THE SETUP. ProPublica's Jake Pearson had never placed a sports bet. His editors gave him $22,000 and a mandate: bet like a problem gambler, on purpose, and see what the sportsbooks do.
He didn't improvise the behavior. He assembled a panel — professional bettors, recovering addicts, addiction specialists, academics — and asked them to define the specific patterns that flag a gambler in trouble: escalating stake size, increasing frequency, late-night betting, chasing losses, parlays, and wagering on events you don't otherwise watch. Then he executed the list.
He did it on FanDuel and DraftKings, the duopoly that controls most of legal American sports betting. It took one night to get into the VIP program.
WEEK ONE. April, NBA playoffs, Knicks. Pearson opened with $20 to $50 bets specifically to train the algorithm — "to show the algorithm that we're sort of dumb and inexperienced." A $50 bet on a Yankees win, $20 on Aaron Judge to homer, $200 on the Knicks to win the conference.
Then Game 3 in Atlanta. The panel's advice: ratchet up to $100, add in-game wagers, increase the number of bets. The Knicks lost. He deposited $750 and put all of it on the Cavaliers to make it back the same night.
"I'm gonna put $750 more in my account right now and just put it all on the Cavs and try and make up for the losses that I just incurred on the Knicks. And then go to bed. This is ridiculous."
He lost about $1,800 that night. Five consecutive losses, each one chased.
THE EMAIL. The next day he was at a conference at Northeastern — a room full of researchers and advocates discussing the public-health damage from online betting. He was, at that exact moment, giving a speech about his descent into addiction.
Mid-conference, his phone buzzed. DraftKings. Congratulations, you've been accepted into the VIP program.
"It's like I literally couldn't script it better than this."
One night. One losing streak. A rep with a name and a phone number, assigned to a man who had been betting for eight days.
WHAT A VIP IS. Millions of people use these apps; a tiny fraction get in. The reporting calls them what they are: an elite crowd of lucrative losers the company wants to keep. The program is framed as loyalty rewards. A recent lawsuit alleges it rewards precisely the customers least able to control their activity.
There's a precedent, and it's the one American regulators keep ignoring. In the U.K., investigators found that something like 2 percent of depositors accounted for more than 80 percent of company revenue — numbers so indefensible that the U.K. rewrote its rules for VIP programs. The U.S. hasn't.
THE CONTROL GROUP. Pearson placed the same bets at the same odds on the same days on FanDuel. No VIP invitation. He pivoted the experiment to DraftKings, which is itself a finding: the two products behaved identically, and only one of them decided the pattern was worth monetizing.
THE HONEST PARTS. The responsible-gaming tools actually work, and the reporting says so. When Pearson hit the wall, he set deposit limits and time limits, and he could not circumvent them. When he took a cool-off — the most restrictive option short of self-exclusion, minimum three days — the app locked him out.
Then it kept pushing. Even during his limits, the push alerts and boost offers kept arriving. And when the cool-off ended, something changed: the notifications stopped, the promos got less juicy. "Brake-pumping."
For about a week. Then his VIP rep, who had ignored a 1 a.m. text asking for a deposit match, delivered the largest one-time bonus he'd been offered the entire experiment — days after the cool-off that was supposed to signal he wanted to stop.
THE NUMBER THAT ENDS THE ARGUMENT. Over the whole run, Pearson lost $24,000 in three days at his coldest stretch. And his account received roughly 32 messages about responsible gaming.
When he put that to DraftKings — was I ever flagged? — the answer was that the system "worked the way it was supposed to." The company said it conducted 92,000 manual reviews of about 5 million accounts last year, that it closes accounts "all the time," and that a customer who bets $40 repeatedly and writes in saying he has a problem gets closed permanently.
What it wouldn't say: how many of those 92,000 reviews ended with an account closed for problem gambling.
"Would the company close a user's account absent them saying, I have a problem?" "Yes, we do it all the time." "Like how often?" "We do it all the time."
THE HUMAN COMPARISON. The experiment is almost too neat, so the reporting leans on people it happened to for real. A man who put $300,000 to $400,000 through DraftKings between March and December 2024 on a $70,000–$80,000 salary. Net loss: about $110,000.
"I was at the Boston Garden. I got to play basketball on the court with DraftKings, which was amazing. But meanwhile, my life's falling apart. I get diamond-tier status and I'm homeless. I needed help, like, desperately. I wanted to die."
And another, on the thing the industry insists is just marketing: "I didn't start taking money that wasn't mine, and doing all these things that were so against my character and my morals and values, until I had a host. And I was the one who did it. I understand. But that got me there a whole hell of a lot faster."
THE NEXT PRODUCT. Sports betting is the entry point, not the business. The real money is iGaming — casino games on the phone, live in about half a dozen states, with an icon sitting right on the home screen of the same app. An industry figure puts it plainly: iGaming is in only 11 percent of the U.S. population, so "there's still a ton of room just by state expansion alone."
Pearson tried it. "I am betting so much more. Double, double, double, double. Lose, lose, lose, lose, lose. You can just tap your phone as fast as you can and spend as much money as you can." That's the slot machine the panel warned about, wearing a sportsbook's logo.
THE ENDGAME. Eight years after the Supreme Court struck down the federal ban, Americans have wagered roughly $600 billion legally, diagnoses are up 61 percent since 2018, and the states that legalized are now dependent on the tax revenue — which is a conflict of interest disguised as a budget line.
And the next fight is already here. Prediction markets — Kalshi and its competitors — are live in all 50 states, including states where sports betting is illegal, on the theory that they're trading rather than betting, and therefore subject to a lighter regulatory regime. New York is suing Kalshi. The company's position is that its rules are superior. Multiple states say they'll see it in court, and it's headed to the Supreme Court.
SPLASH'S READ: The finding isn't that DraftKings ignored the red flags. It's that the red flags are the recruitment criteria.
Everything Pearson did was, by the industry's own public standard, a warning sign: chasing, escalating, betting late, parlaying, betting sports he doesn't watch. Those are the exact behaviors the panel defined — and the exact behaviors that got him a human host within a night. A VIP program is not a bug in the responsible-gaming system. It's the monetization layer sitting on top of it, and it treats a customer's loss of control as a proxy for lifetime value.
The 32 responsible-gaming messages are the perfect illustration of how this is designed. The company can point to all 32 and say it did its job. None of them required a human to pick up a phone and ask whether the guy depositing $750 at midnight after a losing parlay was okay. The messages satisfy the legal obligation; the VIP host captures the revenue. Both are automated, and only one of them calls you.
The U.K. already ran this experiment and found the number that should end the debate: 2 percent of depositors, 80 percent of revenue. If that ratio holds here — and there's no reason to think Americans are built differently — then the industry's growth model is not "more customers." It's more from the customers who can't stop, and every state that legalizes iGaming without a hard cap on VIP programs is choosing tax revenue over the people generating it.
One more thing, and it's the part that should scare regulators most: the reporter had $22,000 of someone else's money, a panel of experts, and a plan. He still ended up staring at a chart of his own losses saying "it's jarring." Nobody without those advantages has a chance.