Cash-Out Prompts at 2 Losses Cut Chase Deposits 34% in 500 Slips
Across a sample of 500 withdrawal slips logged between 3 February and 28 March 2025, players who received a mandatory cash-out prompt after their second consecutive losing session went on to deposit 34% less over the following fortnight than a matched control group that received no prompt. The finding, drawn from a mid-sized operator's behavioural analytics export rather than a controlled trial, suggests that a single well-timed interruption can meaningfully dampen the deposit escalation that typically follows a losing run. It also raises a harder question: whether that reduction reflects healthier play or simply displaced activity onto other platforms.
What the intervention actually was
The prompt under study was deliberately modest. After a player closed a second losing session within 24 hours — defined as a session ending with a net loss and at least 15 minutes of active play — the client displayed a full-screen interstitial on next login. The interstitial stated the two-session loss in currency terms, offered a one-tap 24-hour cool-off, and required an explicit "continue playing" confirmation to proceed. There was no bonus, no free spins, no "come back tomorrow" incentive. Crucially, it fired on session outcomes, not on deposit amounts, so it did not discriminate between a player down €40 and one down €4,000.
The 500 slips were split 250/250 between prompted and unprompted players, matched on prior 30-day deposit volume, session frequency, account age, and game vertical (slots made up 61% of the sample, live casino 22%, sportsbook 17%). Matching was done on a coarse four-band deposit scale, which is the study's first soft spot: within-band variance was not controlled.
Why two losses and not a deposit threshold
Deposit-triggered interventions are the industry default because they are easy to wire into the payments stack. But they arrive after money has already moved, and players read them as friction at the worst moment. Session-outcome triggers land earlier in the behaviour chain — before the chase deposit, not after it. The two-loss threshold appears to be the operative detail: a single-loss prompt tested in an earlier internal run showed no statistically meaningful deposit difference, and players described it as nagging.
The numbers behind the 34%
The headline figure is a comparison of mean deposits in the 14 days following the trigger event. The prompted cohort averaged €118.40; the control cohort averaged €179.40. That is a 34.0% reduction, with a 95% confidence interval of roughly 27% to 41% — wide, because a 250-per-arm split of a heterogeneous population is not a lot of statistical power.
| Metric | Prompted (n=250) | Control (n=250) |
|---|---|---|
| Mean 14-day deposits | €118.40 | €179.40 |
| Median 14-day deposits | €60.00 | €95.00 |
| Share making zero deposits | 31.2% | 18.8% |
| Mean session count, 14 days | 6.1 | 7.9 |
| 30-day retention | 71.6% | 76.4% |
Two things stand out beyond the headline. First, the median gap (€35) is proportionally smaller than the mean gap, which tells you the effect is concentrated among higher-spending players — the top decile of the prompted group cut deposits by more than half. Second, retention fell 4.8 percentage points. Some of the deposit reduction is players who simply left.
The displacement problem
That retention figure is the crux. A 31.2% zero-deposit rate over two weeks does not distinguish between "stopped gambling" and "gambled elsewhere." The dataset contains no cross-operator visibility, so any claim that the prompt reduced total gambling expenditure is unsupported. What it demonstrably did is reduce expenditure with this operator. In a market with meaningful licence friction and shared self-exclusion registers, displacement is a smaller concern; in a fragmented grey market, it may be the whole story.
Where the effect is strongest — and weakest
Segmenting the 500 slips by vertical and by prior loss-chasing history produces a more useful picture than the aggregate.
- Slots players with no prior chase flag: 22% deposit reduction. The prompt functioned as a speed bump; most confirmed and carried on at a lower intensity.
- Slots players with a prior chase flag (two or more same-day re-deposits in the previous 60 days): 51% reduction, but also a 9-point retention drop. This is the group the intervention is designed for, and the group most likely to be lost entirely — which, depending on your regulatory posture, is either the point or a commercial problem.
- Live casino: 19% reduction, weakest of the three. Live dealers create longer continuous sessions that blur the "session end" definition, so the trigger may simply have fired less reliably.
- Sportsbook: 28% reduction, but highly sensitive to fixture calendar. A two-loss prompt during a midweek lull behaves nothing like one during a Champions League week, and the sample period included both.
The chase-flag segmentation is the most defensible result in the dataset. Prior chase behaviour is observable at account level, it is stable over 60 days, and the effect size in that subgroup is large enough to survive the sample's statistical noise.
What the 500 slips cannot tell us
Three limitations deserve more weight than the 34% headline.
Selection. Players were assigned to prompt or control by a rule-based system that was not random — the operator rolled the prompt out by account-ID range, and the matching was retrospective. If prompted accounts skewed toward a particular acquisition cohort, the deposit gap partly reflects who they were, not what they saw.
Novelty. The prompt was new to every player in the sample. Behavioural responses to interruptions decay; a player who has seen the same interstitial eleven times may confirm through it reflexively. The 34% figure is a first-exposure estimate and should not be projected forward as a steady-state effect.
The counterfactual is missing. There is no arm in which players received a different intervention — a deposit limit suggestion, a reality-check timer, a plain informational message. We know the prompt outperformed nothing. We do not know it outperformed cheaper alternatives, and a full-screen interstitial on every second losing session carries real UX cost: 6.8% of prompted players in a follow-up survey described it as "patronising," and 2.1% said it made them actively want to deposit out of irritation. That is a small number and a real one.
Responsible-gambling tooling has a long history of being evaluated on adoption rather than outcome. Prompts get measured by how many players click the cool-off link, which rewards intrusive design. The useful thing about this dataset is that it measures the downstream money, not the click — even if it does so imperfectly. Operators weighing whether to ship a two-loss trigger should treat 34% as an upper bound from a favourable sample, and should instrument the displacement question before quoting it in a licence application. The open question is not whether the prompt works, but whether a 4.8-point retention cost is a price the operator is willing to pay, or one regulators will eventually make non-optional.