Kalshi cleared $2.42 billion in volume on the busiest college football Saturday on record, then broke it again a day later with $2.46 billion on the first NFL Sunday. Stack up the six busiest prediction exchanges, Kalshi, Polymarket US, DraftKings, the Crypto/FanDuel/Fanatics trio, Novig, and Robinhood's Rothera, and the weekend cleared $6.2 billion. Kalshi alone owned 63% of sports volume and 79% of everything traded.
Those numbers are the headline everyone in this space has already screenshotted, and they're not fake. What they are is misleading if you use them the obvious way, which is to hold them up next to legal sportsbook handle and conclude prediction markets are suddenly playing in the same weight class. They're not there yet. But the more interesting story isn't the gap, it's how fast these platforms are spending to close it anyway.
The record numbers are inflated by design, not by accident Notional volume
counts both sides of every trade, so the crude fix is cutting the total in half. Prediction markets inflate the number a second way on top of that: a $10 bet at odds equivalent to 10-1 buys 100 shares worth a dollar each, turning $10 of real money into $100 of volume. The honest discount is deeper than 50%, not exactly it.
Halve the weekend's sports volume anyway and $5.72 billion becomes $2.86 billion. Legal sportsbooks took in about $15.5 billion last September among the 54% of over-21 Americans with access to them, or $3.8 billion a week across four NFL weekends that each also carried a full MLB slate. On the surface, prediction markets look close to pacing with legal sports betting nationally. They aren't. Fifty-five percent of that prediction volume, $3.45 billion, came from parlays, the most inflated bet type there is. Straight football bets, the cleaner comparison, totaled just $1.26 billion for the entire weekend.
The season-long projection is the number that actually matters
Gambling research firm Eilers & Krejcik has done the harder work of modeling the full season, projecting prediction operators will generate roughly $8.4 billion in handle-equivalent volume across the NFL year, about 21% of a combined $40 billion market shared with sportsbooks. Sportsbook handle itself is projected up 8% year over year, with parlays making up 40% of it and in-play betting another 30%. That framing puts Kalshi's rise in its proper place: a real, fast-growing slice of a bigger pie, not a takeover in progress.
The celebrity ad spend is the tell that these companies believe their own hype
If the volume math is murky, the marketing spend is not. Polymarket rolled out a Peter Berg-directed campaign stacked with LeBron James, Derek Jeter, and Eli Manning, plus Sue Bird, Reggie Bush, Rajon Rondo, and Spike Lee. Kalshi answered with Marshawn Lynch going full Beastmode in a Fast and Furious homage, after already running a Pete Sampras spot ahead of the U.S. Open semis. Novig's Sydney Sweeney ad broke the internet on release and again during the backlash, which likely helped the app more than hurt it.
That's tens of millions in celebrity-driven spend chasing a market whose real size is still being sorted out in back-of-envelope math. Volume is the right tool for ranking Kalshi against Polymarket, DraftKings, and the rest of the field. It's the wrong tool for measuring any of them against a sportsbook industry they haven't caught yet, no matter how loudly the record-breaking weekends suggest otherwise.
The billion-dollar Sundays are real, the growth curve is real, and the ad campaigns prove the companies chasing this market believe the trajectory more than the current total. Prediction markets are building something adjacent to sports betting, quickly, not replacing it, and that distinction is worth more than any single record-setting weekend.
