Volatility and hit frequency: why slots with the same RTP play completely differently
Two slots can both be built to return 96% of what is staked and still behave like different games. One pays something back often, in small amounts. The other pays nothing on most spins and, occasionally, a large amount in one go. RTP cannot tell them apart. Volatility and hit frequency can, though the four studios covered below publish them in different ways, when they publish them at all.
What slot volatility measures
RTP is the share of everything staked that a game is built to pay back over millions of spins. Our RTP guide covers what that figure means and why one game often carries several of them. What RTP leaves out is the shape of the returns: whether the 96% arrives as a steady trickle of small payouts or as rare, large ones.
That shape is volatility. The researchers Kevin Harrigan and Mike Dixon describe a game’s volatility index as a measure of how much its payback percentage varies over a given number of plays. A 2025 paper by Nigel Turner and colleagues explains how the industry calculates it: from the standard deviation of the outcome, the statistical measure of how far single results tend to land from the average.
The two games below are made up and deliberately basic, but nothing in the table is estimated: every figure is worked out exactly from the probabilities.
| Game A | Game B | |
|---|---|---|
| What it pays | 1.92x the stake on 50% of spins | 96x the stake on 1% of spins |
| RTP | 96% | 96% |
| Hit frequency | 50% | 1% |
| Standard deviation of one spin | 0.96x the stake | 9.55x the stake |
| 100 spins: chance of no payout at all | practically 0% | 36.6% |
| 100 spins: middle 90% of results, as a share of the total staked | 80.6% to 111.4% | 0% to 288% |
| 1,000 spins: middle 90% of results, as a share of the total staked | 91.0% to 101.0% | 48.0% to 144.0% |
| 1,000 spins: chance of getting back less than half the total staked | practically 0% | 6.6% |
Both games have a median result of exactly 96% of the total staked, after 100 spins and after 1,000. What changes is everything around that middle. Game A keeps its results in a narrow band, while Game B spreads them across a band more than nine times wider, so it produces more sessions that end far above 96% and more that end far below it. Neither is the better game: averaged over every possible session, both return 96%.
Volatility leaves the RTP untouched. It also says nothing about the next spin. Each spin is drawn independently, so a quiet run does not make a large payout any more likely.
Hit frequency: how often a spin pays anything
Hit frequency is the share of spins that return something, of any size. Harrigan and Dixon, quoting an earlier source, define it as the percentage of time a machine gives any payout. They also note that a manufacturer’s probability sheets can state it per line rather than per spin, so two figures quoted for the same game can differ without either being wrong. Paylines, ways and clusters count wins differently too, as how slots pay explains.
A payout is also not the same thing as a win. On one 15-line game in their study, 33.52% of spins paid something, and 60.73% of those payouts were smaller than the amount staked. The machine still greets them with the sights and sounds of a win. Researchers call these losses disguised as wins. In a 2010 study of 40 novice players published in Addiction, they produced larger skin conductance responses, a measure of physical arousal, than ordinary losses did.
A third made-up game, Game C, shows how far apart paying and winning can sit:
- It pays 0.5x the stake on 32% of spins and 10x on 8%, which gives an RTP of exactly 96%.
- Its hit frequency is 40%: four spins in ten pay something.
- Four in five of those payouts are half the stake, so 92% of all spins hand back less than was staked.
A high hit frequency can describe a game that mostly returns less than the stake. And neither figure is the RTP. As Nigel Turner and Roger Horbay pointed out in 2004, a 90% payback is no promise of a payout on 90% of spins.
Four studios, four formats
Four of the studios behind the games here handle volatility and hit frequency in four different ways. Here is what each one puts on its own website.
NetEnt shows volatility as a graphic on its game pages and prints hit frequency as a percentage. The three figures below come from netent.com:
Gonzo’s Quest
41% hit frequency · 95.97% RTP
The highest hit frequency of these three NetEnt games.
Gonzo’s Quest demo →
Starburst
23% hit frequency · 96.08% RTP
Almost the same RTP as Gonzo’s Quest, and a little over half its hit frequency.
Starburst demo →
Mega Joker
17% hit frequency · 99% RTP
A 3x3 classic with five fixed lines and a Supermeter game above it. NetEnt shows it as high volatility.
Mega Joker demo →Gonzo’s Quest and Starburst sit 0.11 of a percentage point apart on RTP and 18 points apart on hit frequency. Mega Joker pairs the highest RTP of the three with the lowest hit frequency, so a high RTP is no sign that a game pays often. NetEnt’s pages do not say how a hit is counted, and each percentage is best read as NetEnt’s own measure.
Habanero uses words. Each of the 102 game pages we read on its site carries one of five ratings, from very low to very high, and the RTP guide shows how those games split across them.
Play’n GO shows no volatility in the visible text of its Book of Dead page. The data behind that page rates the game’s volatility as high, and the studio’s rules document for the game lists a volatility of 7 without saying what scale that number sits on. Book of Dead demo →
Pragmatic Play prints an RTP on its game pages, 96.50% for Gates of Olympus for instance, but none of the Pragmatic Play pages we read carried a volatility rating. A rating quoted for one of those games comes from somewhere other than the studio’s own page.
NetEnt’s percentage is a hit frequency, a different measure from a volatility rating. The three studios that do rate volatility each use their own scale, and none of them publishes a way to convert one into another, so we have not ranked games across studios.
Where volatility comes from
Turner and Horbay describe volatility as a natural result of a slot’s variable prize structure, and a 2011 simulation study by Turner looked at how prize structure shapes it. The arithmetic is simple. Once the RTP is fixed, every point of it set aside for rare, large prizes is a point that cannot go to frequent, small ones.
A paytable whose best combinations pay far more than the rest puts more of the return into outcomes that seldom land. A game that can pay thousands of times the stake has to make room for that ceiling inside the same RTP as everything else. When much of the return comes from a bonus round, the base game has less to pay out between triggers.
Two Habanero games show the pattern side by side. The studio publishes nearly the same RTP settings for both, about 86%, 92%, 94% and 98%, plus 97% on one and 96% on the other. It rates them at opposite ends of its scale.
Hot Hot Fruit
Very high volatility · 5,087x maximum multiplier
15 paylines, and free games in which up to 14 symbols from winning combinations lock in place for the rest of the round.
Hot Hot Fruit demo →
Arcane Elements
Very low volatility · 608x maximum multiplier
50 paylines, and 10 free spins in which you pick two elements to act as wilds.
Arcane Elements demo →The ceiling on Hot Hot Fruit is more than eight times higher. Habanero does not publish how much of either game’s return comes from its free games, so the pair illustrates the pattern rather than measuring it.
What a few hundred demo spins can and cannot show you
A demo runs on virtual credits, which makes it a good place to watch how a game behaves. A few hundred spins can show you:
- The pace of the game: whether something lands on most spins or long gaps are normal.
- What a payout smaller than the stake looks like, and how the game celebrates it.
- What the features do, if one triggers in that time.
What they cannot show you is the RTP or the volatility rating: a few hundred spins are far too few to measure either. For a land-based game with a 92.5% payback, Harrigan and Dixon’s figures put the middle 90% of results at 59.45% to 125.71% of the amount staked after 1,000 plays, narrowing to 91.54% to 93.63% only after a million. In the table above, Game B goes 100 spins without a single payout 36.6% of the time, while returning 96% over the long run.
Impressions mislead as well. In a lab study, 48 students played simulated machines set to differ in both payback and how often they paid. How well a player could tell the machines apart by how often each one paid was linked to prior learning and beliefs, the gambler’s fallacy in particular, and not to the machines’ payback. Players also placed more bets on the machine with the lower payback. A slot has no memory: a long dry spell does not make a payout due, and a run of payouts does not mean it is about to end.
The demo shows how a game plays. It says nothing about what a real-money session would return, and it is not practice for one.
Quick answers
- What does volatility mean in slots?
- How widely results spread around the RTP over a stretch of play. Low-volatility games tend to pay smaller amounts more often and high-volatility games larger amounts less often, but not always: how often a game pays is a separate figure, its hit frequency.
- What is hit frequency?
- The share of spins that pay anything at all, including payouts smaller than the stake.
- Does high volatility mean a higher RTP?
- No. They are separate measures: Habanero publishes nearly the same RTP settings, about 86% to 98%, for a very low and a very high volatility game.
Try it on virtual credits
Every game on this site runs on virtual credits, with nothing deposited and nothing paid out. The paytable inside each game shows its prize structure.
If gambling has stopped being fun for you or someone close to you, the responsible gambling page lists where to find help in South Africa.
Sources
- NetEnt game pages for Gonzo’s Quest, Starburst and Mega Joker.
- Habanero game pages for Hot Hot Fruit and Arcane Elements.
- Play’n GO: the Book of Dead game page and its game rules and settings.
- Pragmatic Play: the Gates of Olympus game page.
- Harrigan, K. A. and Dixon, M. (2009). PAR sheets, probabilities, and slot machine play. Journal of Gambling Issues, 23. doi:10.4309/jgi.2009.23.5
- Turner, N. and Horbay, R. (2004). How do slot machines and other electronic gambling machines actually work? Journal of Gambling Issues, 11. doi:10.4309/jgi.2004.11.21
- Dixon, M. J. and others (2010). Losses disguised as wins in modern multi-line video slot machines. Addiction, 105(10). doi:10.1111/j.1360-0443.2010.03050.x
- Turner, N. E. (2011). Volatility, house edge and prize structure of gambling games. Journal of Gambling Studies, 27(4). doi:10.1007/s10899-011-9238-0
- Coates, E. and Blaszczynski, A. (2014). Predictors of return rate discrimination in slot machine play. Journal of Gambling Studies, 30(3). doi:10.1007/s10899-013-9375-8
- Turner, N. E. and others (2025). Use of computer-based mathematical simulations to understand gambling games and potential risks for problem gambling. Journal of Gambling Issues, 56s. PDF