchange position and your sportsbook position are not netted against each other. There is no prime broker sitting between Kalshi and DraftKings to say "these two offset, post margin once." You fund the full stake on the exchange and the full stake at the book, simultaneously, for the same event. That doubles the capital each opportunity ties up and drags down your real return on the money at work, even when the gap itself is positive.

Stack those together and the illustrative 3.5 cents is often a fraction of a cent, sometimes zero, occasionally underwater. None of that is a reason to dismiss the strategy. It is the reason speed matters more than the size of any single gap, which brings us back to the number we promised.

How To Actually Find It: Speak One Language

Remember the two price languages. You cannot compare 53 cents to +130 by eye, and that is precisely why most people never see these gaps. The fix is the calculation we promised: convert each venue's price into what it costs, right now, to return a dollar on that outcome. A prediction-market contract already reads that way, since a 53-cent contract costs 53 cents to return $1.00. The sportsbook price converts too, because +130 costs about 43.5 cents to return the same dollar. Line the two executable costs up, and a gap that was invisible becomes a subtraction problem. That cost-to-return-a-dollar comparison, not the fair probability, is what tells you an arbitrage exists.

A second number tells you whether that gap is worth trusting. Strip the vig out of the sportsbook line and you get its no-vig fair price, the book's honest estimate of the outcome once its margin is removed. That is the value read: it separates a genuine cross-venue inefficiency from a stale or limit-only quote that will vanish the moment you try to size up. You find the arb with executable prices; you judge it with fair odds.

Doing both by hand across every game, every book, and two exchanges is not realistic, which is where our tools come in. This is our lane, not because we run a prediction market, but because pricing sportsbook lines at scale is exactly what OddsShopper already does:

• The Arbitrage tool scans every major U.S. sportsbook in real time and surfaces crossed prices where the two sides of a market do not agree. This is a live, multi-book scanner, not a two-input calculator you feed by hand.
• Our no-vig, fair-odds pricing shows the true implied probability behind any posted line, so once you spot a book-versus-exchange gap you can tell whether the contract is genuinely mispriced or just noise.
• The +EV screen flags when a sportsbook price is out of line with the fair market, using expected win rate and expected ROI rather than a gut feel.
• The Sharp Action tool reads where exchange and prediction-market sharp money is moving, which is useful context for whether a gap is a real inefficiency or a stale quote about to snap shut.

The workflow is the story of this whole guide in three steps: translate both venues into the cost to return $1.00, find where those costs sum to under a dollar, and decide whether the gap survives the fees, the slippage, and the doubled collateral we just walked through. Free OddsShopper tools cover the core odds comparison; OS Pro unlocks the full real-time arb feed and the deeper de-vig and Sharp Action reads, with a 7-day free trial. New OS Pro members can take 20 percent off with code PMARB20.

Kalshi Against Polymarket: The Same Idea, One Venue Removed

Everything above pairs a prediction market with a sportsbook, but the cleanest version of prediction market arbitrage is exchange against exchange. If Kalshi prices a given outcome's YES contract at 55 cents and Polymarket prices the opposite outcome such that covering it costs less than 45 cents, the two contracts sum to under a dollar, and the same structure exists without a sportsbook leg at all.

Two things make the exchange-versus-exchange version distinct. First, both venues quote in cents already, so you skip the de-vig translation entirely. Second, exchanges quote tighter spreads than sportsbooks post holds, so the raw gaps are usually thinner, and fees and liquidity matter even more. A gap you spot on paper can evaporate if the contract you need is thinly traded and your order walks the book. Liquidity is no footnote on an exchange. It is the difference between a fill and a phantom.

The Legal Line: These Are Contracts, Not Bets

This is the part you do not get to hand-wave, and it is why this article carries a review hold before it publishes.

Prediction markets like Kalshi and Polymarket are CFTC-regulated event contracts, not sports betting. That is not a marketing distinction. It is a legal one, and it cuts in a genuinely counterintuitive direction: prediction markets can be available in states where sportsbooks are not, and blocked in states where sportsbooks are live. Do not assume that because you can bet at a sportsbook where you live, you can trade event contracts there, or the reverse. That said, the framing is actively contested: several state regulators are challenging how some sports-related contracts are treated, and court rulings have been flipping enforcement on a state-by-state basis, so the picture below can change between the time this is written and the time you read it.

Here is the map as of July 2026, and it moves, so treat these as pointers, not gospel, and confirm your own state against each platform's live eligibility page before you act:

• Kalshi is a CFTC-designated contract market available in 40-plus states, with restrictions, limits, or paused sports contracts in Arizona, Massachusetts, Maryland, Michigan, Montana, New Jersey, Nevada, and Ohio.
• Polymarket returned to the U.S. through a regulated acquisition and a CFTC order, with its U.S. exchange live since late 2025. As of July 2026 it is app-based on iOS and blocked in Arizona, Illinois, Massachusetts, Maryland, Michigan, Montana, Nevada, and Ohio.
• Minnesota is the hard stop. A state law criminalizing these markets is scheduled to take effect August 1, 2026, carrying felony-level exposure, and while it faces a federal court challenge, do not count on that going your way. If you are in Minnesota, do not trade these markets. Full stop.

Because the rules are shifting quarter to quarter as the CFTC and individual states work through jurisdiction, the only reliable source for whether you personally can trade is each platform's own eligibility check at the moment you sign up. Verify it there, every time. And the usual footing applies: this is a strategy that ties up real capital with real downside, so trade only what you can afford to lose, and never chase a loss with a bigger position.

Frequently Asked Questions

Is prediction market arbitrage legal in the U.S.? Trading CFTC-regulated event contracts is legal in the states where Kalshi and Polymarket are authorized to operate, and prohibited in others, with Minnesota criminalizing it as of August 1, 2026. The arbitrage itself is just holding offsetting positions across venues; the legality question is really "can I access these platforms where I live." As of July 2026 the answer is state-specific and changing, so confirm on each platform's eligibility page before trading.

How risky is it, really? Anyone who sells it as a done deal is selling you something. The math aims to return the same amount regardless of the outcome, but trading fees, spreads, slippage between your two legs, sportsbook limits, and capital tied up on both platforms all cut into or erase the gap. Treat it as a low-margin, low-risk-in-theory strategy with real execution risk, not a promise of profit.

What is the difference between prediction market arbitrage and regular sportsbook arbitrage? Sportsbook arbitrage crosses two books that both quote in American odds. Prediction market arbitrage crosses a venue that quotes in cents (an exchange) with either a sportsbook or another exchange, which adds a translation step, a fee structure, and a collateral cost that sportsbook-only arbing does not have. Our sportsbook arbitrage explainer covers the simpler two-book case.

How do I compare a Kalshi price to a sportsbook line? For the arbitrage check, convert each to what it costs to return $1.00. A 53-cent contract costs 53 cents; a +130 sportsbook price costs about 43.5 cents. If the two sides sum to under a dollar, a gap exists. Separately, OddsShopper's no-vig fair-odds pricing strips the book's margin to show the true probability, which tells you whether a contract is genuinely mispriced rather than just whether a raw book-versus-exchange gap exists.

Why do prediction market and sportsbook prices disagree at all? Different crowds, different mechanics, different clocks. The book posts a line with a built-in margin and adjusts to recreational money; an exchange price is set by traders buying and selling contracts against each other. When those two pools value the same outcome differently, the same question ends up cheaper on one venue than the other. That disagreement is the raw material for the whole strategy, and how you can see it depends entirely on the structural differences between the two.

The Bottom Line

Prediction markets did not invent arbitrage. They opened a new surface for it by pricing the same events a second way, in cents, set by a separate crowd trading on its own schedule. The gap that creates is genuine, and it is exactly the kind of cross-venue inefficiency that used to live only between sportsbooks. But it is thinner than the highlight-reel screenshots suggest, and once you account for fees, slippage, limits, and the collateral you have to post on both platforms at once, the winner is whoever can price the disagreement before it closes.

That speed comes down to the one move we kept returning to: getting both venues into the same language. A sportsbook line only becomes comparable to a Kalshi contract once you read both as the same thing, the cost to return a dollar, and then strip the vig to learn whether the gap is a real edge or a mirage. Doing that across every book and market in real time is the difference between spotting a gap and watching it close. That is the work our tools are built for, and it is the work worth getting right before you risk a dollar on either side.

Compare the true no-vig price on every sportsbook line, side by side, with the free OddsShopper odds tools, and unlock the full live arb feed and Sharp Action read with OS Pro, 20 percent off with code PMARB20 and a 7-day free trial.