A controlled trial of 1,842 slot players found that requiring a voluntary limit increase at spin 60 — rather than permitting raises at any point in a session — reduced average session stakes by 9.1% among participants who requested an increase, and by 4.3% across the full sample. The intervention did not reduce the number of players who eventually raised their limits; it delayed the raise and shortened the interval between raise events. The finding complicates the assumption that friction at the point of limit-setting is inherently protective.
The study, conducted over 14 weeks ending March 2025, drew on a panel of U.S. adults aged 21 to 64 who had deposited at least $200 in the preceding 90 days. Participants were randomized into three arms: a control group with no change to the limit-raise flow, a "cooling-off" arm requiring a 24-hour wait before a raise took effect, and the spin-60 arm, in which any raise request submitted before the 60th spin of a session was deferred until that threshold was crossed. The primary outcome was total amount wagered per session, measured in dollars rather than in spins or time-on-device.
The Mechanism: Delay, Not Deterrence
The 9% figure requires unpacking. It is not a reduction in the number of raises. In the spin-60 arm, 31.4% of sessions included at least one limit raise, against 33.1% in the control arm — a difference that fell outside the study's pre-registered significance threshold (p = 0.11). What changed was when the raise occurred and what happened in the intervening spins.
Median time from session start to first raise was 4 minutes 12 seconds in the control arm and 11 minutes 38 seconds in the spin-60 arm. That seven-minute gap is where the stake reduction lives. Players who raised early tended to raise again within the same session; the spin-60 cohort, by contrast, often exhausted the session's initial budget before the raise took effect and either stopped or reduced bet size to extend play at the lower limit.
The behavioral reading is straightforward: the intervention inserted a delay between impulse and capacity. It did not change the impulse. This distinction matters for regulators who have treated all limit-setting friction as equivalent.
Why the Effect Was Larger Among Raise-Requesters
The 9.1% figure applies only to the 578 participants who requested at least one raise during the study period. Within that subgroup, the effect was concentrated among players whose pre-raise average bet was between $0.50 and $2.00 — a band that accounted for 61% of raise requests but only 38% of total handle in the control arm. Higher-stakes players showed a 2.7% reduction, statistically indistinguishable from zero.
One interpretation: low-stakes players are more sensitive to the friction because their sessions are shorter and their budgets tighter. A seven-minute delay represents a larger fraction of a 25-minute session than of a 90-minute one. Another: high-stakes players had already internalized limit management and the raise request was less impulsive to begin with. The data cannot separate these.
The 24-Hour Arm Performed Worse Than Control
The cooling-off arm is the study's most uncomfortable result. Participants required to wait 24 hours before a raise took effect showed a 1.8% increase in average session stakes relative to control — not significant, but the wrong direction.
Two explanations have been offered. The first is anticipatory wagering: players who knew a raise was pending the next day wagered more aggressively in the current session, treating the existing limit as a deadline rather than a constraint. The second is selection: 22% of cooling-off participants who requested a raise abandoned the request before it took effect, and those who persisted were disproportionately high-volume players. The arm's composition shifted over the study period.
Neither explanation is satisfying. What the result does establish is that friction duration is not monotonically protective. A 24-hour delay is long enough to change behavior in ways a 60-spin delay is not, and not all of those changes point the same direction.
Comparison to Deposit-Limit Research
The finding runs against a body of deposit-limit literature suggesting that mandatory waiting periods reduce subsequent deposits. That research, however, largely concerns first deposit limits set at account opening, not intra-session raises. The population differs: new accounts versus established players mid-session. The spin-60 result should not be read as a refutation of deposit-limit cooling-off periods, but it does suggest the two interventions operate on different behavioral levers.
What the Numbers Do Not Show
The study measured wagered amounts, not losses. A 9.1% reduction in stakes is not a 9.1% reduction in harm. Players who wagered less per session may have simply played more sessions; the study tracked session-level data over 14 weeks but did not aggregate to monthly handle per participant, a limitation the authors acknowledge.
Nor does it address the substitution problem. If a player cannot raise a slot limit, they may move to a different vertical — table games, sports betting, or a second operator — where the limit architecture differs. The panel was restricted to a single operator's ecosystem, so cross-platform substitution is invisible.
There is also the question of novelty. The spin-60 threshold is arbitrary. It was chosen because 60 spins approximates the median session length in the operator's historical data, not because 60 has any behavioral significance. A threshold of 40 or 100 might produce a different effect, and nothing in the design tests sensitivity to that parameter.
The Regulatory Question
Several U.S. states have moved toward mandatory limit-setting prompts, and the federal SAFE Bet Act proposal — reintroduced in 2025 — includes provisions for session-level interventions. The spin-60 result suggests that where the prompt sits in the session matters more than whether it exists. A prompt at account opening, or a raise flow that permits immediate increases, produces measurably different outcomes than a mid-session gate.
The open question is whether a 9% stake reduction justifies the implementation cost and the player-experience friction. Nine percent is not nothing. It is also not the 30–40% reductions observed in some deposit-limit studies, and it comes with a subgroup pattern — no effect for high-stakes players — that maps awkwardly onto the populations regulators are most concerned about.
If the mechanism is delay rather than deterrence, then the policy lever is timing, and timing is cheap to adjust. The harder question is whether operators have any incentive to adjust it in the direction the evidence supports.