Michigan Mailed 33 Letters Arguing That Only a Licensed Book Will Honour a Self-Exclusion List. It Is the States' Best Argument and Their Most Fragile One

Editor's note: TrueEdge builds odds tools and earns affiliate commissions from licensed sportsbooks, including two named in this piece. Weigh it accordingly.

The Michigan Gaming Control Board sent 33 cease-and-desist letters on Thursday. The recipients are a parade of names that sound invented — Azartzona Casino, Sloto Tribe, Magicianbet, Genie Jackpot, L0tt0, Lucky Charm Sweep — and most of them will ignore the letter, and Michigan knows it.

The reason to read the release anyway is the argument the board chose to make. Executive Director Henry Williams did not lead on tax leakage or on protecting licensees from unfair competition. He led on exclusion: "These 33 letters put operators on notice that our residents are not an untapped market, and that anyone who thinks they can ignore our protections will hear from us." The protections the board enumerated are specific — self-exclusion tools, deposit limits, independently tested games, data safeguards, a mechanism for recovering winnings the house refuses to pay. And one line does the real work: a Michigan resident who has put themselves on the state's self-exclusion list can still gamble on all 33 of those sites, because an unlicensed operator never checks the list.

That is the same argument thirty-nine attorneys general made to the Supreme Court two days ago, in the brief supporting New Jersey against Kalshi. States, they wrote, "prevent minors from betting, create exclusion lists to assist gambling addicts, apply consumer protections to prevent unfair practices and protect the integrity of sports, limit gambling spending, and more." Strip the states of jurisdiction and you do not get a different regulator. You get no regulator holding the list.

It is the best argument in the case, because it is the only one that describes something a reader can picture. And this week it got tested.

The ten-week experiment

On September 26, ProPublica published the results of an experiment by reporter Jake Pearson, who spent ten weeks deliberately betting the way a person in trouble bets — chasing losses, escalating stakes, playing late — on DraftKings and FanDuel, to see what the safeguards actually did.

DraftKings invited him into its VIP programme. The invitation arrived a day after he chased roughly $1,800 in losses in a single night. Over the full run he deposited $21,600 and finished down $10,702 net. The pattern he was simulating is the pattern the screening is supposed to catch, and the system read it as a sales signal.

FanDuel came out of the comparison better, which is worth saying plainly because the industry's critics usually do not bother. It extended no VIP invitation under comparable conditions, and during one extended session it put six "Reality Check" prompts in front of him reporting his time in the app and the amount wagered. Six interruptions in one sitting is a real intervention, not a compliance fig leaf.

FanDuel has its own problem in the same file. On September 24 it sent Congress a second letter defending its VIP programme, with Cory Fox, its senior vice president for public policy and sustainability, answering eight questions — including about a personalised Bryce Harper video sent to a bettor reported to have gambling problems. Fox put the number of personalised athlete and entertainer videos sent to VIP customers at roughly 30, and denied the company uses perks to stop customers from closing accounts. A lawsuit has been filed on behalf of the bettor.

ProPublica is now trying to replace the anecdote with a sample. Its uploader asks bettors to submit their own transaction histories, by browser extension, direct data request or a Pikkit export. As of Friday the page showed 57 submissions against a target of 250. If they hit it, this becomes the first outside dataset on how these apps behave as a customer deteriorates, and every regulator quoting self-exclusion at a courthouse will have to read it.

Two tracks

Set the two enforcement stories side by side and the shape of the year is visible.

Michigan's 33 letters went to offshore casinos. Those work, more or less. An operator with no US banking relationships and no licence to protect either geoblocks the state or accepts that the attorney general is next; the board sent 45 such letters across four months ending in April and the sky did not fall. This is ordinary regulatory housekeeping, and it is effective precisely because the targets have nothing to lose by leaving.

Ohio's letters went somewhere else. On October 2 the Casino Control Commission, under chair Thomas Stickrath, sent ten cease-and-desist notices to Coinbase, Gemini Titan, Moomoo Financial, Novig, Plus500US, Polymarket, Prophet X, Robinhood Derivatives, Underdog Predict and Webull Financial, giving them until October 16 to confirm in writing that they have stopped offering sports event contracts in Ohio. Several of those companies hold federal registrations and the published view of their federal regulator is that Ohio has no authority here at all. Missouri's six letters in September produced one visible result: PrizePicks, which was not even among the recipients, pulled its sports contracts out of the state on its own.

Same instrument, two entirely different outcomes, and the variable is a federal licence. A sweepstakes casino in Curaçao folds on receipt. A CFTC-registered designated contract market files a lawsuit, and increasingly the United States files one alongside it.

The counterargument, which is good

The obvious reply to all of this is that we are grading on the wrong curve. A flawed regulator beats no regulator. Michigan's self-exclusion list is imperfectly enforced and DraftKings' VIP screening failed a reporter who was trying to make it fail — but the list exists, the reporter could point at the failure, a congressional committee could write a letter, and a lawsuit could be filed. None of that machinery exists for Sloto Tribe, and none of it would exist for a sports market the states cannot touch. The argument from imperfection is an argument for better enforcement, not for no jurisdiction.

That is right, and we will not pretend otherwise. The existence of a forum in which a failure can be named is not a small thing; it is most of what regulation is.

But it does not fully survive the specifics, for two reasons. The first is that Pearson's test did not find a gap in coverage. It found a licensed, state-regulated operator in a mature market identifying an escalating-loss pattern and responding by routing the customer toward higher-touch marketing. That is not a protection that failed to apply. That is a commercial incentive running in the opposite direction from the protection, inside the system the states are asking the Court to preserve.

The second is that the states are making an empirical claim in a legal brief, and empirical claims can be checked. "We have the particular expertise to address the risks associated with gambling" is a sentence that invites someone to measure it. The CFTC conceded the point in 2024 and has since reversed; if the states want that concession to carry weight in 2026, the evidence has to be better than a list of statutory powers.

What we would watch

Three dates. October 16, when Ohio's deadline lands and we find out whether any federally registered operator blinks — Underdog and Novig are the plausible candidates, because their businesses are not purely exchange-shaped. Whenever ProPublica clears 250 submissions, because the industry's standing defence is that the reporter's behaviour was unrepresentative, and a few hundred real transaction histories will either support that or demolish it. And November 9, when Kalshi answers New Jersey.

The thing that would change our reading: if a state produced audited numbers showing that its exclusion list actually blocks a meaningful share of attempted sessions by enrolled problem gamblers at licensed books, the "only we can protect them" argument stops being a brief-writing flourish and becomes evidence. We have not seen a state publish that. Michigan did not publish it on Thursday, and it was the natural day to.