DraftKings AI Loss Ranking
DraftKings Used Machine Learning Model to Rank Casino Customers by Likely Financial Losses
Anonymous
Key Points
1
DraftKings reportedly collected about $8.7 billion in gross gambling revenue in the prior year and distributed roughly $3 billion in promotions, underscoring the financial scale of its incentive program.
2
Former DraftKings data analyst Jayden Butts said, “The best investment would be a problem gambler,” while another former analyst described the approach as “as predatory as it sounds.”
3
DraftKings reportedly reviewed a separate responsible-gambling model developed by data scientist Nestor Hernandez in 2024 to flag customers moving toward a crisis, but executives reached a “collective decision” not to pursue predictive tools because they were considered insufficiently “evidence-based.”
4
The investigation comes amid worsening indicators of gambling-related harm: the National Council on Problem Gambling says more than 31,000 Americans contact its helpline each month, while a 2026 survey found that 60% of online sports bettors reported chasing losses and 26% said gambling losses were causing them financial problems.
5
DraftKings disputed the characterization of the model’s use, saying its promotions target customers with “sustained, engaged use” rather than customers selected because of their losses, and calling Butts’s test “preliminary and inconclusive.”
DraftKings built an artificial intelligence system in 2023 to identify and target customers most likely to lose money after receiving promotional offers, The New York Times reported. The company used a machine-learning model called an "elasticity" score to rank online casino players by how much they would spend when given free bets, directing roughly $3 billion in annual promotions based on these predictions.
Former employees said the system could favor people showing signs of problem gambling. DraftKings shelved a separate tool designed to identify customers at risk of addiction, deeming it insufficiently evidence-based. The findings raise concerns about using AI to encourage betting without comparable safeguards for vulnerable players.
How DraftKings Ranked Customers by Predicted Losses
DraftKings analyzed playing frequency, account balances, and losses relative to wagers to score each customer, The New York Times reported. The elasticity model measured how responsive players were to incentives — essentially predicting how much additional money they would spend if offered promotions. Higher-scoring customers received more generous free bets and bonuses.
The company collected roughly $8.7 billion in gross gambling revenue the prior year, with promotions totaling around $3 billion. Former data analyst Jayden Butts told NY Post that "the best investment would be a problem gambler," suggesting the system deliberately targeted financially vulnerable players.
Abandoned Plans to Identify At-Risk Gamblers
In 2024, data scientist Nestor Hernandez developed a separate predictive model to flag customers moving toward a gambling crisis. DraftKings executives reviewed the responsible-gambling tool but made a "collective decision" not to pursue it, citing insufficient evidence. Audacy reported that the company rejected a tool specifically designed to protect vulnerable players.
The rejected model stood in stark contrast to the elasticity system, which remained in use. While DraftKings invested resources in targeting high-loss customers, it declined to deploy technology for identifying and helping those at risk of addiction.
Rising Gambling Harm and Addiction Indicators
Problem gambling is accelerating across the United States. Psychology Today reported that the National Council on Problem Gambling now receives more than 31,000 calls monthly from Americans struggling with gambling losses. A 2026 survey found that 60% of online sports bettors admitted to chasing losses, while 26% said gambling was causing them serious financial problems.
These rising harms coincide with the expansion of AI-powered promotional targeting. Critics argue that companies like DraftKings deploy sophisticated tools to maximize spending from the most vulnerable customers, even as industry safeguards lag behind.
DraftKings Defends Its Promotion Model
DraftKings disputed NY Post's characterization of the elasticity model, saying promotions target customers with "sustained, engaged use" rather than those selected for losses. The company called Butts's analysis "preliminary and inconclusive" and defended the system as a standard marketing practice.
The company has not explained why it rejected the responsible-gambling tool or acknowledged using elasticity scores to identify financially vulnerable players. Meanwhile, attorney Jennifer Hoekstra has filed lawsuits against both DraftKings and FanDuel, alleging that gambling apps are designed to exploit and addict users.
Sentiment
Mostly NegativeNEGATIVENEUTRALPOSITIVE
Publishers
20
Articles
9
Reach
29
Topics
A
AnonymousAuthor
