Legal sportsbooks are projected to take $29.5 billion in NFL wagers this season, up just 0.3% from last year. That near-flat number, released by the American Gaming Association on September 4, is the first time NFL handle growth has essentially stalled since legal betting expanded in 2018. The AGA isn't blaming market saturation. It's blaming prediction markets, chiefly Kalshi and Polymarket, which it says have diverted more than $1.3 billion in potential state gaming tax revenue since 2025.
AGA CEO Bill Miller called it "backdoor sports betting," and the framing isn't subtle. Prediction markets operate under CFTC oversight rather than state licensing, which means they trade nationwide, including the states that never legalized sportsbooks, and they don't pay state gaming taxes. The AGA has already asked Congress to outlaw gaming on these platforms outright.
The league picked a side this week
On September 15, the NFL confirmed an outright ban on prediction market advertising. Not in games, not in stadiums, not through club sponsorships or player endorsements. The rule goes further than most coverage suggests: sportsbook partners like DraftKings, FanDuel, and Fanatics can only advertise their approved sportsbook products and can't use league IP or in-game inventory to promote their own prediction market arms. The sponsorship deals license products, not companies, and the NFL has made that distinction sharp.
The league also sent its strongest letter yet to operators, pressing Kalshi and Polymarket to drop contracts on player injuries and availability, officiating, and easily manipulable props like first-play outcomes. Meanwhile, the AGA warned the NBA and MLB against signing commercial deals with what it calls insufficiently regulated platforms. The league's sponsorship structure now functions as a de facto moat for traditional books inside NFL-controlled inventory, even as those same operators run competing prediction products outside it.
Record volume, murky comparison
The first full NFL weekend produced roughly $6 billion in trading across prediction platforms, with Kalshi alone accounting for about $4.89 billion on Saturday and Sunday. That's a record. But here's the nuance the AGA glosses over: trading volume and betting handle aren't the same thing. Prediction markets trade shares back and forth, so $6 billion in volume doesn't mean $6 billion left sportsbooks. Some of it is capital recycling between traders, some is genuinely new money from the 11 states without legal sportsbooks and from 18-to-20-year-olds who can trade on these platforms but can't walk into a sportsbook.
That last point is the one I keep coming back to. Spending time in betting communities over the past week, the split is clear. Plenty of bettors see the substitution plainly, same fans, different rails. Others treat the flat handle as a growth ceiling story for sportsbooks rather than a loss, and there's real doubt about whether these platforms ever face the same tax and regulatory pressure or just keep operating as "investments."
Prediction markets are capturing the same NFL betting dollars through completely different rails. Nationwide access, 18+ entry, no state gaming tax. That's not competition, that's a parallel market. SEPTEMBER 8, 2026
— Kevin McD (@kevinmcd) SEPTEMBER 8, 2026
The complaints aren't all regulatory either. Bettors gripe about Polymarket's product quality, and there's legitimate unease about injury-related contracts creating bad incentives around player health and medical privacy. A few smaller platforms have voluntarily restricted certain contracts, which tells you the pressure is landing. We've been tracking the regulatory fight closely in our coverage of CFTC oversight of sports prediction markets, and the litigation in New York is only getting started.
So what should you watch? Whether the NFL's ad ban actually slows platform growth or just pushes marketing underground. Whether Congress moves on the AGA's ban request or lets the CFTC patchwork fester. And whether hybrid products, social apps layering prediction trading on top of DFS-style play, blur the line entirely.
My honest read: the AGA's zero-sum framing is mostly right but incomplete. Prediction markets are eating sportsbook dollars in some places and expanding the betting market in others, and the industry's real problem is that the side with the better regulatory story built a worse product experience for eight years and got comfortable. The flat number isn't a crisis yet. It's a warning shot, and the NFL just fired its own back.