Slot Session Length Drops 16% When Autoplay Caps at 25 Spins
A cap of 25 autoplay spins cuts average slot session length by roughly 16% compared with sessions where autoplay runs to 100 spins or more, according to operator-side data shared by three European-facing platforms and one U.S. social-casino operator over a rolling 90-day window ending in March. The drop isn't uniform: it concentrates in the first 10 minutes of play, and it barely shows up among players who set a loss limit before they spin. The finding matters because autoplay has become the quiet lever in the U.S. responsible-gambling debate, and the number 25 is now doing real regulatory work.
Where the 25-spin number comes from
Autoplay caps aren't new. The UK Gambling Commission pushed operators toward limits on auto-spin features in 2021, and several European jurisdictions settled on 25 as a ceiling — enough to keep the convenience of unattended spins, short enough to force periodic re-engagement. When the American Gaming Association's responsible-marketing code was updated in 2023, it recommended caps but didn't name a number. Individual operators filled the gap. FanDuel's social-casino arm and several tribal-adjacent real-money apps landed on 25. DraftKings' social product went to 50. A handful of smaller platforms left autoplay uncapped entirely until late 2024.
The data behind the 16% figure comes from a comparison operators almost never publish: matched cohorts of players on the same game, same average bet, same account age, split only by whether their autoplay was capped at 25 or allowed to run to 100-plus. Sessions were defined as continuous play until a 15-minute inactivity gap or explicit logout. Average uncapped session: 41 minutes, 18 seconds. Average capped session: 34 minutes, 41 seconds. That's a 16.1% reduction, which rounds to the headline number.
The sample is imperfect. It's four operators, not forty. It skews toward slots with medium volatility (games rated 3 out of 5 by the vendors' own volatility scales), and it excludes live-dealer and table games. Two of the four operators adjusted their autoplay defaults mid-window, which means part of the effect may reflect novelty rather than a stable behavioral shift. Treat 16% as a directional estimate, not a fixed constant.
Why the effect shows up in the first 10 minutes
Session-length curves for capped and uncapped players are nearly identical for the first 8 to 12 minutes. Then they diverge. Uncapped players keep spinning through what researchers call the "decision-free zone" — the stretch where the reels turn without any input. Capped players hit the 25-spin wall, and somewhere between 30% and 40% of them, depending on the operator, don't restart. They cash out, switch games, or close the app.
That restart friction is the whole mechanism. It's not that capped players gamble less per spin. Their average bet size is actually 4% higher, likely because the pause gives them a beat to reconsider stake. It's that they stop more often. A 25-spin cap inserts a mandatory pause roughly every 90 seconds on a typical 3.5-second spin cycle. Over 40 minutes, that's about 26 decision points where an uncapped player has zero.
What the cap does and doesn't change
The industry's preferred framing is that autoplay caps protect problem gamblers. The data is more interesting than that.
| Metric | Uncapped (100+) | Capped at 25 | Change |
|---|---|---|---|
| Avg. session length | 41m 18s | 34m 41s | −16.1% |
| Avg. bet per spin | $0.94 | $0.98 | +4.3% |
| Sessions per player per week | 3.1 | 3.4 | +9.7% |
| Avg. daily loss (net) | $22.40 | $21.10 | −5.8% |
| Players setting a loss limit | 11% | 19% | +8 pts |
Read that table carefully. Sessions get shorter, but players come back slightly more often. Net daily loss drops modestly — 5.8%, not the 16% the session-length headline implies. And the biggest single behavioral change isn't session length at all: it's that capping autoplay nearly doubles the share of players who set a loss limit before they start.
That last number deserves scrutiny. It could be causal — the cap forces a pause, and the pause prompts reflection. It could also be selection. Operators that capped autoplay at 25 were often the same operators that redesigned their deposit screens in the same release, adding a loss-limit prompt that didn't exist before. Three of the four operators confirmed the prompt was added within two weeks of the autoplay change. So the 8-point jump in loss-limit adoption is probably a package effect, not an autoplay effect.
The problem with treating 25 as magic
Twenty-five is a round number chosen for regulatory convenience, not behavioral science. The research on slot session length and continuous-play features — most of it from Australian and Canadian gambling studies published between 2018 and 2023 — points to pause frequency mattering more than pause count. A 25-spin cap on a fast game with 2-second spins creates a pause every 50 seconds. The same cap on a slower game with 5-second spins pauses every 2 minutes. If the mechanism is "interrupt the decision-free zone," the cap should scale with spin speed, not sit at a fixed number.
No U.S. operator currently scales autoplay caps to spin speed. The 25 figure is applied uniformly across a game library that includes titles running from 1.8 to 6.2 seconds per spin. That's a design gap, and it means the 16% figure likely understates the effect on fast games and overstates it on slow ones.
The U.S. regulatory picture is messier than Europe's
In the UK, autoplay caps arrived through a combination of regulator pressure and operator self-regulation, and by 2023 most major brands had settled on either 25 or 100 spins with mandatory loss-limit prompts. The U.S. has no equivalent. Online casino is legal in seven states as of early 2025 — Connecticut, Delaware, Michigan, New Jersey, Pennsylvania, Rhode Island, and West Virginia — and each has its own rules. None currently mandates an autoplay cap. New Jersey's Division of Gaming Enforcement raised the topic in a 2024 responsible-gambling review but stopped short of a rule. Michigan's Gaming Control Board has been the most active, and its 2024 annual report flagged autoplay as an area "warranting further study" — which in regulatory language means it's on a list somewhere.
Sports betting operators, which vastly outnumber casino-only platforms in the U.S., have largely stayed out of the autoplay conversation because their products don't have the feature. That leaves the fight to a handful of casino platforms and the social-casino apps, where autoplay is common and lightly regulated.
The practical result: a player in Pennsylvania can set autoplay to 500 spins on some platforms. The same player, on the same game, on a platform licensed in the UK, can't exceed 100. If the 16% figure holds, that gap represents a meaningful difference in how long the average session runs — and by extension, how much the average player loses per sitting, even if the daily net loss difference is closer to 6%.
What operators say privately
Two compliance officers at U.S.-licensed casino platforms, both of whom asked not to be named because they weren't authorized to speak to press, described autoplay caps as a "cheap win" — a feature change that costs almost nothing in revenue and buys goodwill with regulators. One estimated the revenue impact of a 25-spin cap at "under 2% of slot GGR" for a typical operator. The other put it closer to 4% but noted that the loss is concentrated among high-volume players, who are also the players most likely to generate complaints.
That's the tension. The players most affected by an autoplay cap are the ones who use autoplay most — and those are disproportionately high-stakes, high-frequency players. A 16% session-length reduction for the average player might translate to a 25% or 30% reduction for the top decile of players by volume. Operators know this. It's why the caps arrived slowly, and why some platforms set them at 50 or 100 instead of 25.
What a 16% drop actually means for a player
Strip out the operator framing and the number is smaller than it sounds. If your typical slot session runs 40 minutes and you're betting $1 a spin at 3.5 seconds per spin, you're making about 685 spins and cycling roughly $685 through the machine. A 16% shorter session means about 575 spins and $575 cycled. At a 96% RTP, expected loss drops from about $27.40 to about $23.00. That's a $4.40 difference per session. Over a year of twice-weekly play, it's roughly $450.
That's not nothing. It's also not transformative, and it's an expected-value calculation — actual results swing far wider than $4 either way on any given session. The more meaningful change for most players is the loss-limit adoption number, if it holds up. A player who sets a $50 loss limit and sticks to it has capped their downside regardless of how long autoplay runs.
The honest read: autoplay caps are a modest, well-targeted intervention that produces a real but small reduction in play time and an uncertain reduction in harm. The 25-spin figure works because it's simple and defensible, not because it's optimal. The 16% session-length drop is real in the data, but it's a proxy for harm, not a measure of it. Session length and gambling harm correlate, but they aren't the same thing — a player who plays 34 minutes and chases losses is worse off than one who plays 41 minutes and stops at a preset limit.
The open question: does the cap scale, or does it just move the behavior?
Here's what nobody has measured yet. If a 25-spin cap shortens sessions by 16% but increases session frequency by nearly 10%, the net effect on weekly play time is closer to 7% than 16%. And if players adapt — learning to restart autoplay the moment it stops, which takes about two taps — the cap's friction disappears within a few weeks of exposure. The 90-day data window in this sample is too short to know whether the effect decays.
That's the question regulators should be asking before they write 25 into a rulebook: does the cap reduce harm, or does it just relocate the same amount of play into more, shorter sittings? The early data can't answer it. A 12-month longitudinal study, tracking the same players before and after a cap is imposed, would. Nobody has funded one in the U.S. yet. Until someone does, the 16% figure will keep getting cited — by operators as proof they're responsible, by advocates as proof the number should be lower, and by regulators as a reason to act without quite knowing what they're acting on.
If you're playing slots tonight, the cap on your platform — or its absence — is worth a look before you spin. So is the loss limit. One of those two settings is doing more work than the other, and it isn't the one the industry is arguing about.