A mandatory deposit-limit prompt shown at the moment a player attempts a third same-session buy-in reduced completed rebuys by 22.4% among accounts capped at $50 per deposit, according to a twelve-month field study of 41,600 US-facing real-money accounts. The effect held across slots, blackjack, and crash-style titles, though it was weakest in live-dealer games. The finding matters less for the headline number than for what it implies: friction placed at a specific decision point outperforms friction placed at account opening, where most operators currently concentrate their responsible-gambling architecture.
The study, conducted between March 2023 and February 2024 across three licensed operators, used a staggered rollout design. Roughly half the sample encountered the prompt immediately; the other half served as a control for the first six months before receiving the same treatment. That structure lets us separate the prompt's effect from seasonal variance and from the general decline in session length that typically follows a new regulatory disclosure. Rebuy behavior, not deposits, was the primary endpoint — a distinction that matters because deposit limits can be gamed by players who simply reload more often at smaller amounts.
What the Prompt Actually Did
The intervention was deliberately narrow. When a player attempted a third buy-in within a rolling four-hour window, a modal appeared stating the account's remaining deposit allowance for the day and offering three options: proceed, set a lower cap, or end the session. No player was blocked. No default was changed. The prompt required a click to dismiss, which added roughly 2.8 seconds to the transaction.
That 2.8-second delay produced the 22.4% reduction. Among the $50-cap cohort, completed third buy-ins fell from a baseline of 1.31 per active session to 1.02. Fourth and fifth buy-ins fell harder — 31.7% and 38.2% respectively — which suggests the effect compounds rather than dissipates within a session. Players who dismissed the prompt and proceeded anyway showed no subsequent change in wager size, stake per spin, or session length. The prompt did not make engaged players more cautious; it appears to have intercepted players who were on the margin of stopping.
The $50 Threshold
Caps above $50 showed a weaker response. At the $100 cap, rebuy reduction was 11.8%; at $250, 6.3%; at unlimited, 2.1%, which is within the study's confidence interval and should be treated as noise. The authors attribute this to a straightforward mechanism: a $50 cap binds. A player who has already deposited $40 and wants to buy in for $25 faces an actual constraint, and the prompt makes that constraint salient at the exact moment it bites. A player with a $250 cap rarely encounters the ceiling mid-session, so the prompt reads as informational rather than restrictive.
This is the study's most useful practical finding, and it cuts against how many compliance departments think about limits. A limit that never binds is a disclosure. A limit that binds is a decision aid. Operators who set generous default caps and count them as player-protection measures are, under this data, mostly counting disclosures.
Where the Effect Broke Down
Live-dealer tables showed a 4.9% reduction, well below the slots cohort's 24.6%. The authors offer two explanations, neither fully satisfying. First, live-dealer sessions have longer natural intervals between buy-ins — a player at a $10 minimum table with a $200 stack may go forty minutes before needing to reload, so the four-hour rolling window catches fewer decision points. Second, and more speculative, live-dealer players in the sample skewed toward higher-income brackets, where a $50 cap is less likely to bind.
Poker showed a 17.3% reduction, but the sample is thin — 2,140 accounts — and the buy-in structure differs enough that I would not lean on the number. Sports betting was excluded entirely, which is a real limitation given how many US accounts now span verticals. A player who hits a deposit cap on slots may simply move the same dollars to a same-game parlay, and this study cannot see that substitution.
The Substitution Question
That gap deserves more attention than the headline figure. The study measured rebuys within the vertical where the prompt appeared. It did not measure cross-vertical migration, and it did not measure whether players who stopped a slots session at buy-in three simply returned the next day with a fresh limit. The 22.4% figure is a within-session, within-vertical number. Anyone citing it as a net reduction in gambling spend is overreading the data.
The authors are careful about this, noting that total monthly deposits across the treatment group fell 8.6% — real, but far smaller than 22.4%. The difference between those two numbers is the study's honest boundary. Friction at the third buy-in changes what happens in that session. It does not, on this evidence, change what happens that month by nearly as much.
Regulatory Context and the 2024 Timeline
The timing is not incidental. Several US states moved toward mandatory deposit-limit disclosure during the study window, and the American Gaming Association's updated responsible-gaming code, published in February 2024, explicitly recommends "just-in-time" interventions over static disclosures. The study's results arrived roughly six weeks after that code update, and at least two of the three participating operators have since moved the prompt to a default-on setting rather than an opt-in.
That shift raises a question the study cannot answer: whether the effect survives when the prompt is universal and expected. Novelty is a confound in any staggered rollout. A player who sees a modal for the first time in month one may respond differently than a player who has seen it at every third buy-in for a year. The authors acknowledge this and note that the reduction attenuated slightly — from 24.1% to 20.8% — across the treatment group's twelve months. Attenuation, not extinction, but the trend line is worth watching.
Design Details That Mattered
Two implementation choices appear to have driven much of the effect. The prompt displayed the remaining daily allowance in dollars, not as a percentage of the cap. And it required an active dismissal rather than timing out. A pilot variant that auto-dismissed after ten seconds produced a 9.4% reduction — less than half the main result. Passive prompts get ignored; the friction has to be real to work.
The second detail is that the prompt offered a "set a lower cap" button directly in the modal. Only 3.7% of players used it, but those who did reduced subsequent buy-ins by 61.2%. That is a small, self-selected group and the number should not be generalized, but it hints that the prompt's value may be less about stopping the current buy-in than about surfacing an option players did not know existed.
What Operators Do With This
The commercial read is uncomfortable for anyone selling engagement. A 22.4% cut in third buy-ins is a direct hit to session revenue, and the study's own total-deposit figure — down 8.6% — suggests the net cost is smaller than the rebuy number implies but not trivial. Operators weighing the prompt against their margins are effectively asking whether the intercepted players were profitable long-term. The study does not model lifetime value, and that omission is the largest gap in the analysis.
For regulators, the finding argues for specificity. "Encourage responsible play" is not an intervention. A dollar-denominated cap, shown at the third same-session buy-in, with a required click, is. The 22.4% figure will get quoted widely; the more durable takeaway is the mechanism — friction at the binding constraint, not at account opening.
The open question is whether this scales down. Most of the sample sat at the $50 cap because that was the modal default. If operators lower defaults to $20 or $25, does the prompt still find a decision point to interrupt, or does it simply push players into more frequent, smaller deposits that the four-hour window never catches? That is testable, and nobody has run it yet. Until someone does, the 22.4% figure describes one design at one cap level, and the industry should resist treating it as a general law of player protection.