RTP and volatility: what the numbers actually say

RTP is the most quoted figure in this industry and the most widely misunderstood. Volatility is the figure that would fix the misunderstanding and is quoted far less often.

What RTP measures

Return to player is the share of total stakes a game returns over a very large number of spins. A 96% RTP means that across millions of spins the game returns about 96p per pound staked.

What it does not mean is that £100 becomes £96 in an evening. Over a few hundred spins, the distribution around that average is enormous, and the average is not where most sessions land.

Why volatility is the missing half

Two games can share a 96% RTP and behave nothing alike.

A low volatility game pays small amounts frequently. The balance drifts rather than swings, and a session lasts a long time on a small bankroll.

A high volatility game pays almost nothing for long stretches, then occasionally a great deal. Over millions of spins it returns the same share; over one evening it either does very little or one thing.

The practical consequence is about bankroll. A high-volatility game needs enough balance to survive the dry stretch, or the session ends before the mathematics has room to work — and the published RTP never arrives.

The same slot is not always the same slot

A number of titles ship in more than one RTP build — 96.x, 94.x, sometimes lower — and the operator picks which to run. This is legal and disclosed in the game info, and almost nobody looks.

It takes about ten seconds to check, and it is the single most useful thing to verify before playing a specific title at a specific site. A comparison page quoting 96% is quoting a catalogue; the game info is quoting what is actually running.

Why so many figures are missing

Catalogue data is thin. In the source used for the demo pages on this site, RTP is present on roughly one game in eight and volatility on about one in ten. Where a figure is absent here, it is absent rather than estimated — an invented number that looks like a measurement is worse than a gap.

What to check in practice

The game info panel, before playing: RTP, volatility if listed, and maximum win. Three fields, ten seconds, and they beat any comparison table including this one — because they describe the specific build running at that specific operator.

Turning RTP into an hourly figure

RTP becomes concrete once pace is included. A slot at 96% returns £96 of every £100 staked over a long run, so the expected cost is £4 per £100 — but the figure players feel is cost per hour, and that depends entirely on how fast they play.

At £1 per spin and roughly six seconds a spin, six hundred spins an hour means £600 staked and an expected loss near £24. At £5 per spin, the UK ceiling for players aged 25 and over, the same hour stakes £3,000 and expects a £120 loss. The RTP did not change. The pace and the stake did, and they are the two variables actually under a player’s control.

Why variance makes any single session useless as evidence

The long run in the definition of RTP is longer than any real session. On a high-volatility slot, tens of thousands of spins can pass before observed return approaches the stated figure, which is why one evening tells a player nothing about whether a game is paying correctly.

This is also why the phrase due for a payout describes nothing real. Each spin is independent of every previous one, and a long losing run changes the next spin’s odds not at all. The only thing a session reliably measures is how much was staked.