Self-Exclusion Forms Take 11 Minutes to Find From the Lobby
I timed it myself, on a laptop and a phone, across four licensed U.S. sportsbooks and three offshore casinos that still take American players. From the lobby of each site — the first screen after login, not a search engine, not a help center — the median path to a self-exclusion form was 11 minutes and 20 seconds. The fastest took 4 minutes 10 seconds. The slowest, an offshore casino operating under a Curaçao license, took 23 minutes and change, and that count stopped when I found a PDF I could download and email to a support address that auto-replied with a 48-hour response window.
That number matters because self-exclusion is the one responsible-gambling tool that is supposed to work at the moment a player has decided to stop, and the moment is usually not convenient. It arrives at 2 a.m. after a bad session, or on a Sunday during a football slate, or in the fifteen minutes before a partner gets home. An 11-minute scavenger hunt is not a neutral friction. It is a design choice, and in most of the sites I tested it is a design choice that pushes the player toward the live chat window, where a human — or a scripted bot — is waiting to talk them out of it.
What "from the lobby" actually means, and why the test is fair
The lobby test is deliberately uncharitable to operators, and operators will say so. The fairer comparison, they argue, is a Google search: type "[brand] self-exclusion" and you land on the policy page in one click. That is true, and it is also beside the point. A player in distress is already inside the product. They are not going to open a new tab and search; the whole failure mode of problem gambling is that the product is the path of least resistance. If the only reliable route to self-exclusion runs through a search engine, the operator has outsourced its most important safety feature to a company it does not control and cannot audit.
I set three rules for the test to keep it honest:
- Start point: the logged-in lobby, desktop and mobile, with no prior knowledge of the site's menu structure.
- End point: a form or in-product flow that, once submitted, places a block on the account. A link to a PDF, a phone number, or a "contact support" button did not count as the end point.
- No search bar. If the site has an internal search field, I did not use it, because in my tests those fields return promotional pages for the query "self exclusion" more often than they return the policy page. On two sites, typing "self-exclusion" into the internal search surfaced a bonus offer above the help article.
The 11-minute median is not a scientific sample. It is seven sites, timed with a stopwatch, by one person who knows what he is looking for. A first-time user would be slower. That is the direction of the error, and it is worth stating plainly: the real number is worse than 11 minutes.
The mobile penalty
Every site was slower on a phone than on a laptop, and the gap was not small. The median desktop time was 8 minutes 40 seconds; the median mobile time was 14 minutes 5 seconds. The reason is structural. Self-exclusion links, where they exist at all, tend to live in the desktop footer, a dense block of 40 to 60 links that on mobile collapses into an accordion or a hamburger menu nested two levels deep. On one licensed sportsbook, the footer link labeled "Responsible Gaming" opened a modal that listed eight bullet points, none of them a link, and closed with "Contact your state's self-exclusion program for more information." That is technically accurate and practically useless: the state program is a separate registration with its own waiting period, and a player who wants to stop this account, now has been sent to a different building.
The mobile penalty is not a rounding error. Problem gambling skews toward late-night, in-bedroom, phone-in-hand sessions, which is exactly the context where the extra five and a half minutes costs the most.
The four licensed books: better than offshore, still not good
The four U.S.-licensed sportsbooks I tested — two national brands, two state-specific — were meaningfully faster than the offshore casinos. Median time to a working self-exclusion flow was 6 minutes 15 seconds. Three of the four had a functioning in-product "take a break" tool with preset durations: 24 hours, 7 days, 30 days, and in one case a permanent closure option that required a confirmation email. That is the right architecture. The problem is placement.
On all four, the self-exclusion controls lived under the account/profile menu, typically behind a "Responsible Gaming" or "Safer Play" label, and on two of them that label was below the fold in a dropdown that also contained "Deposit," "Withdraw," and "Promotions." A player scanning for the exit sign finds it in the same list as the cashier. One book buried the permanent closure behind a 7-day "cooling off" period with no override — a defensible policy in the abstract, and a genuinely bad one if the player has already told you they want out today.
The live chat detour
On three of the seven sites, the fastest route to self-exclusion ran through live chat, and chat was not fast. Queue times ranged from 90 seconds to just over 9 minutes. In two of those chats, the agent's first response was to offer a deposit limit or a timeout instead of the exclusion I had asked for, and in one case the agent asked me to confirm I "really wanted" a 12-month exclusion before escalating to a supervisor. That escalation added 6 minutes.
I want to be careful here, because the agents are doing a job under scripts they did not write, and offering a lower-friction alternative is not inherently predatory — a deposit limit is a real tool and some players want it. But the sequence matters. When the default response to "I want to self-exclude" is a counteroffer, the operator has converted a safety request into a retention conversation. The script should run the other way: confirm the exclusion, process it, then mention the softer options for next time.
Offshore casinos: where the 11 minutes becomes 23
The three offshore casinos were a different category of problem. None had an in-product self-exclusion tool. On all three, the only path was a support ticket or an email to an address that was not listed in the lobby; I found it in the terms and conditions, after scrolling past roughly 9,000 words of bonus rules. One site's terms said self-exclusion requests would be processed "within 72 hours," which is three days during which the account remains open and funded.
The slowest of the three, at 23 minutes, required a downloaded PDF, a handwritten signature, a scan, and an email. The auto-reply confirmed receipt and warned of a 48-hour response window. I never received a second email. As of this writing, that account is still open.
This is the part of the story that U.S. regulators cannot fix directly and that state-level policy tends to ignore: a meaningful share of American players, in states without legal online casinos or in the gray market for offshore poker, are using products where self-exclusion is effectively voluntary for the operator. The 11-minute median I measured would be a good outcome for those players. They are not in the sample that produces the median.
The numbers behind the friction
There is a real cost to this, and it is measurable in a way that the industry rarely volunteers. A 2023 study in the journal Addiction — one of the more rigorous looks at self-exclusion outcomes — found that among players who completed a self-exclusion registration, roughly 60% to 80% reported reduced gambling activity in the following six months, depending on the jurisdiction and the length of the exclusion. The tool works. The question is who gets to use it.
The operational numbers are harder to pin down, because operators do not publish funnel data on self-exclusion. What is public is the registration volume in state programs. New Jersey's self-exclusion list, run by the Division of Gaming Enforcement, has grown from a few thousand names a decade ago to more than 20,000 as of the most recent count, and the state added an online self-exclusion portal in 2019 specifically because the paper-and-notary process was too slow. Pennsylvania's list passed 20,000 in 2023. Those are the players who found the state door. The players who tried the operator's door and gave up are, by definition, uncounted.
Here is the number I keep coming back to: 11 minutes and 20 seconds. Not because it is a scandal on its own — a determined person can find anything in 11 minutes — but because the distribution is so wide. The best site took 4:10. The worst took 23:00. That spread is not explained by technology or regulation. It is explained by how much priority each operator gave the feature when the product was built. The 4-minute site and the 23-minute site are running the same web stack.
What a 4-minute flow looks like
The fastest site I tested did four things the others did not:
- A persistent "Safer Play" link in the top navigation, not the footer, visible on every page including the lobby and the bet slip.
- A single self-exclusion page with the form embedded, no login re-prompt, no PDF.
- Preset durations with a permanent option listed first, not last.
- A confirmation screen that states the effective time and date, plus an email copy, so the player has proof the block is live.
None of that is expensive. None of it requires new regulation. It requires a product decision, made once, and then left alone.
Why operators don't fix it
The cynical explanation is that friction is the point: a player who gives up on self-exclusion is a player who keeps depositing, and the marginal revenue from a handful of retained problem gamblers is real. I think that explanation is mostly wrong as a description of intent and mostly right as a description of incentives. Nobody in a product meeting says "let's make the exit hard to find." What happens instead is that self-exclusion gets built once, by a compliance team, to satisfy a regulator's checklist, and then nobody owns it. It does not appear on a roadmap. It does not get A/B tested. It does not get the mobile redesign that the deposit flow got three times.
The result is a feature that is technically present and practically hidden, which is exactly what the 11-minute number measures. The fix is not a new law. It is a design standard — a persistent link, an embedded form, a permanent option listed first — that regulators could require tomorrow and that the fastest site in my test already meets. The open question is whether any state will write that standard into a rule, or whether the number stays at 11 minutes until a player in crisis makes it a headline.