Share 
State of Play’s TL;DR
- A report summarizing a New York Times investigation says DraftKings used a machine-learning model to score users on “elasticity,” in an effort to target players prone to losing.
- It has raised responsible gambling questions.
A report summarizing a New York Times investigation says DraftKings built a machine-learning model to identify which customers were most likely to generate more revenue after receiving free bets and other promotions.
According to the report, the model scored users on “elasticity,” or how responsive they were to incentives. The New York Times said its reporting was based on interviews with more than 40 former DraftKings employees and internal documents.
Former employee Jayden Butts told the Times his task was to help answer a basic question: whether a user would give DraftKings more than the company gave that customer. If the answer was yes, Butts said, the company could “open the floodgates” with promotions.
Report focuses on profitability-based promotion targeting
The Times reported DraftKings built the elasticity model in 2023. It said the system measured how strongly users responded to incentives such as free bets and other offers. The report also said slots players were found to be more elastic, and that early 2024 data showed highly elastic slots players spent more money than less elastic ones.
DraftKings disputed the implication that its marketing unfairly targets customers.
In a statement cited by the Times, the company said it “rejects any implication that its marketing practices are unfair or improperly targets customers.” DraftKings also said promotions are aimed at users who show sustained engagement on the platform, not at customers based on their losses.
Separate crisis-prediction efforts were reportedly shelved
The same report said former employees also tried to build a model that could predict when a user was heading toward a crisis and might need intervention. Former employee Jake Shanin said an internal crisis-prediction model was showing promise, according to the Times.
But a planned presentation on that work in early 2025 was canceled, the report said. Two other efforts to build similar algorithms were also shelved, according to two former employees cited by the Times.
DraftKings Chief Responsible Gaming Officer Lori Kalani told the Times the company needs customers betting within their means and said DraftKings monitors for potentially risky behaviors.
Kalani also said evidence showed that risk-modeling technology was not helpful.
For players, the report adds to broader questions around how operators use customer data, promotions, and responsible gambling tools. It does not allege a new regulatory action, but it puts fresh attention on how betting companies balance retention strategies with player protection.
Based on reporting by Mike Pearl for AOL.
