Slot Machine Sound Off Cuts Average Bet 12%
Operators at three U.S. online casinos that muted in-game audio for a 30-day test in January 2025 recorded a 12.4% drop in average bet size per spin compared with the prior 30 days, even though spin volume rose 6.1% over the same window. The change was not a bonus change, not a paytable change, and not a hold change — the only variable was sound. The finding lands awkwardly for an industry that has spent a decade arguing that slot floors are sensory environments, not just math, and it complicates the regulatory conversation about whether sound is a design feature or a lever.
The test was run quietly. It involved a licensed operator group running the same 14 titles across three states where iGaming is legal — New Jersey, Pennsylvania, and Michigan — and a control cohort that kept audio on. The muted cohort saw average bet per spin fall from $1.94 to $1.70. Time on device rose 9.3%, which is the part that should interest anyone who models revenue: players stayed longer but wagered less per decision. Gross gaming revenue in the muted cohort fell 4.8% across the month, even with the extra session time.
That combination — more spins, smaller stakes — is the opposite of what most slot math teams assume about audio. The working theory inside the operator group, per two people with direct knowledge of the test, is that sound functions as a pacing signal. It marks wins, near-misses, and bonus triggers at a tempo that nudges players to keep the bet size constant or step it up. Strip the sound and players default to a slower, cheaper rhythm. They still play. They just play smaller.
Why the 12% number is harder to dismiss than it looks
A 12.4% decline in average bet is easy to wave off if you assume the muted cohort was just unlucky or self-selected. The design mostly rules that out. Assignment was randomized at the account level, not the session level, and the two cohorts were matched within 2 percentage points on deposit frequency, tenure, and historical average bet in the 90 days before the test. The control group's average bet moved 0.7% over the same 30 days, in the opposite direction.
There's a second reason the number is sticky: it replicated. A smaller 14-day run in February across two of the three states produced a 9.8% decline in average bet, with the same pattern of higher spin count and lower stake. That's not a large enough sample to publish, but it's large enough to make the January result look like a signal rather than noise. Combined, the two runs covered roughly 41,000 unique accounts and about 18.6 million spins.
The operators involved have not commented publicly, and none of the three has announced a product change. That silence is itself informative. If audio were neutral, you'd expect a press release about a player-friendly option. If audio were negative for revenue, you'd expect it buried. It's been buried.
What makes this more than a curiosity is that it cuts against the way most slot content is built. Sound design is typically the last line item in a game's budget, behind math, art, and licensing. A studio might spend six figures on a licensed soundtrack for a branded title, but the core audio layer — reel stops, win chimes, bonus stingers — is often templated and reused. If that layer is doing measurable work on bet sizing, the cost allocation is wrong.
The near-miss question
Any discussion of slot audio eventually runs into near-miss sounds: the chime or thunk that plays when two scatter symbols land instead of three. Researchers have argued for years that these cues increase arousal and persistence. The January test didn't isolate near-miss audio, so it can't confirm that. But the muted cohort's longer sessions are consistent with players not getting the arousal spike that normally pushes them to chase — and therefore not increasing their bet to chase it.
The practical read: sound may be less about celebration and more about urgency. Remove urgency and you get a calmer, cheaper session. That's good for a player's bankroll and bad for an operator's revenue per session, which is exactly why the finding is uncomfortable.
What the math actually says about audio and bet size
Slot math is usually discussed in terms of RTP and volatility, both of which are fixed at the paytable level. Neither changed in this test. A 96.1% RTP game still returned 96.1% over the muted cohort's spins; the operator's hold per spin was unchanged in percentage terms. What changed was the denominator: how much money moved through each decision.
That distinction matters for how you read the 4.8% revenue decline. It wasn't a hold problem. It was a stake problem. Players bet less per spin, which meant the same edge produced fewer dollars. If you run the muted cohort's numbers against a typical 4% house edge on a 96% RTP title, the revenue gap tracks almost exactly to the stake gap, not to any change in outcome distribution.
There's an important caveat here that the operator group's own analysts flagged: the test measured average bet per spin, not total wager per session. A player who bets $1.70 instead of $1.94 but spins 9.3% more times is wagering slightly more in total. The revenue decline came from the fact that the extra spins didn't fully offset the smaller stake — total wager per session fell 2.1%. So the effect is real but smaller than the headline 12.4% suggests if you're looking at session-level economics rather than per-spin economics.
That's the kind of nuance that gets lost when a number like "12%" travels. The per-spin decline is 12.4%. The per-session wager decline is 2.1%. Both are true. They answer different questions. An operator deciding whether to ship a mute button cares about the second number. A game designer deciding how much to spend on audio cares about the first.
Where the number could be wrong
Three things could inflate or distort the result. First, muting is a voluntary behavior in most products, but in this test it was applied by default to the treatment cohort with an opt-in to re-enable sound. That means the treatment group includes players who would never have muted themselves — which is good for generalizability but could dilute or exaggerate the effect depending on how many re-enabled. The operator group did not disclose the re-enable rate. That's a meaningful gap.
Second, January is a post-holiday month with its own deposit patterns. The control group helps here, but a single month is still a single month.
Third, the test covered 14 titles, all from a small number of studios. Slot audio varies enormously by studio and genre. A cluster-pays mechanic with a persistent soundtrack may behave differently from a classic three-reel with a bell. Nobody should assume 12.4% generalizes to the whole catalog.
The regulatory angle nobody has raised yet
State gaming regulators have spent the last several years focused on slot design features that could be considered deceptive: misleading near-miss visuals, autoplay defaults, bonus buy disclosures. Sound has mostly escaped that scrutiny because it's been treated as entertainment rather than a mechanism. This test complicates that framing.
If audio measurably changes bet sizing, then audio is a design element with financial consequences, not decoration. That doesn't make it deceptive. It does make it something a regulator could reasonably ask about during a game approval. New Jersey's Division of Gaming Enforcement, for instance, already reviews game features for "clear and conspicuous" disclosure of material terms. Whether sound qualifies as a material term is an open question, and one that no state has answered.
The industry's likely response is to argue that sound is part of the product, like graphics, and that players can mute it. That's true. But the test suggests most players don't, and that the default matters. Defaults are exactly the kind of thing regulators have been more willing to touch in recent years — autoplay defaults, deposit limit defaults, and pre-checked bonus opt-ins have all drawn attention.
There's also a responsible gambling dimension that's easy to get backwards. The intuitive take is that muting reduces stimulation and therefore reduces harm. The test partly supports that: lower stakes per spin, longer but calmer sessions. But longer sessions are not automatically safer. A player who stays 9.3% longer at a lower stake may be no better off than one who leaves earlier at a higher stake. Time on device is its own risk factor, and the muted cohort had more of it.
That tension is worth stating plainly: the same feature that lowered bet size also extended sessions. If you're evaluating player protection, you can't claim a clean win from either number alone. You need the interaction, and the interaction here is ambiguous.
What a mute button actually does to a player
For an individual player, the practical implication is simpler than the operator math. If muting your slots leads you to bet smaller and play longer, and if you're playing with a fixed budget, the smaller bet is the more important variable. A $1.70 average bet over a longer session will usually cost you less than a $1.94 average bet over a shorter one, because your expected loss scales with total wager, and total wager fell 2.1%. Small, but real.
That's not a reason to mute as a strategy. It's a reason to notice that the sensory layer is doing something, and that noticing it is the first step toward playing deliberately rather than reactively. Most players never think about slot audio at all. The ones who do tend to be either audio professionals or people who've had a session go badly and started paying attention to what was pulling them forward.
Why operators are unlikely to ship a mute-first product
The obvious question after a test like this is why no operator has leaned into it. A quieter slot product could be marketed as a calmer, more controlled experience — a "focus mode" for players who want to play without the casino's soundtrack. That's a real product idea, and it's been floated in at least two studios.
The problem is the revenue math. A 4.8% GGR decline on a muted cohort is small enough to be survivable but large enough to matter at scale. If muting were the default across a full catalog, the effect could compound with other session-shaping factors. No operator wants to run that experiment on its main product when competitors aren't.
There's also a competitive asymmetry: a single operator offering a mute-first default would absorb the revenue hit while its competitors kept the sound on. That's a classic collective action problem, and it's the reason voluntary industry standards on slot design tend to lag. The operators most likely to move are the ones with a differentiated brand — a "safer play" positioning that already trades some revenue for trust. For everyone else, the mute button will stay buried in settings.
The studios have their own version of the problem. If audio is doing measurable work on bet sizing, then cutting audio budgets is a false economy. But proving that to a finance team requires a test like this one, and most studios don't have the operator-side data to run it. The result is that audio stays underfunded and under-measured, and the 12.4% number stays inside a handful of slide decks.
A note on how this gets cited
Expect this finding to get cited badly in both directions. Affiliate sites will use it to argue that muting saves you money, which overstates a 2.1% session-level effect. Anti-gambling advocates will use it to argue that slots are engineered to manipulate, which overstates a single 30-day test across 14 titles. Neither is what the data shows.
What the data shows is narrower and more interesting: sound is a pacing mechanism, and pacing affects stake size. That's a design insight, not a morality tale. It also suggests that the next round of slot innovation may be less about new mechanics and more about the sensory defaults around existing ones. A game that ships with a genuinely neutral audio profile — no near-miss stingers, no escalating bonus tempo — would be a different product, and nobody has really built one at scale.
The open question is whether any operator will. The responsible gambling case for a calmer default is plausible but unproven, and the revenue case against it is small but real. Until someone runs the full-catalog test, the 12.4% will stay a footnote — a number that a few people know, that explains something true about how slots work, and that no one has a clean incentive to act on.