A slot listed at 96% RTP is mathematically designed to return $96 for every $100 wagered, but only over a very large number of spins, at 1,000 spins of $1 each, the expected loss works out to $40, and the actual result of any given session can land far from that figure. Volatility is the separate number that describes how bumpy the road to that average is, and it matters just as much as RTP for how a session actually feels.
Two numbers get thrown around casually in the same sentence, and they measure completely different things. One is an average taken over an enormous sample. The other is how far individual results tend to wander from that average before the sample gets large enough to matter.
What RTP actually measures
Return-to-player is the percentage of total wagers a game is built to pay back, calculated across a sample large enough that short-term luck washes out, in practice, simulations run into the millions of rounds during independent testing. A game’s RTP is fixed into its paytable and confirmed by a certification lab before it launches under a regulated licence, as covered in How Online Casinos Work. It is not a setting that adjusts session to session, and it says nothing about whether the next ten spins will be good ones.
Theoretical RTP versus what a specific machine shows
There’s a second figure worth knowing about, separate from the certified theoretical RTP: some jurisdictions’ land-based reporting also tracks actual payback on physical machines over a given period, a number that reflects real play on real hardware rather than a lab simulation. Online, this distinction matters less because every round runs against the same certified paytable regardless of who is playing it, but the underlying idea is the same one this whole guide comes back to: a percentage measured over a long enough run behaves very differently from the same percentage measured over an afternoon.
It’s worth being precise about what “theoretical” means here, since the word can sound like a hedge. It isn’t one. The theoretical RTP is the exact figure the game’s mathematics produce when run to completion, the number the certification lab confirms by simulating enough rounds that the result stabilises to within a tight tolerance of the paytable’s design. Nothing about a real session changes that underlying math; what changes is only how much of it has had a chance to play out by the time a player stops.
A worked example
The table below shows expected loss at different RTP levels over 1,000 spins of $1 each, a total of $1,000 wagered in every row. The math is simple: expected loss equals the amount wagered multiplied by the house edge, where house edge is 100% minus RTP.
| RTP | House edge | Total wagered (1,000 spins × $1) | Expected loss |
|---|---|---|---|
| 98% | 2% | $1,000 | $20 |
| 96% | 4% | $1,000 | $40 |
| 94% | 6% | $1,000 | $60 |
| 92% | 8% | $1,000 | $80 |
“Expected” is the operative word. It describes the average outcome across many players running many such sessions, not a forecast for any one of them. A single player’s 1,000 spins might finish well ahead or well behind $40 down, and that gap is exactly what volatility describes.
At 96% RTP and a stake of 1 per spin, 1,000 spins carry an expected loss of 40. The chart shows how far a single session wanders from that average; every slider change re-runs it.
What volatility means
Volatility (sometimes called variance) describes the size and frequency of a game’s payouts relative to the stake, independent of RTP. A low-volatility game tends to return small wins often, so a session’s balance moves in small steps and rarely swings far from its starting point. A high-volatility game pays out rarely, but when it does the payout can be many multiples of the stake, so a session’s balance can sit flat for long stretches and then jump sharply.
An older land-based analogy makes this concrete: a mechanical three-reel slot with a modest top jackpot behaves very differently from a progressive machine chasing a jackpot that only pays once in a very long while, even if their long-run RTPs happen to land close together.
Volatility isn’t usually published as a precise number the way RTP is; operators and game studios more often describe it in bands, low, medium, high, sometimes with a numeric scale layered on top for a specific studio’s catalogue. That looseness is deliberate. Volatility depends on the entire distribution of a game’s possible payouts, not a single summary statistic, so a tidy one-number label would hide more than it reveals. A game studio’s own internal volatility index is a reasonable rough guide, but it isn’t standardised the way RTP certification is, and figures aren’t directly comparable studio to studio without checking how each one defines its scale.
Why volatility matters for bankroll planning
Two players with identical stakes and an identical session length can have very different experiences purely because of the volatility setting, independent of anything either one does. A low-volatility game with a small bankroll tends to produce a long, gradually declining session, useful for someone who wants to play for a set amount of time on a fixed budget. A high-volatility game with the same small bankroll carries a real chance of running out of funds before a meaningful payout ever lands, precisely because its wins are rarer and larger rather than frequent and small. Neither approach is “better” in the abstract; they suit different goals, and confusing the two is one of the more common reasons a session doesn’t go the way a player expected.
How a slot’s RTP compares to a fixed-odds table game
A slot’s RTP and a table game’s house edge are the same underlying math, both describe the average share of wagers a game keeps, but they get there differently. A roulette wheel or a blackjack shoe has a house edge dictated entirely by fixed physical or mathematical odds, described bet by bet in Roulette Bets Explained. A slot’s RTP, by contrast, is a designed figure: the studio building the game sets the paytable and symbol weightings to land on a target percentage, then has that target verified, not derived from a naturally fixed set of odds like a wheel’s 37 or 38 pockets. Both numbers converge the same way over a large enough sample; only the mechanism that produces them differs.
How the two interact
RTP and volatility are independent axes, not two names for the same thing. A 96% RTP low-volatility game and a 96% RTP high-volatility game will converge on roughly the same long-run return, but the paths they take to get there look nothing alike. A player who wants steady, predictable sessions and one who wants a shot at a rare large payout are, mathematically, looking for different points on this grid, even at an identical RTP.
Why the same game can list different RTPs at different casinos
This surprises a lot of players who assume a slot has exactly one RTP everywhere it’s offered. Many studios build a game with several selectable RTP configurations from the start, and license each configuration separately to operators, meaning the same title can genuinely carry, say, a 96% build at one casino and a 94% build at another, both certified, both real. This is disclosed in the game’s own information screen if you check it, which is worth doing before assuming a title behaves identically everywhere it appears.
Where the certified figure comes from
A slot’s stated RTP isn’t a marketing claim; it is checked by an independent lab as part of licensing, alongside the RNG that drives the game, before the figure can be published or displayed in-game. The House Edge Explorer puts this same figure alongside blackjack, roulette and other table games for direct comparison, and the RTP and Volatility Simulator lets a reader run a bankroll through many spins at a chosen setting to see the spread first-hand.
The honest caveat: no simulation, however many spins it runs, tells a player what their own next session will do. It only shows what the distribution looks like once enough sessions have been added together, which is precisely the number that stays invisible while you’re still inside it.
The law of large numbers, in plain terms
The statistical idea underneath all of this has a name: the law of large numbers, which says that as a sample grows, its average gets closer to the theoretical expected value. It says nothing about small samples, which is exactly why a 30-spin losing streak or a lucky first ten rounds tells you almost nothing reliable about a game’s true RTP. If you’ve ever wondered why a “hot” or “cold” machine feels so real in the moment and yet the whole industry insists RTP is fixed, this is the answer: both things are true. The short-term streak is real and the long-run percentage is also real; they’re simply answers to different-sized questions.
Frequently Asked Questions
What is a good RTP for a slot game?
Published slot RTPs commonly range from roughly 92% to 98%, and figures at the higher end of that range are generally considered better for the player over the long run. RTP is only half the picture, though; a game's volatility changes how that average feels session to session.
Does RTP apply to a single spin or a single session?
Neither. RTP is calculated over a very large number of rounds, typically millions in simulation. A single spin either pays or it doesn't, and a single session can land well above or well below the long-run figure by chance alone.
What's the difference between RTP and volatility?
RTP describes how much of all wagered money a game is designed to return over time. Volatility describes how that return is distributed: a low-volatility game pays small amounts often, while a high-volatility game pays rarely but can pay much larger amounts when it does. Two games can share the same RTP and feel completely different to play.
Can a casino change a game's RTP after it's certified?
Not on the fly. RTP is built into the game's paytable and certified by an independent testing lab as part of licensing; changing it typically requires a new build and a new certification, not a setting an operator can quietly flip.
Why did I lose more than the RTP suggested I should?
Because you experienced one outcome from a distribution, not the average of that distribution. Over 1,000 spins the numbers start to resemble the stated RTP; over 50 spins they frequently don't, especially on a high-volatility game.
Where can I check a specific slot's RTP before playing?
Regulated operators are generally required to disclose RTP somewhere in the game's information screen or help menu. If it isn't listed anywhere, that is itself worth noting before playing.