H2 Gambling Capital projects the first year-over-year decline in NFL sports betting handle in nine years of legal wagering.
Key Takeaways
- H2 projects a $31.4-billion handle during the NFL season.
- Revenue is expected to fall 2% after last year’s operator “luck.”
- September could be a strong indicator of prediction market impact.
The marketing and data firm said in a release Wednesday that it estimates U.S. sportsbooks will generate $31.4 billion this NFL season, down 0.8% from last year and impacted by the rise of prediction markets heading into Wednesday’s New England Patriots-Seattle Seahawks game.
H2’s report said NFL U.S. sports betting still has underlying demand growth of around 4%, but prediction markets such as Kalshi and Polymarket are expected to cut into sportsbook handle by 4% to 6% in competitive legal states.
H2’s projections include Florida, which does not publish regulator data on sports betting handle, and are fairly in line with the American Gaming Association’s recent estimate.
“The American Gaming Association reaches the same conclusion by a different route: it puts legal NFL handle at $29.4 billion for the 2025 season, forecasts $29.5 billion for 2026 and describes it as the first season without meaningful growth, against $27.6 billion for 2024,” H2’s report said. “Our reporting-state figures of $28.7 billion and $28.8 billion are within 3% of the AGA on both years.”
Enjoying Covers content? Add us as a preferred source on your Google account
Jurisdictional impacts
The firm’s estimates account for Missouri’s first full season of legal sports betting after the state launched wagering in December 2025. Missouri is projected to account for $2 billion of the overall handle.
Including college football, H2 projects $42.1 billion in handle across the full market and $38.5 billion in reporting states.
H2 says total sportsbook handle in reporting states has been flat so far in 2026, with prediction markets taking around 4% of sportsbook handle. Florida, New Jersey, and New York have the heaviest overlap between sportsbooks and prediction markets. Kalshi and partner Robinhood have concentrated user bases in those jurisdictions.
“We assume that rate holds,” the marketing and data firm said. “Sports event contracts on Kalshi and other prediction market providers are now, on our handle-equivalent measure, about 35% the size of the combined sportsbook and prediction-market wagering pool, with 70-80% of that volume in the states without regulated online betting. In the competitive legal states, prediction-market consumer handle is equivalent to 8-11% of sportsbook handle.”
H2 said a significant reason for those percentages is that prediction market sites allow 18- to 20-year-olds to trade in jurisdictions where licensed sportsbooks generally require bettors to be 21 to open an account.
The firm also believes self-excluded and closed accounts, brokerage-native users, quasi-investment positions, and younger users account for 50% to 60% of the substitution. H2 expects the resulting 4% to 6% displacement of sportsbook handle to continue to rise.
Revenue down too?
H2 projects U.S. sportsbooks will generate a hold of nearly 11%, leading to a gross revenue of $3.5 billion, down 2% from the previous year, from August to February. The report does not expect operators to match the 2025-26 season’s 11.4% hold.
H2 said bettors had a great run early last season but forked funds back over to operators in November and December due to “luck.”
“Football itself ran hotter than the market last season: football hold, with parlays allocated, was 11.4% for August to February against 8.6% in 2024-25 and 8.9% in 2023-24, and relative to the market it was 1.11 times against 0.94 and 0.98 in the two prior seasons,” H2 said. “We assume a ratio of 1.05, the middle of the last four seasons, which gives an NFL hold of about 10.9% for 2026-27. Every 0.05 on that ratio is worth about $150 million of NFL gross win.”
H2 factors parlay margins into revenue projections. Multi-leg win rates continue to rise for sportsbooks, but that leads to lower recycling of winnings and impacts the amount wagered.
Market indicators
H2 said September will provide the first test of prediction market impact on handle. NFL handle is expected to fall around 25% year over year because the month has three Sundays instead of four.
“A fall materially larger than that would point to a bigger prediction-market bite than we have assumed; a smaller one would suggest promotional spend has been heavier than the earnings calls implied,” H2 said. “The hold in September will be flattered by free-bet stakes and should be read with the net win figure alongside it.”
The firm also believes the handle will be higher in December and January, which includes Week 17 and Week 18 this year, than last season. However, another stretch of operator-friendly results could result in better gross win but less wagering.
“Either way, the relationship between one month's hold and the next month's handle is now strong enough that monthly forecasts should be updated as each month's hold becomes known, and H2 will do so in the weekly NFL results assessment and the monthly state data updates through the season,” the report said.
