The $2.4 billion number, and what it is actually measured against
A record handle is not a record month for the companies taking the bets. New York just proved the point: the New York sports betting handle topped $2.4 billion in September while operator revenue for the same broad period went backwards.
The New York State Gaming Commission reports bettors wagered more than $2.4 billion for the month, a 5% increase on September 2025 and an uptick of nearly $114 million. Through September, 2026 volume is running 3.7% ahead of last year, roughly $697 million in additional wagers. Across the state’s 2026–27 fiscal year, the first six months from April produced almost $12.6 billion in handle, up 6.9% on the same stretch in 2025.
One clarification worth making before the number gets repeated too confidently. The comparison the commission’s figures actually support is year over year, September against September. They do not establish that this was the largest single month in the state’s history, and I am not going to assume it was. Treat $2.4 billion as the best September New York has reported, not as an all-time peak.
Where the extra wagers came from
What the commission’s report counts
The $2.4 billion is online volume from New York’s licensed mobile sportsbooks, which NY.gov lists as eight: Bally Bet, BetMGM, Caesars Sportsbook, DraftKings Sportsbook, theScore Bet, Fanatics Sportsbook, FanDuel Sportsbook and Resorts World Bet. Mobile sports betting in NY is what drives these headline totals; retail sportsbooks at upstate casinos are accounted for separately and are not the source of this figure.
If you see a precise mobile-versus-retail percentage attached to this September number, it did not come from this report. The split exists, but putting a decimal point on it here would be invention.
September is the first full NFL month
Timing explains most of the seasonal shape. September is when the NFL regular season starts and college football fills the Saturdays, which reliably lifts handle across every legal state. What it does not explain is the size of the year-over-year gain, and 5% is the profile of a mature market growing steadily rather than one adding customers in a rush.
The usual suspects get named at this point: parlay mix, bonus bets, higher bet frequency per account. All plausible, all genuinely influential in US sports betting trends generally, and none of them visible in this dataset. New York’s monthly release does not break out bet type or promotional spend, so crediting parlays for the record would be a guess dressed up as analysis.
Prediction markets and the argument over where the money went
Prediction markets are exchanges where users buy and sell event contracts: yes-or-no positions on an outcome, priced in cents and settling at a fixed value once the event resolves. Platforms including Kalshi and Polymarket list contracts on game results, which lets users take positions on sports under federal Commodity Futures Trading Commission oversight rather than a state sportsbook licence.
How an event contract differs from a sportsbook bet
With a sportsbook, you bet against the house. The operator sets the price, builds a margin into it and carries the risk on the outcome. On a prediction market, participants trade against each other and the platform earns from fees and spread rather than from a built-in overround. That distinction is the whole basis of the platforms’ defence: they argue users are trading, not wagering against a bookmaker. Regulated operators argue the customer experience is close enough to a bet that the label is cosmetic. If you want the mechanics in more depth, this plain-English breakdown of prediction markets versus sports betting covers the structural differences.
The regulatory question is open, and nobody should pretend otherwise
The American Gaming Association’s position is blunt: prediction platforms are “offering illegal sports betting nationwide outside the state and tribal regulatory frameworks that protect consumers,” bypassing licensing, consumer protections and taxes. The AGA estimates states have lost roughly $1 billion in gaming tax revenue to prediction markets, and some AGA figures cite up to $1.6 billion. State-level enforcement is live too, including Utah’s criminal charges against Kalshi’s founders.
Here is where the evidence stops. The $1 billion figure is an estimate published by a trade association representing the companies that lose if prediction markets win, not an audited figure, and New York’s rising handle is the counterpoint market proponents reach for. Neither claim is settled by a monthly handle report, because the thing both sides are arguing about is a counterfactual: how much would have been wagered with the eight licensed platforms had event contracts not existed. That number is not observable. Anyone quoting it to a decimal place is modelling, not measuring.
Record handle, smaller revenue: how hold works
Handle is the total amount wagered. Revenue is what operators keep after paying out winning bets. The two move independently, and in New York’s fiscal half-year they moved in opposite directions: handle rose 6.9% while gross gaming revenue fell 5.8% to $1.1 billion, because the average win percentage, or hold, dropped from 10% to 8.8%.
| Metric (April–September, NY fiscal 2026–27) | Figure | Change vs 2025 |
|---|---|---|
| Sports betting handle | Nearly $12.6 billion | +6.9% |
| Sportsbook gross revenue | $1.1 billion | −5.8% |
| Average win percentage (hold) | 8.8% | Down from 10% |
That 1.2-point drop in hold is the entire story. On roughly $12.6 billion of handle, 1.2 percentage points is about $151 million of revenue that did not materialise, which is why volume can set a record while US sportsbook revenue declines. The commission’s explanation is simply that bettors fared better over the six months. Results, in other words, not a structural shift.
Where competitive pressure enters is in what comes next rather than what already happened. When two sets of platforms chase the same customer, the standard response is more promotional spend and tighter pricing, both of which compress hold further. That is an expectation based on how competitive markets behave, not something this dataset demonstrates. Watch hold percentage over the next few reporting periods; it will tell you more about the threat than handle ever will.
What New York can and cannot tell you about the rest of the market
New York is the largest of the 39 states with legal sports betting, which makes it the most quoted and the least representative. The trend lines elsewhere do not point the same way.
| State | Handle | Period | Change |
|---|---|---|---|
| Illinois | $9.3 billion online, pro and college | Through August 2026 | +5.7% from $8.8 billion |
| New Jersey | $8.3 billion | Through September 2026 | −4.1% from $8.6 billion |
Two of the three largest markets grew, one shrank. New Jersey’s 4.1% decline is the figure both camps will reach for, and it has several candidate explanations, none of which these reports isolate. Attributing it cleanly to prediction markets is the same counterfactual problem in a different state.
What to watch instead of the headline
The honest summary of New York’s September is narrower than the headline suggests: volume is still growing in the biggest regulated market, margins got worse for reasons the regulator attributes to betting results, and the legal status of the platforms competing for that volume is unresolved in the courts. The sports betting record and the competitive threat are both real, and neither one proves anything about the other yet.
For anyone betting rather than tracking the industry, the hold figure is the useful part. An 8.8% average win percentage means the eight licensed platforms kept close to nine cents of every dollar wagered over six months, and that edge does not disappear in a good month for bettors; it reasserts itself over time. Set a deposit limit, treat the money as spent when it is staked, and use the self-exclusion and cool-off tools the state requires operators to offer. If betting has stopped being entertainment, the New York Problem Gambling HOPEline is free and confidential.