RTP, Volatility and Variance: the Same Three Numbers, Explained With Arithmetic

Q: What does 96% RTP actually mean?

It means the game is designed to return 96 pounds for every 100 pounds staked, averaged over a very large number of game cycles. It applies to total turnover rather than to your deposit, and it says nothing about any individual session.

Return to player, volatility and variance get used as though they were three words for one idea. They are not. RTP is an average, variance is the spread around it, and volatility is the label a studio puts on that spread. Two games can share an identical RTP and behave so differently that no amount of play would let you confuse them.

Here is what each one is, with the arithmetic done rather than asserted, and with the regulatory requirements that determine what you are entitled to see before you stake.

What you are entitled to know before you play

In Great Britain this is not left to the operator. The Gambling Commission’s Remote Technical Standards, RTS 3A, requires that “an explanation of the applicable rules must be easily available to the customer before they commit to gamble.” RTS 3C requires disclosure, before commitment, of the information needed to understand the chance of winning — the implementation notes give the acceptable forms as the “house edge (or margin),” the “return to player (RTP) percentage,” or the probability of each winning event (Gambling Commission, RTS 3, last updated 7 May 2024).

So the number is meant to be findable. It is usually in the game’s information panel or “how to play” screen. If it is not there at all, that is a fact about the site.

RTP is a long-run average, and the long run is longer than you think

A 96% RTP means that across an extremely large number of game cycles, the game is designed to return £96 for every £100 staked. The complement — 4% — is the house edge, and it is applied to turnover, not to your deposit. That distinction is where most misunderstanding starts. Stake £1 a hundred times and you have generated £100 of turnover from what may have been a £20 deposit.

The Commission’s own quarterly data lets us see the aggregate at work. In its market overview published 21 May 2026, covering January to March 2026, total online gross gambling yield “increased 7 percent… to £1.55 billion,” with “Slots GGY increased 12 percent year-on-year to £773 million.” The behavioural figures in the same release:

Metric, GB online slots	Jan–Mar 2026	Year earlier
Spins per session	124	136
GGY per session	£3.82	£4.01
Average session length	15 minutes	17 minutes
Total sessions	202 million	+18% year on year
Sessions over one hour	4.4%	5.9%

Source: Gambling Commission — market overview, operator data to March 2026.

One derived figure, flagged as derived: £3.82 of yield across 124 spins is about 3.1 pence lost per spin on average. If the underlying games ran at a 4% house edge, that would imply an average stake in the region of 77 pence. That is one aggregate divided by another rather than a measured stake, so treat it as an order of magnitude — but it is a useful corrective to the idea that these are large stakes. The edge does its work through repetition, not through size.

Variance, computed on two games with identical RTP

Take two hypothetical slots, both stake £1 per spin, both with an RTP of exactly 96%.

Game A returns £2 on 48% of spins and nothing otherwise.
Game B returns £100 on 0.96% of spins and nothing otherwise.

Both have an expected return of £0.96 per £1 staked. Now the spread:

	Game A	Game B
Expected return per £1	£0.96	£0.96
Variance	0.998	95.08
Standard deviation	≈ 1.0 × stake	≈ 9.8 × stake

Identical RTP, roughly a tenfold difference in dispersion. That is volatility, and it is the single most important number for what a session actually feels like — and for how much money you need for the session to resemble the average at all.

The waiting time follows directly. For Game B, the number of spins at which you have a better-than-even chance of hitting at least one £100 return is:

n ≥ ln(0.5) ÷ ln(1 − 0.0096) ≈ 72 spins

That is manageable. Now scale it to the kind of headline feature a modern high-volatility slot advertises. Suppose a top prize occurs on average once in 100,000 spins — an illustrative assumption, not a measured figure for any specific game. The same formula gives about 69,315 spins for a 50% chance of seeing it once.

Since RTS 14D requires that “it must be a minimum of 2.5 seconds from the time a game is started until the next game cycle” (Gambling Commission, RTS 14), 69,315 spins is at least 48 hours of continuous play. At the implied average stake above, it is well over £50,000 of turnover. Nothing about that is a criticism of the game. It is simply what “1 in 100,000” means when you convert it into time and money, which marketing copy never does.

Why a losing spin can look like a win

Two RTS rules exist because game design was exploiting exactly this gap between the average and the experience.

RTS 14F: “The gambling system must not celebrate a return which is less than or equal to the total stake gambled.” This targets the losses-disguised-as-wins effect — a £2 stake returning £0.40 accompanied by celebratory sound and animation. It is a loss. Under the current standard it must not be presented as anything else.

RTS 14D and 14G set minimum speeds: 2.5 seconds per slot game cycle, and a minimum five-second interval for other casino games. Speed multiplies the number of times the edge is applied per hour, so a speed floor is an exposure limit dressed as a design rule.

Alongside those, the stake ceiling is now statutory: since 9 April 2025 the maximum stake per online slot game cycle in Great Britain is £5 for players aged 25 and over, and since 21 May 2025 it is £2 for players aged 18 to 24, under The Gambling Act 2005 (Operating Licence Conditions) (Amendment) Regulations 2025.

Five things the numbers do not mean

RTP does not predict your session. It is the limit of an average over a sample far larger than any individual will ever play.
A game is never “due”. Each cycle is independent. A game that has not paid in 500 spins has exactly the same probability on spin 501 as on spin 1.
Volatility does not change the edge. High and low volatility versions of a 96% game both keep 4% of turnover. They differ only in how lumpy the path is.
A higher RTP is not a winning strategy. 97% and 95% are both losing propositions in the long run; one loses more slowly.
Published RTP is theoretical. It is a design parameter verified in testing, and some titles ship in multiple RTP configurations, which is why the number in the information panel is the one that matters rather than the number in a review.

What we could not verify

We could not find a regulator-published distribution of actual RTP settings deployed across the GB market, nor an official dataset of hit frequencies or top-prize probabilities for specific titles. The 1-in-100,000 figure used above is stated as an illustration and should not be read as a measurement of any real game. Where you need a specific game’s numbers, the game’s own information panel is the authoritative source, per RTS 3.

Frequently asked questions

What does 96% RTP actually mean?

It means the game is designed to return £96 for every £100 staked, averaged over a very large number of game cycles. It applies to total turnover rather than to your deposit, and it says nothing about any individual session.

What is the difference between volatility and RTP?

RTP is the average return; volatility is how widely results scatter around it. Two games can both have a 96% RTP while one has a standard deviation of about one times the stake and the other about ten times, which produces completely different session experiences on the same underlying edge.

Is a slot ever “due” to pay out?

No. Each game cycle is independent and generated randomly. A long losing streak does not raise the probability of the next spin, and no sequence of past results carries information about future ones.

Why does a slot celebrate when I win less than I staked?

Under UK rules it should not. RTS 14F provides that the gambling system “must not celebrate a return which is less than or equal to the total stake gambled.” If you see it happening on a GB-licensed site, that is a compliance issue rather than a feature.

How long does it take to hit a rare feature?

Convert the probability into spins. For an event with a 1-in-100,000 chance per spin, you need roughly 69,315 spins for a better-than-even chance of seeing it once, which at the regulatory minimum of 2.5 seconds per spin is about 48 hours of continuous play.

Sources

Related reading: our RTP explainer, slot volatility guide and live dealer vs RNG comparison.

Responsible gambling

Every number above describes a product designed to retain a percentage of what passes through it. Understanding the mathematics makes the outcome more predictable; it does not make it favourable. Decide the amount before you start, and treat it as spent.

If gambling is affecting you or someone you know, contact GamCare on 0808 8020 133 (free, 24/7) or visit BeGambleAware.org. UK players can self-exclude across all UKGC-licensed operators via GAMSTOP. You must be 18+ to gamble.

Editorial analysis. 18+. Please gamble responsibly.