Per a New York Times investigation published September 19, 2026, DraftKings built a machine learning model in 2023 that scored online casino customers on how much they’d likely lose per promotion. Internally, the score was called elasticity.
Low scorers, tagged inelastic, got fewer offers. High scorers kept getting them. Similar sports betting models followed and are still in development, per the report.
A separate in-house model meant to assign problem gambling risk scores was shelved. A presentation to leadership was canceled the day of.
DraftKings says promos go to regular, active users, not losing ones, and rejects any suggestion its marketing is unfair.
FanDuel, Fanatics and PrizePicks have signed on to third-party risk scoring. DraftKings has not.
What the New York Times Found
Everything here comes from the Times investigation by Alex Klavens, Walt Bogdanich and Jenny Vrentas, built on research memos, presentations, Slack messages and betting records from experiments run on customers.
In 2023, DraftKings fed customer betting records into a machine learning model built to answer one question: who bets more, and loses more, once a promo lands? Testing it fell to data analyst Jayden Butts. He told the paper the financial logic pointed somewhere ugly: on a spreadsheet, the best return on a bonus comes from a problem gambler.
The model scored each customer on their habits. Higher score, bigger expected loss for each promotion offered. It weighed dozens of data points per gambler, including:
- How often they played
- What sat in their account day to day
- Typical losses measured against total bets
- A second model’s estimate of how likely they were to quit gambling
Six ex-employees who worked on these methods told the Times that DraftKings has kept sharpening them, using data science to aim bet-more promotions at losing gamblers.
Inside the Casino Test
DraftKings made its name on sports, but the Times reports online casino games account for close to a third of company revenue. That’s where, in mid-2023, the company set out to predict how promotions translate into a player’s wins or losses.
Trained on historical play, the model took in a customer’s recent activity each week and produced the DraftKings elasticity score. Below average and you were inelastic, in line for fewer offers. The elastic bettors stayed on the list.
In September 2023, Butts tested the score on promotions for about 5,000 casino players, then widened the test. He told the Times he assumed the goal was saving money. His supervisors corrected him, he said: the company wanted to redeploy promo spending, not shrink it. He took that to mean more offers for the biggest losers.
A 2023 company memo found slots revenue more elastic than other casino games, the Times reports, meaning online casino slots promotions drove play especially well.
One analyst, who the Times says quit in 2024 and requested anonymity, needed five words: “as predatory as it sounds.”
The Model DraftKings Reportedly Didn’t Build
The same data infrastructure that powered elasticity scoring could have flagged customers at risk of gambling harm, the Times reports. But a parallel project to create protective risk scores was shelved.
DraftKings told the Times it “rejects any implication” that its marketing unfairly targets customers and said promotions are directed to users who show sustained, engaged platform use rather than people ranked by predicted losses. The company also said data science and analytics improved promotion-driven sportsbook margins by 13% in 2025 and that AI helped personalize hundreds of millions of dollars in promotional spending.
DraftKings’s chief responsible gaming officer, Lori Kalani, told the Times the company monitors customers for risky behavior and that it declined to deploy a risk-prediction tool because it was not yet sufficiently evidence-based.