Deposit Cap Sliders Cut Average Top-Ups 26% in 500 Wallets

A 500-wallet pilot run across three licensed operators between January and March 2024 found that exposing depositors to a draggable daily cap slider reduced average top-up size by 26.4% relative to a matched control cohort, from €47.10 to €34.68 per deposit. The effect persisted after controlling for deposit frequency, which rose 8.1% over the same window, leaving net monthly staking broadly flat. The result complicates the assumption, common in both operator compliance teams and regulator consultation papers, that frictionless deposit interfaces and harm-reduction design are in direct tension.

The intervention and its measurement problem

The slider in question replaced a fixed default cap — typically €250 or €500 per day, set at registration and rarely revisited — with a persistent, editable control surfaced at the point of deposit. Users saw their current cap, a suggested lower band derived from their own 90-day deposit history, and a one-tap option to reduce it. Raising the cap required a 24-hour cooling period and a re-authentication step. Lowering it was instant and, notably, irreversible for seven days.

Measuring the causal effect is harder than it sounds. Deposit cap changes correlate with self-selection: players who lower caps are disproportionately those already moderating their play, so a naive before-and-after comparison overstates the intervention's bite. The pilot addressed this with a 1:1 matched control drawn from the same operators' pre-period populations, matched on deposit frequency, average deposit size, tenure, and game mix. The 26.4% figure is the difference-in-differences estimate; the raw treated-group decline was 31.2%, and the control group drifted down 4.8% on its own, plausibly reflecting post-holiday seasonality.

Why the slider, not the cap, moved the number

The interesting finding is not that lower caps reduce deposits — that is arithmetic. It is that voluntary, reversible cap reductions produced a larger behavioural shift than the fixed caps they replaced, even when the chosen ceiling was higher than the operator default. Roughly 41% of treated users set a cap above €250 at least once during the pilot, yet their average deposit still fell. The mechanism appears to be salience rather than constraint: the slider forced a monthly-ish confrontation with a number most players had never consciously chosen.

Where the 26% comes from — and where it doesn't

Disaggregating by deposit size band matters, because the headline figure hides a bimodal distribution.

The heavy tail drives the aggregate. A 41.9% reduction among the top 4% of wallets accounts for roughly half the total euro-volume decline, even though those wallets are a small minority. This is consistent with the harm-reduction literature's general finding that intensive interventions on high-intensity users yield the largest absolute effects, but it also means the 26.4% figure is fragile: exclude 20 wallets and it drops to about 17%.

Two caveats deserve emphasis. First, the pilot covered three operators in two jurisdictions, both with mature self-exclusion registers and mandatory affordability checks above €2,000 monthly. Results in markets without those backstops may differ. Second, the 90-day follow-up window is short. A 26% reduction in deposit size is not the same as a 26% reduction in losses, and it is certainly not the same as a 26% reduction in harm. Players who deposit less per transaction may simply deposit more often — the 8.1% frequency increase hints at exactly that, and the net staking figure, while flat, carries a confidence interval wide enough to accommodate a modest increase.

The frequency rebound

The frequency increase is the most policy-relevant number in the dataset and the one most likely to be misread. On its face, an 8.1% rise in deposit count against a 26.4% fall in deposit size implies gross deposits fell around 20%. But the operators in the pilot report deposits in euros, not in sessions, and a meaningful share of the frequency increase came from users splitting what would have been a single €100 deposit into two €50 deposits to stay under a self-imposed cap.

This is not necessarily irrational. Splitting deposits can function as a crude commitment device: each individual transaction is smaller, and the friction of a second deposit creates a decision point. But it also means the intervention shifted behaviour without much changing total exposure, which is precisely the outcome a purely volumetric evaluation would miss. Any regulator using average deposit size as a harm proxy should treat this pilot as a warning about that metric's sensitivity to interface design.

What the operators got out of it

Commercially, the picture is mixed. Gross gaming revenue per treated wallet fell 11.3% over the pilot, but 90-day retention rose 6.4 percentage points, and the share of treated users completing affordability verification — a costly manual process — rose from 12% to 29%. Two of the three operators have since made the slider permanent. The third reverted to fixed caps, citing the revenue delta and, according to its published note, "insufficient evidence that the retention gain offsets margin compression." That is a defensible reading of a 500-wallet sample.

What a larger trial would need to settle

The pilot raises more questions than 500 wallets can answer. Does the effect decay? Cap-setting behaviour in the first two weeks of any new interface tends to be unrepresentative, and a 12-month panel would show whether the 26.4% is a level shift or a transient. Does it survive jurisdictional variation? Both pilot jurisdictions mandate cooling periods on cap increases; in markets where a cap can be raised instantly, the slider's friction is weaker and the effect may shrink or invert. And does it hold for sports bettors, who were only 19% of the sample and whose deposit patterns are far more event-driven than casino players'?

There is also an unresolved design question. The pilot's slider presented a suggested lower band derived from the user's own history — a nudge, in the Thaler-Sunstein sense, toward the player's existing median rather than toward any externally set safe level. A regulator-mandated slider might anchor differently, perhaps to an affordability-derived figure, and the behavioural response to a prescriptive default is not obviously the same as to a descriptive one. The 26.4% figure belongs to the latter. Whether it transfers to the former is the question that should determine how quickly any of this gets written into licence conditions.