Slot Volatility vs RTP: What Each One Actually Changes

Slot volatility and RTP measure two different things. RTP (return to player) is the long-run percentage of all money staked that a slot is designed to pay back, so it sets the average price of each spin. Volatility describes how that money comes back: in frequent small amounts or in rare large ones. RTP decides how much a long run of play costs on average. Volatility decides how wide the range of results is for any one session, including how likely you are to lose your whole budget before you finish. Over a few hundred spins, volatility usually has much more effect on what you actually experience than a difference of a few RTP points.

The basic definitions are in our guide to how online pokies work, and the question of whether certified games are honest is covered in are online slots rigged?. This guide does something different: it puts numbers on what each figure changes. We use slot design documents that researchers in Ontario obtained through a freedom-of-information request, and a simulation we ran for this page of 100,000 sessions on each of three slot maths models.

The Short Version: Two Numbers, Two Different Jobs

	RTP	Volatility
What it measures	Average share of stakes paid back over a very large number of spins	How widely results spread around that average
What it changes	Your average cost per spin, and how long a budget lasts on average	Chance of losing the whole budget, size and rarity of wins, how far one session can stray from the average
What it does not change	What happens in any single short session	The average cost over time
Where it is published	Game help or info screen; UK rules expect RTP, house edge or win probability to be shown	Sometimes on the info screen as a studio’s own label; no regulator-defined scale in the rules we read

What RTP Actually Changes: The Price of Each Spin

RTP is easiest to understand as a price. A slot with a 96% RTP keeps, on average, 4% of everything staked on it. Your expected cost is just arithmetic:

Expected cost = stake per spin × number of spins × (100% − RTP)

Stake per spin	Spins	At 96% RTP	At 94% RTP	At 92% RTP
0.20	1,000	8	12	16
1.00	500	20	30	40
2.00	1,000	80	120	160

The figures are in whatever currency you play in. A small-looking RTP difference becomes a large difference in cost: going from 96% to 92% doubles the expected cost of the same play, because the house’s share goes from 4% to 8%.

The research behind this guide shows the same effect on real games. Kevin Harrigan and Mike Dixon of the University of Waterloo obtained PAR sheets, the maths documents that describe a slot’s design, for games in use in Ontario through a freedom-of-information request. They found that the province had approved several versions of the same five-reel video slot that look identical to the player but range from 85% to 96.2% RTP. On a $1.00 spin the 96.2% version costs about 3.8 cents on average and the 85% version about 15 cents, nearly four times as much. A player who starts with $10 and plays until it runs out would get about 263 one-dollar spins on average from the 96.2% version, but only about 67 from the 85% version.

So RTP mostly decides how long your money lasts on average. That matters. But it is an average across many sessions, and it tells you very little about one session.

What RTP Does Not Change: Your Next Few Hundred Spins

The same PAR sheets show how slowly a slot’s actual return settles towards its RTP. For a 92.5% single-line reel game, the manufacturer’s sheet gives a “volatility index” and the 90% range that the player’s actual return is expected to fall within for a given number of spins:

Spins played	Actual return expected within (90% of the time)
1,000	59.45% to 125.71%
10,000	82.11% to 103.06%
100,000	89.27% to 95.90%
1,000,000	91.54% to 93.63%
10,000,000	92.25% to 92.92%

After 1,000 spins, the range is more than 60 percentage points wide. A gap of three or four RTP points between two games is small next to that. It only becomes clearly visible after hundreds of thousands of spins, far more than a typical player ever plays on one title.

Harrigan and Dixon then tested this in their 2010 follow-up study. They simulated 2,000 first-time players, each starting with $100, playing an 85% and a 98% version of the same game until their money ran out. The typical (median) player’s experience was similar on both versions. The differences appeared in the tails, among the luckiest players: on the 98% version, many more players saw their balance peak above $1,000 (54 against 5). The authors suggest this is a concern rather than a comfort, because a large early win can be what draws someone into heavier play.

What Volatility Actually Changes: The Shape of a Session

Volatility is about spread. Statistically it is linked to the standard deviation of a single spin’s result, and some manufacturers turn that into a single figure. The PAR sheet in Harrigan and Dixon’s paper defines the volatility index as the standard deviation of the game multiplied by 1.65, the z-score for a 90% confidence interval. The 92.5% reel game above has a standard deviation of about 6.35 and a volatility index of about 10.48. The range for any number of spins is then the RTP plus or minus the volatility index divided by the square root of the number of spins. That is where the table above comes from.

Two things follow from that formula:

• A higher-volatility game needs many more spins to settle. Double the volatility index and you need four times as many spins to narrow the range by the same amount.
• Volatility changes the spread, not the centre. The midpoint of the range is always the RTP. A volatile game is not cheaper or more expensive on average; its results are just spread more widely around the same cost.

What players notice is how the game behaves: how often anything pays, how big the wins are, and how long the dry spells last.

Hit frequency is not the same as volatility

Hit frequency is the share of spins that pay anything at all. Low-volatility games tend to have higher hit frequencies, but the two are separate design figures. The PAR sheets also show that a “hit” is often not a win. On the 15-line game Harrigan and Dixon analysed in full, played on all lines, 33.52% of spins paid something, but 60.73% of those payouts were smaller than the amount staked on that spin. The machine celebrates them with lights and sounds while the player’s balance goes down. The paper describes these as wins that are in fact losses, and they can make a slot feel less volatile, and more generous, than it is.

Our Simulation: Same RTP, Different Volatility

To show volatility in isolation, we built three simple slot maths models and ran 100,000 sessions on each. These are illustrative models we designed for this page, not copies of any real game:

• Low volatility, 96% RTP: pays something on 47.0% of spins; top prize 30× the stake.
• Low volatility, 92% RTP: the same game with its top prize made rarer; still pays on 46.9% of spins.
• High volatility, 96% RTP: pays something on 16.6% of spins; prizes up to 5,000× the stake.

Each session started with a budget of 100 units, staked 1 unit per spin and stopped after 500 spins or when the balance dropped below one stake.

Result per session	Low vol, 96%	Low vol, 92%	High vol, 96%
Standard deviation per spin	2.6	2.4	30.8
Ran out of budget before 500 spins	12.5%	18.9%	85.5%
Finished ahead of the starting 100	34.4%	21.8%	10.4%
Median finishing balance	76.5	56.5	under 1
Top 1% of sessions finished at or above	231	197.5	1,355.5
Finished at 300 or more	0.06%	0.01%	6.13%
Median spins until the 100 ran out (no 500-spin cap)	1,347	911	158

Compare the columns in pairs:

• Changing only RTP (first two columns) moved results gradually. The typical session ended about 20 units lower, the chance of finishing ahead fell from about one in three to about one in five, and the budget lasted about a third fewer spins. The game still felt the same: it paid on almost half of all spins.
• Changing only volatility (first and third columns) changed the whole experience. With the same 96% RTP, the high-volatility game used up the full budget in more than four out of five sessions, and the median session ended with nothing left. Yet it also produced almost all the big results: about one session in sixteen finished at 300 or more, which the low-volatility game almost never did.

That is the main point of this guide. RTP moved the average. Volatility moved almost everything else.

Same Game, Different RTP: Why the Label Matters

The Ontario documents also show how operators can get different RTPs from the same title: designers change how often ordinary symbols appear on the reels while leaving the top-paying symbols and the number of reel positions unchanged. In the case Harrigan and Dixon examined, hit frequency barely moved between versions (4.9% and 5.2% per line). So the lower-RTP version does not look or feel obviously tighter. It mainly takes slightly more on average from a lot of ordinary spins.

That is why the RTP shown in the game you open matters more than a figure quoted for the title elsewhere. We cover the wider question of multiple certified versions in the rigged guide linked above; the practical point here is to check the number in the game’s own help screen before you play.

Where to Find Each Number

RTP. Under the UK Gambling Commission’s technical standard RTS 3, licensed operators must make available, before a player commits to a game, information that helps them make an informed decision, such as the house edge, the RTP percentage or the probability of winning. The guidance says this can sit in the game’s artwork, its help or “how to play” pages, or other supporting material, and should describe the average theoretical return. Other regulators set their own rules, but the help or paytable screen is the place to look first.

Volatility. None of the regulatory documents we read for this guide requires a volatility rating. When a game shows one, as a word like “high” or a scale of bolts or stars, it is the studio’s own label. One studio’s “medium” may not mean the same as another’s, and there is no shared definition behind it. If a game gives a maximum win as a multiple of the stake, that is a useful hint: a very large top prize has to be paid for with smaller or rarer ordinary wins, which usually means a more volatile game.

Neither number is a forecast. RTP is a theoretical average over a huge number of spins. A volatility label describes spread. Neither tells you what your next session will do.

Using Both Numbers to Set a Budget, Not to Win

Neither figure is a way to win. What they can do is make the cost of play easier to predict:

1. Use RTP to estimate the average cost. Multiply stake × spins × (100% − RTP) before you start. If that number is more than you are happy to spend on an evening’s entertainment, lower the stake or play less.
2. Use volatility to estimate the worst likely case. On a low-volatility game, a session will usually end somewhere around the average cost. On a high-volatility game, losing the whole session budget is the most likely single result; in our simulation it happened in most sessions. Only stake what you can lose in full.
3. Do not change games to chase a loss. A run of losses on a high-volatility slot is exactly what its design produces. Switching to a higher-RTP or “looser-feeling” game does not win back what is gone.

If you are comparing slots with other casino games on cost alone, our ranking of the lowest house edge casino games puts slot RTPs next to table games. A lower house edge still means a slower loss over time, not a likely profit.

Volatility and Bonus Wagering

Wagering requirements make the difference between the two numbers very clear. The expected cost of meeting a requirement depends on RTP: wagering 3,500 units at 96% RTP costs about 140 units on average (3,500 × 4%), more than a 100-unit bonus is worth. Volatility decides how likely you are to reach the end of that wagering at all. We ran the two 96% models again with a 100-unit balance and 3,500 units of wagering to complete at 1 unit per spin. The low-volatility game completed the wagering in 19.8% of 50,000 runs and the high-volatility game in 3.8%. In both, the typical run hit zero first.

Game weighting, maximum-bet caps and excluded titles usually matter as much as the maths, so read the full terms before opting in. Our casino bonuses comparison sets out the wagering terms side by side.

How We Worked Out These Figures

The PAR sheet figures, the Ontario versions and the hit frequencies come from Harrigan and Dixon’s 2009 paper in the Journal of Gambling Issues, and the 85% against 98% player simulation from their 2010 paper in the Journal of Gambling Studies. The display rules come from the UK Gambling Commission’s RTS 3. The three-model simulation is our own. We built each paytable to an exact RTP, used a fixed random seed so the results can be repeated, and ran 100,000 sessions per model (50,000 for the wagering test). The models are deliberately simple, so treat the percentages as an illustration of how RTP and volatility behave, not as a measurement of any real slot.

Keep It Affordable

Slots are built to keep more than they pay out over time, whatever their RTP or volatility. Set a budget and a time limit before you start, treat the whole budget as the price of the entertainment, and stop if play stops being fun or starts to feel like a way to make money. If that has already happened, the free and confidential help on our responsible gambling page is a good place to start. Online gambling is for adults only (18+, or older where local law requires it).