Deferred Bonus Wagers Cut Chase Deposits 24% in 500 Accounts

A 500-account field study run between 3 February and 28 April 2025 found that deferring the release of bonus funds until wagering thresholds were met reduced average "chase" deposits — deposits made within 48 hours of a bonus balance hitting zero — by 24.1%, from €212 to €161 per account per month. The effect held across three operator verticals and survived controls for player tenure, deposit frequency, and bonus size. The design change itself was modest: instead of crediting bonus funds at opt-in, the participating operators held them in a pending ledger and released them in tranches as wagering progressed.

The finding matters less for the headline number than for what it implies about why players chase. If deferred release works, the mechanism is probably not persuasion but friction — and friction, unlike messaging, can be measured, priced, and audited.

Study design and the 500-account sample

The sample was drawn from three licensed operators sharing a common wallet infrastructure, which allowed the researchers to apply the same ledger logic across all accounts. Inclusion criteria were narrow: accounts had to be at least 90 days old, have opted into at least two deposit bonuses in the prior quarter, and show at least one historical chase deposit (defined as a deposit within 48 hours of bonus balance exhaustion, followed by wagering within the same session). That last criterion is what makes the sample unrepresentative of the general player base — and also what makes the result interpretable. These were not casual players; they were accounts with a demonstrated pattern.

Accounts were matched into treatment and control cohorts on five variables: 90-day deposit total, average session length, bonus opt-in rate, game mix (slots share vs. table share), and prior chase frequency. Matching reduced the effective analytic sample to 486 accounts after 14 dropped for inactivity or self-exclusion during the window. The 500 figure in the headline refers to the recruited cohort; the 486 is what the regressions actually used, and the difference is disclosed in the appendix.

What "deferred release" meant operationally

Treatment accounts received bonus funds in three tranches: 30% at opt-in, 40% at 50% wagering completion, and 30% at 100%. Control accounts received the full amount at opt-in, as before. Maximum bonus size, wagering requirement (35x on slots, 40x on table games), and game weighting were identical across cohorts. The only variable was timing.

This matters because a large share of the responsible-gambling literature conflates bonus structure with bonus size. Here, size was held constant. Players in the treatment arm were not offered less; they were offered the same amount on a different schedule. Any behavioural difference is therefore attributable to timing, not generosity.

The 24% figure, decomposed

The headline reduction of 24.1% in average chase deposit value is an aggregate. Broken down, the effect was uneven in instructive ways.

The slots-heavy segment drove most of the result. That is consistent with a friction interpretation: slots players churn through bonus funds faster and hit the zero-balance moment more abruptly, so a tranche structure interrupts the transition more often. Table players, whose bonus funds deplete more slowly and whose sessions are longer, experienced the intervention as less of a break in continuity.

The researchers also tracked a secondary measure: the rate of chase deposits, not just their value. Here the effect was smaller — a 14.7% reduction in the number of chase deposits, against a 24.1% reduction in value. In plain terms, deferred release made chase deposits somewhat less frequent and substantially smaller when they happened. That pattern is more consistent with a cooling-off effect than with deterrence.

Why timing beats messaging

A useful comparison point is the messaging literature. Operator-funded "responsible gambling" prompts — the pop-up that appears after a loss streak, the email after a large deposit — have been studied extensively, and the consensus is unflattering: prompts produce small, short-lived effects on session length and almost none on deposit behaviour. A 2023 meta-analysis of 41 prompt interventions put the median effect on deposit size at under 3%, with wide confidence intervals.

Deferred release is not a prompt. It is a structural change to the product, and it operates whether or not the player reads anything. That distinction is the most transferable finding in the study. Behavioural interventions that require the player to process information at the moment of highest arousal — precisely when processing is worst — tend to fail. Interventions that change the shape of the offer do not depend on that processing.

There is a commercial caveat the paper is careful about and that operators reading it should be too. Deferred release reduces chase deposits, but chase deposits are, from the operator's perspective, revenue. The study reports a 6.8% decline in total deposits per account in the treatment arm, not just a decline in the chase subset. The authors frame this as the cost of the intervention and note that it was not offset by higher retention — retention was flat. Whether a 24% reduction in a problematic behaviour is worth a 6.8% reduction in deposits is a policy question, not an empirical one, and the paper does not pretend otherwise.

The confound the authors flag

One limitation deserves prominence. Deferred release changes not only timing but also perceived bonus value. A tranche structure is harder to evaluate at opt-in; players may simply value the offer less and engage with it less. The study attempted to control for this by surveying a subsample on perceived value, but the survey was administered after the intervention window and response rates were low (n = 71). The authors concede that a "reduced perceived value" mechanism cannot be ruled out and would produce a similar deposit pattern. A follow-up with a same-size, same-timing-but-differently-framed offer would help separate the two.

Implications for bonus design and player protection

The practical takeaway is not that every operator should switch to tranche release tomorrow. The sample was selected for chase behaviour, the window was 12 weeks, and the effect could decay as players adapt to the new structure — a possibility the study cannot address because it did not run long enough. What the study does establish is that bonus timing is a live variable with measurable behavioural consequences, and that it has been under-studied relative to bonus size and wagering requirements.

For regulators, the result offers something rarer than a new restriction: a lever that appears to work without banning anything. Deferred release is compatible with competitive bonus markets, does not require limiting stakes, and can be implemented at the wallet layer. That is a more attractive policy instrument than a cap, and it is testable at scale.

The open question is durability. If deferred release works because it interrupts a moment of high arousal, players who learn the structure may simply plan around it — depositing in anticipation of the tranche boundary rather than after exhaustion. The 12-week window is too short to see that adaptation, and no study in this literature has run long enough to answer it. Until one does, the 24% figure should be read as a real but provisional effect, measured in a population chosen precisely because it was already exhibiting the behaviour the intervention targets. Responsible-gambling tools that work on the already-affected are valuable; tools that work before the pattern sets in would be more valuable still, and that is the experiment nobody has run.