Developing A Reporter Lost $10,702 on DraftKings and Got a VIP Host. The Number That Should End the Argument Is 14 of 71

The number everyone will quote from ProPublica's DraftKings investigation is $10,702 — what reporter Jake Pearson lost, net, across ten weeks of deliberately betting like someone in trouble. It is the wrong number.

The right one is 14 of 71. Pearson gambled on 71 days. A responsible-gambling prompt appeared on fourteen of them. Over the same stretch the app sent him as many as six push notifications a day telling him what to bet next.

That ratio is the entire argument, and it is not an argument about intent. Nobody at DraftKings decided to hurt Jake Pearson. The product simply has one system running continuously and another running occasionally, and the occasional one is opt-in.

What the test actually did

Pearson built his betting pattern with a panel: addiction specialists, a recovering gambler, professional bettors. The behaviours he reproduced are the ones clinicians use as markers — chasing losses within a session, escalating stake size, switching sports late at night, leaning on parlays and microbets. Then he turned on the guardrails DraftKings provides. Deposit cap. Session cap.

After roughly three weeks and $5,800 in deposits he was invited into the formal VIP programme: a personal representative, deposit matches, bonus bets. The invitation arrived the day after he chased a loss of nearly $1,800 in a single night on basketball. By week eight he had deposited $21,600. At one point a $5,000 deposit was matched with $1,250 in DK Dollars.

The first warning he received came only after he had lost roughly a month of minimum-wage earnings — and it was a legally mandated New York disclosure triggered by crossing a $2,500 deposit threshold, not a judgment by anyone about how he was betting.

DraftKings' Chief Responsible Gaming Officer Lori Kalani told ProPublica: "I think as a business, as an industry, we're doing a good job of educating people, of raising awareness." She also said, of the episode, "The system worked the way it was supposed to."

Taken at face value, that second sentence is the most damaging thing in the piece. It is probably also true.

The part the industry will not say

Here is the strongest defence of DraftKings, stated properly, because it deserves to be.

Pearson is not a problem gambler. He is a salaried reporter executing a script with a budget and an exit date, and a system calibrated to catch genuine distress should arguably not fire on him — he never missed a payment, never requested a limit increase he could not fund, never contacted support in crisis. Behavioural detection that flags everyone who loses money flags most customers, which is not detection. DraftKings operates state by state under rulebooks that differ on what a book may do unprompted with a customer's account; a book that unilaterally closes accounts is also a book deciding, with no due process, that an adult may not spend their own money. And the 8 per cent figure cuts both ways: if only 8 per cent of Massachusetts DraftKings users opt into the tools, the honest reading includes "most customers do not want them."

All of that is fair. None of it touches 14 of 71.

The question ProPublica's test poses is not whether DraftKings should have closed the account. It is why a company with the product sophistication to deliver six personalised betting prompts a day — selected, timed and priced by systems that know exactly what this customer responds to — applies none of that sophistication in the other direction. The same machinery that identified Pearson as a high-value customer within three weeks had, by construction, identified every behaviour that made him one. The asymmetry is not a failure of detection. Detection worked. It was pointed one way.

DraftKings declined to say how often it proactively closes accounts for problem gambling. It employs 51 people on manual review. For context, that is a smaller team than most of its state-level marketing operations.

Tuesday's letter, Wednesday's threat

On October 6, Senator Richard Blumenthal joined Representatives Paul Tonko and Valerie Foushee in telling FanDuel that its September 24 response to their VIP inquiry was inadequate. That response ran four pages and disclosed that FanDuel had distributed roughly 30 personalised athlete and entertainer videos to customers over two years without saying which athletes, to whom, or on what basis. The underlying episode is a Bryce Harper Cameo video bought by FanDuel in November 2024 and sent to a VIP customer with a gambling addiction.

"They owe us more specifics," Blumenthal told Front Office Sports on Wednesday. "They owe the public more specifics." Then he widened it: "We will be asking questions of DraftKings as well as FanDuel. We'll be asking the questions about prediction markets."

That last clause is the one with teeth, and it has little to do with the prospects of any federal gambling bill, which remain poor in this Congress. Kalshi and Polymarket have spent eighteen months arguing to courts and to the CFTC that they are not sportsbooks. A Senate inquiry that treats them as a fourth and fifth name on a list of gambling platforms with VIP practices worth examining is a different kind of problem than a cease-and-desist: it is a factual record, built under oath or close to it, that the exchanges host the same behaviour. A judge reading that record next year is reading it in a preemption case.

What we would watch

Three things, in order of how much they would tell us.

Whether DraftKings publishes a proactive-closure number. It has never done so and ProPublica asked directly. A company confident in its safeguards releases that figure; the refusal is information.

Whether the next congressional letter asks about automation rather than about athletes. The Cameo story is vivid and small. The reason Pearson got six prompts a day is that targeting is machine-driven and continuous, and no state regulator currently requires a book to disclose how its promotional targeting treats a losing account differently from a winning one. That is the disclosure that would change behaviour, and nobody has asked for it yet.

And whether FanDuel's Illinois surcharge, switched back on this week, shows up in anyone's thinking. The same industry arguing it cannot reliably identify customers who should bet less has just demonstrated, in one state, that it can identify customers it would rather not have and price them out to the cent. The capability is not in question. Only the direction it points.