Developing FanDuel Says Its Median Bet Is $5. The Only Account-Level Study Ever Published Says the Top 5 Percent of Bettors Pay Two-Thirds of the Bill

Editor's note: TrueEdge builds odds tools and earns affiliate commissions from licensed sportsbooks, including FanDuel. This piece argues a position on how a sportsbook should answer its regulators. Weigh it with that in mind.

The median FanDuel bet is $5. The median FanDuel bettor spends about $63 a month, less than a fifth of what the average American spends eating out. Eighty-four percent of sports bettors wager monthly or less often. Ninety percent of same-game parlays are for $30 or less. FanDuel published all of that on Friday, October 16, in a post titled "New research: sports bettors in America," and Covers and Casino.org wrote it up on Monday.

Nothing in it is false, as far as we can tell. The problem is that none of it is responsive. A median describes the customer in the middle of the queue. The three members of Congress who wrote to FanDuel on October 6, and whose answers are due on Tuesday by PlayUSA's account of the letter, were not asking about the middle of the queue. They were asking about the man at the end of it who lost $1.5 million and got a video from Bryce Harper.

What Friday's post says, and whose numbers they are

Read closely, the post is two documents stapled together. One is FanDuel's own data: the $5 median wager, the $14 average same-game parlay, the deposit-frequency claim that the average customer deposits less than once a week, the roughly 30 personalised athlete and celebrity videos sent over two years. The other is a published academic paper that FanDuel is citing rather than producing.

That paper is Scott Fulford's "Online sports gambling and financial wellbeing in the United States," published in August in the Journal of Financial Literacy and Wellbeing. It uses the Consumer Financial Protection Bureau's Making Ends Meet survey from January 2025, 2,630 respondents, weighted to the adult population. The headline findings are the ones FanDuel quotes: 60 percent of sports bettors typically wager $20 or less, 68 percent report a largest single-day loss of $50 or less, 84 percent bet no more than monthly, half bet only a few times a year and 22 percent bet once.

Fulford also finds what the post does not quote. Five percent of sports bettors lost $500 or more in a single day. A third of them, 33.8 percent, met his definition of high-intensity gambling: weekly or more, a typical bet over $50, or a single-day loss of $100 or more. Sports gamblers were significantly more likely to carry revolving credit-card debt. He does not find a statistically significant association between sports betting and lower financial wellbeing in the population as a whole, and that finding is genuinely good for FanDuel's argument; we will come back to it. But the paper's shape is a long tail, and the post cites the body and leaves out the tail.

The study that answers the actual question

Nobody in the United States has published account-level sportsbook data showing how operator revenue is distributed across customers. One group has published it anywhere: David Forrest and Ian McHale of the University of Liverpool, who were given 139,152 online accounts by seven British operators covering the year to June 30, 2019, operators that together held 85.5 percent of the regulated British online betting market. Their paper on the data appeared in the Journal of Gambling Studies in 2024.

For betting specifically, 110,203 accounts, the top 1 percent of customers by spend produced 36.42 percent of operator win. The top 5 percent produced 66.72 percent. The top 20 percent produced 88.97 percent. The remaining four-fifths of customers, the ones a median describes, produced 11 percent of the money.

A $5 median and a two-thirds-from-5-percent distribution are not in tension. They are the same distribution viewed from different points. A business whose median bet is $5 and whose revenue depends on its top twentieth is a business whose marketing, hosting and retention spend will, if it is rational, follow the revenue. That is what a VIP programme is. It is not a scandal that FanDuel has one. The scandal the lawmakers are circling is whether the programme's targeting criteria and its player-protection criteria draw on the same data and point in opposite directions, which is the sixth of the six questions ReadWrite lists from the letter, and the one Friday's post does not approach.

Terry Thompson is the case that made the question concrete. The Public Health Advocacy Institute sued FanDuel and DraftKings on his behalf in March in Philadelphia Common Pleas Court, alleging $1.5 million in losses; the Inquirer reported he had wagered $18.5 million with FanDuel. In the video, which the Inquirer surfaced in July, Harper addresses Thompson by name, mentions his young son and says he is reaching out at the request of "your host Bryttanni at FanDuel." Harper's statement on July 13: "I did not know FanDuel would do this, I did not consent to it, and FanDuel had no right to do it." FanDuel's September 24 response to the lawmakers ran four pages and disclosed the figure of roughly 30 such videos without saying who received them. Blumenthal's reply, per Front Office Sports: "They owe us more specifics."

The case for the medians

Here is FanDuel's best argument, which is better than Friday's post makes it.

The Forrest and McHale figures are British, from 2018-19, and from a market with a different product mix, different regulation and no state-by-state licensing. Concentration in an American sportsbook in 2026 might be lower. The Fulford paper, which is American and recent, really does fail to find a population-level link between sports betting and worse financial outcomes, and a company accused of harming its customers is entitled to cite a peer-reviewed null result. And the political demand underneath the letter, that FanDuel stop treating its best customers differently from its worst, is a demand no retail business in any industry meets. Airlines, casinos and brokerages all have hosts.

Two answers. First, the British betting-only figure, 66.72 percent from the top 5 percent, is almost identical to the all-gambling figure of 66.94 percent, and the gaming figure is 70.82. The shape does not move much across products or operators, because it is a property of a low-hold business selling to a population whose stake sizes are heavy-tailed. If FanDuel's American curve is materially flatter, FanDuel can show it, and the showing would end the argument in its favour in a way no median can. Second, the airline comparison fails at the point that matters: the airline's host does not have, and is not legally required to run, a parallel system designed to detect that the customer is being harmed by the flights. FanDuel's is. New York's new at-risk patron rule lists eleven triggers for that system. The lawmakers' question is what happens when a customer trips one of them and has a host.

What would actually answer the letter

FanDuel holds the numbers that would settle this. The share of its net gaming revenue from its top 1 percent and top 5 percent of accounts. The number of accounts in the VIP programme and how many of them had tripped a responsible-gaming flag in the prior twelve months. Whether the 30 video recipients were drawn from a list that any player-protection system had already touched. Those are the data a regulator would ask for and that ProPublica's test of DraftKings approximated from the outside two weeks ago.

What FanDuel published instead was a distribution of the customers who are not the issue, with the concession, to its credit, that medians and averages "only tell part of the story." We agree. Casino.org estimates FanDuel's share of the American online market at roughly 44 percent, which means its answer on Tuesday is, functionally, the industry's answer. If it is the same answer as Friday's post, the next letter will not be from three Democrats. The FTC's review of sportsbook advertising, which we wrote about on October 13, is already open, and Louisiana attached monthly reporting conditions to DraftKings' licence last Thursday. We do not know what FanDuel will send. We know what the median will not tell them.