Football season is quickly approaching, and you've probably seen or heard a million things about prediction markets in the ramp-up. The ads are everywhere, every app suddenly has an opinion about your Sunday, and somewhere in the noise you still have to answer one practical question: which sites should you actually be signed up at right now, and why? So I scored my favorite preferred prediction market platforms, the five names that matter most for a U.S. bettor this fall: Kalshi, Polymarket, Novig, ProphetX, and Underdog Predict.
I graded each one on the four things that decide whether you can actually trade well there: state access, sports coverage, fees, and liquidity. And I'll tell you up front why the grades need to exist at all. The sticker price lies. The same 40-cent contract costs a different real amount on almost every one of these sites once fees touch it, and I'll show you that math further down, in cents, so you can see exactly what your account pays. That's the part every "best sites" list skips, and it's the part that compounds all season.
One framing note before the grades. None of these are sportsbooks. They're exchanges for event contracts, regulated federally by the Commodity Futures Trading Commission (CFTC), where a Yes contract settles at $1 if the event happens and $0 if it doesn't. The price is the probability. If that mechanic is new to you, our guide on how prediction markets work covers it in full.
The Quick Answer
Kalshi is the best prediction market site for most U.S. bettors in 2026: the broadest established state access, the most complete sports menu, and by far the deepest liquidity in the category. Polymarket is the pick if politics and culture markets are your reason for signing up. For price hunters, ProphetX is the sleeper, Novig is the fee schedule betting on itself, and Underdog Predict is the crossover play for pick'em players. The full five-site scorecard, the fee math in actual cents, and the state-access fine print are all below.
Free: The Weekly PM Market Brief — the 8-model panel's graded record, the week's biggest market-vs-model gaps, and what's spiking next. One email, Sundays. These are market prices and model estimates, not predictions of fact and not financial advice. Kalshi is a CFTC-regulated event-contract exchange (18+; availability varies by state).
Start With The Money: Who's Actually Big
Before I hand out grades, look at where the dollars already went. I posted this table on X this week, from an Underdog compilation of notional volume flow since September 2025, and it tells you the shape of the whole industry in five lines:
| Platform | Notional volume since Sept 2025 |
|---|---|
| Kalshi | $86.5B |
| Robinhood | $32.5B |
| Underdog | $6.49B |
| Polymarket | $6.48B |
| DraftKings | $1.2B |
Two rows in that table do most of the talking. First, Kalshi's $86.5B is 65% of everything listed, on its own, and the second-place app launched its prediction hub on Kalshi's rails in the first place (more on Robinhood near the bottom). Second, look at Underdog edging Polymarket. The pick'em company out-trafficked the most famous prediction market brand on earth, which tells you how much of this category's growth is sports, not politics. Hold onto both facts. They explain almost every grade below.
How I Scored The Five
Four dimensions, weighted in the order a bettor actually feels them:
- State access. The best platform on earth is worthless if it can't legally serve you. Availability is volatile and platform-specific, so this grade reads breadth, not a memorized state list.
- Sports coverage. Football sign-ups want game lines, futures, and props, not just an election board.
- Fees. What a trade really costs after fees, not the sticker price.
- Liquidity. A tight, deep order book. A thin market with a seven-cent spread quietly charges you more than any posted fee.
The grades are my read of each platform's documented structure as of mid-August 2026, not a measured dataset, and the rules change fast enough in this category that you should treat every row as a starting point. Where a platform is too new to grade honestly, I say so instead of inventing a number.
The Scorecard: My Five, Four Grades Each
| Site | State access | Sports coverage | Fees | Liquidity | The verdict |
|---|---|---|---|---|---|
| Kalshi | A | A | B | A | The sports-first pick |
| Polymarket | B | B | B | A on marquee markets | The politics-first pick |
| Novig | B | B (sports only) | A | Too new to grade | The fee play, if it deepens |
| ProphetX | A (49 states) | B+ (sports only) | A- | C+ (still building) | The name-your-price play |
| Underdog Predict | B- | B | C+ (unpublished) | B- | The pick'em crossover |
The column that decides the most accounts is the first one, and it's the least glamorous. For plenty of readers the "which site" debate ends before fees or liquidity ever come up, because only two or three of these operate where they live. That's why the legal section near the bottom of this page isn't boilerplate, and it's why our full ranked platform comparison keeps state access as its through-line too. What the scorecard can't show is how differently these five make their money, so let's go one at a time.
Kalshi: The Sports-First Pick
Kalshi wins the overall grade because it wins the two columns bettors feel first. It's a CFTC-designated contract market with the broadest established U.S. footprint in the category, and its sports menu is the most complete on this list: game lines, futures, and a growing props board across the NFL, college football, and the other major leagues. You just saw the liquidity case in dollars: $86.5B in notional flow since last September, 65% of the tracked field by itself, plus whatever slice of Robinhood's flow still routes through its rails. During the 2026 World Cup, prediction markets handled about 27% of U.S. legal sports-betting volume tied to the tournament, up from roughly 9% at the start of the year, per H2 Gambling Capital estimates reported by Bloomberg, and Kalshi's market on the World Cup champion alone traded more than $1.2 billion, the biggest single market in the platform's history. The depth is real, and on headline sports markets the spreads generally reflect it.
The B in the fee column is the honest note. Kalshi charges takers a fee of roughly 7% × price × (1 − price) on most markets, which is small in absolute cents but real, and I'll show you exactly how real in the fee-math section. Our Kalshi fees breakdown keeps the current schedule, and how to bet sports on Kalshi is the full walkthrough if it becomes your pick.
Polymarket: The Politics-First Pick
Polymarket's U.S. exchange went live in December 2025 under CFTC regulation, after it acquired a registered exchange (QCEX) in mid-2025, and it dropped its waitlist in May 2026. Its identity is depth on the biggest questions: elections, geopolitics, culture. When one huge market is the argument everyone is having, Polymarket is usually where the most money is having it, and that concentration is exactly what you want as a trader, because depth means tighter spreads on the contract you actually care about.
The sports grade is a B because the U.S. sports menu is still expanding behind Kalshi's, and the volume table above is the receipt: for all of Polymarket's fame, its flow since September trails a pick'em app's. A football-first reader would be signing up for the second-best sports catalog. A politics-first reader flips that logic completely. On fees, the U.S. exchange charges takers 0.06 × price × (1 − price) per contract, capped per fill, and pays a small rebate to resting maker orders; our Polymarket fees guide runs the current schedule. The two leaders differ enough, state by state and market by market, that the Kalshi vs. Polymarket head-to-head is worth ten minutes before you fund either.
Signing up for Polymarket? Our link carries offer code OS3, a $20 trading bonus after a $10 deposit and your first qualifying trade (credits once the trade is placed) for new users: claim it here. It's a trading bonus you use to take positions, not instant withdrawable cash. Full terms and eligible states are in our Polymarket sign-up bonus guide.
18+ Only. Restrictions and eligibility requirements apply. Not available in all jurisdictions. Trading is risky. 100% loss can occur. See polymarket.com/tos for more information. The Polymarket US App serves as an independent software provider and affiliate of Polymarket US and Polymarket Clearing, the CFTC-regulated exchange and clearing organization. OddsShopper is a Polymarket partner and we may earn a commission if you sign up through our link.
Novig: The Fee Play, If It Deepens
Novig is the newest name on the board and the most aggressive fee story. It relaunched on August 4, 2026 as a real-money, CFTC-regulated exchange (its designation came in June 2026 through Ludlow Exchange, LLC), it lists sports exclusively, and its published schedule is the one that turns heads: maker orders free, pre-game straight-market taker orders free, with formula-based fees on live trading (0.03 × price × (1 − price)) and parlays. Free pre-game taker fills on a regulated exchange is a real price advantage, full stop, and the fee table below shows what it's worth.
Two honest caveats keep it out of the top spot. First, it's weeks old in this form, so its order books haven't had time to prove depth; a zero-fee fill doesn't help if the spread is wide the day you want out. Second, its footprint is still being contested, including a lawsuit it filed against New York within a day of relaunching, and it sets its own minimum age at 21 where the CFTC exchanges above run 18+. If the liquidity fills in behind that fee schedule, this grade climbs. I'd rather tell you that plainly than pretend a weeks-old order book has a track record.
ProphetX: The Name-Your-Price Play
ProphetX is the one I'd tell line shoppers to look at first, and I don't say that lightly. It's a peer-to-peer, CFTC-regulated prediction market exchange (dual DCM and DCO registration in June 2026, nationwide launch June 18, 2026) that runs in 49 states, everywhere but Nevada. That's the single broadest map on this list, and it earns the only other A in the state-access column. The distinction from Kalshi's A is worth naming: Kalshi's footprint is the established one, tested in court and running at scale for years, while ProphetX's 49 states are newer, on-paper breadth from a June 2026 launch. The sports menu is football-ready: NFL moneylines, spreads, totals, player props, and season-long futures, though sports is all it does, so there's no election board here.
The mechanic that makes it different: you don't have to take the posted price. ProphetX lets you post your own, like a limit order, and wait for another trader to meet you. The fee model rewards that patience: no charge to place a trade, and a 2% commission on net winnings on straight trades, so you pay only when you win. Parlays currently carry no commission at all. New users also get a Trade $10, Get $20 offer, credited within 24 hours and subject to terms; our ProphetX sign-up bonus guide walks through it. The knock, and it's the same one Novig wears, is depth: a peer-to-peer book needs two sides, and away from headline NFL markets you'll sometimes post a price and wait. Our full ProphetX review covers the exchange in detail, and the ProphetX vs. Novig head-to-head settles which of the two young books fits you better.
Underdog Predict: The Pick'em Crossover
Underdog is the row in that volume table that should make you look twice, and here's the story behind it. The pick'em giant started routing prediction-market trades through partner exchanges last fall, then went and bought the whole toolchain: in March 2026 it acquired a Designated Contract Market and a Derivatives Clearing Organization (rebranded UDX), and on July 18, 2026 it launched prediction markets on its own exchange, making it the first sports company holding the full DCM, DCO, and FCM license stack. That $6.49B in flow since September is the receipt, third among U.S. operators, ahead of Polymarket.
The grades are middling anyway, and honestly so. State access is a B- because the map covers more than 30 states plus D.C. but excludes the big legal-sportsbook states, New Jersey, New York, Pennsylvania, Arizona, and Massachusetts among them, so check your state before you care about anything else in this section. Sports coverage is a B: the menu launched with baseball and basketball contracts and is expanding into football season, but its own exchange is weeks old. And the fee grade is the one I want you to sit with: Underdog doesn't publish a posted commission for Predict, so your real cost lives in the spread between the two sides of each market.
The line to remember here: unpublished isn't the same as free. When no fee is posted, the spread is the fee, and only the order book can tell you how big it is.
If you're coming from pick'em, our Underdog Predict review is the full translation guide, and the one-line version is this: the price in cents is the probability, and your edge is knowing when that price is wrong.
The Fee Math: What A 40-Cent Contract Really Costs
Here's the promise from the top, paid off. Take the same Yes contract priced at 40¢, a price whose sticker payout converts to +150, and buy it as a taker on each site. Here's the real cost per contract:
| Site | Fee on a 40¢ contract | True cost | Effective odds |
|---|---|---|---|
| Novig (Pre-Game Straight) | $0 | 40.00¢ | +150 |
| ProphetX (Straight) | 2% of net winnings, only if you win: 0.02 × 60¢ = 1.2¢ | 40¢ up front, 58.8¢ profit on a win | +147 |
| Polymarket | 0.06 × .40 × .60 = 1.44¢ | 41.44¢ | +141 |
| Kalshi | 0.07 × .40 × .60 = 1.68¢ | 41.68¢ | +140 |
| Underdog Predict | Unpublished; the spread is the cost | 40¢ + whatever the book is quoting | Depends on the spread |
Same market, same sticker price, and a 10-point swing in effective odds between the cheapest and priciest posted schedules. Two footnotes on precision. ProphetX's row is the win case: its 2% only bills when you cash, while the other posted fees are charged win or lose, so ProphetX's expected cost is actually lower than its displayed number. The table also shows each schedule's underlying rate rather than a billed amount; an exchange can round a computed fee up to the next whole cent on a small order, so a single-contract fill may bill a shade more than the formula implies. That's not a reason to panic over a cent and a half, but it is the correct unit of comparison, and it stacks up over a season of trades.
Two practical notes, and the second one matters more. Resting maker orders dodge most of these fees; every posted schedule here prices makers at zero or better on standard markets, and on ProphetX naming your own price is the whole product. And the fee is only half the true cost. The other half is the spread, which is a liquidity question, which is why a young zero-fee book can still be more expensive than a deep 7%-fee one on the day you need to exit a position. Fees are printed; spreads are earned.
Whichever Site You Pick, Check The Price Before You Take It
The fee table tells you what a contract costs. It still doesn't tell you whether the price is any good, and that's the discipline that separates trading a number because it's in front of you from trading it because it beats the market. A 40¢ contract is the exchange saying 40%. The way to test that claim is to hold it against the sportsbook market for the same outcome, with the books' margin stripped out. OddsShopper's no-vig pricing does that step for you, converting any sportsbook line into a true implied probability, the exact 1¢-to-99¢ unit these contracts trade in, and the live odds screen lets you shop the number across every major book while you're at it. If de-vigging is new, how to remove the vig and implied probability explained teach the mechanic in ten minutes.
The move in real numbers: say Sunday's game shows the favorite at -130 on DraftKings and the underdog at +110 on FanDuel. Read straight off the lines, those prices imply a 56.5% and a 47.6% chance, and they sum to 104.1% because the vig is baked into both sides. Strip it out across the pair and the fair numbers fall out:
| Side | Book price | Implied probability | Fair (no-vig) probability |
|---|---|---|---|
| Favorite | -130 (DraftKings) | 56.5% | 54.3% |
| Underdog | +110 (FanDuel) | 47.6% | 45.7% |
That fair 54.3% is a true 54¢ in contract units. Hold it next to the same outcome's Yes contract on whichever exchange you picked: at 50¢ you're buying under the sportsbook-implied fair price before fees and spread, and at 58¢ the exchange is charging a premium and you pass. Every one of these contracts settles Yes at $1 or No at $0, and the probabilities our own model panel publishes are graded against those settlements in public on our Model Verdict scoreboard, so you can judge the process by its record.
The OddsShopper Liquidity Tool is the second check, and it's built for exactly this category: it tracks the real-money orders hitting prediction exchanges like Kalshi and Polymarket, reads each order to the sharp side, and shows the dollars behind the signal. Price says 40%; the order flow tells you who's pushing it there and with how much conviction. We run this comparison in public every week for football futures in our Kalshi NFL futures vs. sportsbooks breakdown, and the broader prediction markets vs. sports betting guide covers when each venue deserves your dollar.
New to OddsShopper? It turns every sportsbook line into a no-vig true probability and shows the real-money flow behind prediction-market prices, so you can tell a good 40¢ from a bad one before you click. New subscribers get a free 7-day trial, and code LINDY25 takes 25% off your first bill of OddsShopper Pro if you stay on.
Want ready-made ideas while you learn the screens? Browse the free expert picks today and watch how the prices are argued, not just what the picks are.
Who Didn't Make My Five
You noticed Robinhood and DraftKings in the volume table and not in the grades, so let me close that loop. Robinhood's $32.5B is real money, but it's a routing story more than a book of its own: the prediction hub launched on Kalshi's exchange, and since June 2026 Robinhood has been shifting that flow to Rothera, the CFTC-licensed venue it co-owns (the former MIAXdx), with some contracts still routing through Kalshi. If your money already lives there it's a fine on-ramp; just know that which order book you're actually trading on varies by contract. DraftKings Predictions and FanDuel Predicts are the sportsbook brands' event-contract products, rolling out state by state for their own loyalists, and DraftKings' $1.2B row tells you how early that story still is. ForecastEx is Interactive Brokers' venue for macro and rates hedgers, with no sports menu to grade. And PredictIt still gets referenced around the web, but it has operated under prolonged legal uncertainty with a limited political scope. None of them are where I'd send a bettor first this fall, and this page is about where I'd actually send you.
State Access And The Legal Picture, As Of Mid-August 2026
Every platform above operates under CFTC oversight, which is a real federal structure and the reason these aren't offshore books. It is not the same thing as settled law. The rules are being written in real time, and four dated facts frame where things stand:
- April 6, 2026: the Third Circuit (KalshiEX, LLC v. Flaherty) affirmed a preliminary-injunction win for Kalshi, finding the CFTC likely holds exclusive jurisdiction over sports event contracts. A preliminary-posture win, not a final merits decision, but the current law of the land.
- July 24, 2026: the CFTC issued its second advisory of the year, warning exchanges against "blanket" self-certifications that bundle multiple settlement methodologies under one filing.
- July 27, 2026: a federal judge halted Minnesota's law aimed at the companies that operate or advertise these markets, reasoning that federal commodities law likely preempts it. That fight is ongoing, and the law targets operators, not individual traders.
- Still Pending: a Senate bill introduced in March 2026 that would treat prediction markets as gambling has not passed, but it tells you which way part of Congress is leaning.
What that means for you is simpler than the docket: availability is platform-specific, state-specific, and changeable by a single court order. ProphetX's 49 states, Kalshi's broad footprint, and Underdog's 30-plus all sit on different maps, and none of them track sports-betting law, so check eligibility inside each app, on the day you sign up, rather than reasoning from what's legal elsewhere. Our are prediction markets legal explainer goes state by state shape, and none of this is legal advice. One more constant no matter the venue: contracts settle to zero when you're wrong, so only trade money you can afford to lose, and size for the market you're actually in.
Best Prediction Market Sites FAQ
What is the best prediction market site in 2026? Kalshi, for most U.S. bettors: the broadest established access, the most complete sports menu, and $86.5B in notional volume since last September, which is more than the rest of the field combined. If marquee politics and culture markets are the draw, Polymarket takes it instead.
What is the best prediction market site for sports? Kalshi has the most established sports coverage and the deepest books. ProphetX covers 49 states with NFL lines, props, and futures and lets you post your own price. Novig is sports-only with free pre-game taker fills, and Underdog Predict makes the most sense if you're already a pick'em player crossing over.
What is the cheapest prediction market site? Cheapest depends on which cost you count, because the posted fee is only half of it and the spread is the other half. On posted fees alone it's Novig, with free maker and pre-game straight taker orders, and ProphetX runs close behind at 2% of net winnings, charged only when you win, where the other exchanges bill win or lose. But a deep, tight book with a small taker fee, which is Kalshi's profile, can beat a free fill on a thin one the day you need to exit. Compare effective odds on the market you actually trade, not sticker prices.
Are prediction market sites legal in the U.S.? The exchanges in this guide operate under CFTC regulation, and event contracts are 18+ on the major exchanges (Novig sets its own minimum at 21, and Underdog runs 19+ in Alabama and Nebraska). Availability is state-specific, contested in places, and changeable, so confirm eligibility in the app before funding an account. This is not legal advice.
Can I use more than one prediction market site? Yes, where you're eligible, and it's often the sharp move: one site for breadth, another for depth on the markets you care about, and the ability to take whichever venue posts the better effective price on a given contract.
- Bass Vs Raman Odds: The Fake Poll Kalshi Didn't Buy
- Where Does Bitcoin End 2026? Kalshi's Price-Band Odds
- BMW Championship Top 5 Odds 2026: Kalshi Odds
- Draymond Green Next Team Odds: Kalshi Odds
- How To Bet College Football On Kalshi: Markets & Value
The Bottom Line
I've said it before and the volume table says it louder: prediction markets aren't just the future, they're the present. Score my five on what actually hits your account and the picture is clean. Kalshi is the sports-first pick on access, menu, and depth. Polymarket is the politics-first pick on marquee-market liquidity. ProphetX is the price-setter's exchange with the biggest map, Novig is the fee schedule betting on itself to grow a book, and Underdog Predict is the bridge for pick'em players who already think in probabilities. And carry the fee table with you as the tiebreaker: the same 40¢ contract can cost 40.00, 41.44, or 41.68 cents depending on the venue, before the spread says anything. Sign up where you're eligible, price what a trade really costs, and check every contract against the no-vig number before you take it. That habit beats any logo on this list.
Read every price like it's your money, because it is. OddsShopper turns sportsbook lines into no-vig true probabilities and tracks the real-money order flow on prediction exchanges, free for 7 days for new subscribers, then code LINDY25 for 25% off your first bill of OddsShopper Pro.
Start your free trial t=freetrial)
