Slot descriptions often quote a return-to-player figure and a volatility rating side by side, as if they were two views of the same thing. They are not. RTP is a long-run average. Volatility describes how far single results stray from that average. This post builds three invented slots to show the difference. The Play listing for Brasslight Slots states neither number, so nothing below is a claim about that game. The toy slots exist only to make the arithmetic visible.
What RTP measures
Return to player is the total paid out divided by the total staked, over a very large number of spins. A 96% RTP means that for every 100 units staked, 96 come back on average, and the other 4 units are the house edge. It is a property of the pay table and the reel layout, and it can be worked out exactly from them.
It is a statement about the long run, not about a session. It does not say you will get 96 back from 100 staked, or that any particular spin is close to 96%. A single spin returns 0 or a multiple of the stake, and a figure like 96% appears only as an average across enormous numbers of spins.
What volatility measures
Volatility, sometimes called variance, is the spread of results around the average. The usual measure is the standard deviation: take each possible payout, find how far it sits from the mean, square that gap, average the squares weighted by probability, then take the square root. A bigger number means results are more scattered.
Two slots can share an RTP and differ completely in spread, as the next section shows.
Three toy slots, one RTP
Each slot below costs 1 unit per spin and returns exactly 96 units per 100 spins on average. They differ only in how that 96 is packaged.
| Slot | Pays | Hit rate | RTP | Spread per spin |
|---|---|---|---|---|
| A | 3 units | 32 in 100 | 32 × 3 ÷ 100 = 96% | 1.40 |
| C | 9.6 units | 10 in 100 | 10 × 9.6 ÷ 100 = 96% | 2.88 |
| B | 96 units | 1 in 100 | 1 × 96 ÷ 100 = 96% | 9.55 |
The hit rate is how often a spin pays anything. Slot A pays often and pays little, slot B pays rarely and pays a lot, and slot C sits between them. Their RTPs are identical, so RTP cannot tell you which one you are playing.
The spreads come from the definition above. For slot B, the average payout is 0.96. A win sits 95.04 above that (1 time in 100) and a loss 0.96 below it (99 times in 100). The variance is 0.01 × 95.042 + 0.99 × 0.962 = 90.33 + 0.91 = 91.24, and its square root is 9.55. A shortcut gives the same answer: the average of the squared payouts minus the squared average, 0.01 × 962 − 0.962 = 92.16 − 0.92 = 91.24. For slot C the same shortcut gives 0.1 × 9.62 − 0.962 = 9.216 − 0.9216 = 8.29, so the spread is 2.88. For slot A it is 0.32 × 9 − 0.9216 = 1.96, a spread of 1.40.
What 100 spins looks like
Independent spins add their variances, so the spread over 100 spins is 10 times the single-spin spread, since the square root of 100 is 10. That gives 14.0 units for slot A, 28.8 for C and 95.5 for B. The expected result is the same for all three: 100 units staked and 96 returned, a net of −4.
Read the numbers against the stake. With slot A, about 95% of 100-spin sessions land within two spreads of the average, which is roughly between −32 and +24 units. Slot B comes in lumps. The chance of no hit in 100 spins is 0.99100 = 36.6%, which leaves you down the full 100 units. The chance of exactly one hit is 100 × 0.01 × 0.9999 = 37.0%, which leaves you down 4. The remaining 26.4% have two or more hits and finish at +92 or better. The mean is −4, but only about 37% of sessions finish anywhere near it.
RTP tells you the average. Volatility tells you how often a real session lands far from it. Over a short session the scatter is usually larger than the house edge.
What follows from this
- A high RTP does not mean a gentle slot. Slot B has the same 96% as slot A and ten times the spread over 100 spins.
- A high hit rate does not mean a high RTP. A slot that pays on 32 spins in 100 can still return less than one that pays on 1 in 100. Only hit rate multiplied by average win decides RTP.
- More spins shrink the relative scatter, not the absolute one. The spread grows with the square root of the spin count while the expected loss grows with the count itself. After 10,000 spins the spread for slot B is 955 units against an expected net of −400, so the average finally starts to dominate. For slot A it is 140 units, and the average dominates much sooner.
Reading a real slot
A real game has many symbols, a ways count such as the 243 described in what 243 ways to win means, wilds as in how a wild changes the count, and bonus rounds on top. Its RTP is still one weighted average of all those outcomes, and its volatility is still the spread of the same set.
The earlier post on why a slot is never due applies here too. Neither number gives a spin any memory. RTP is not a target the machine steers back towards, and volatility does not mean a quiet stretch is building towards a payout.
Brasslight Slots: Casino Games is a cat-themed 5-reel slot with 243 ways to win, a Yarn Pick bonus, a daily prize wheel and Turbo and Auto spin. It uses virtual coins only, there is no real money in it, and coins and gems have no cash value.