Volatility describes how far a game's payouts can swing around their expected average. In the UK, these games are usually called slots. Lower-volatility games tend to spread returns across smaller, more frequent prizes; higher-volatility games put more of the return into rare, larger payouts. Neither label predicts the next win.
Two slot machines with the same return to player (RTP) can leave players with very different balances after the same number of spins. RTP gives the long-run average return. Volatility describes the variation around that average, which matters when assessing how far a session balance could move.

What does volatility mean in game mathematics?
A slot machine's paytable assigns prizes to qualifying outcomes, while its game mathematics assigns probabilities. Together they form a probability distribution covering losing spins, partial returns, line wins, bonus features and rare top awards. The expected return and its variation come from that distribution. Variance measures the average squared distance of outcomes from the mean. Standard deviation is the square root of variance, expressed in payout units rather than squared units. Providers often call either measure volatility, but a low, medium or high label is usually a classification, not a published variance figure.
For a fixed stake, let X be the gross return per spin in multiples of the stake, including zero for a loss:
- Expected return:
E[X] = Σ(probability × payout multiplier). - Variance:
Var(X) = Σ[probability × (payout multiplier − E[X])²]. - Standard deviation:
SD(X) = √Var(X).
The UK Gambling Commission uses standard deviation in its technical explanation of RTP monitoring. When it is measured in stake multiples, multiplying by the stake gives the spread in the pounds sterling.
Low, medium, high and very high volatility
Low, medium, high and very high are descriptive bands, not universal numerical ranges. One studio's five-point scale need not match another's. Without the underlying payout probabilities, a rating cannot be converted into the likelihood of finishing a session ahead.
| Rating | Typical payout profile | What the label does not promise |
|---|---|---|
| Low | More of the return is distributed through relatively frequent, modest prizes; fluctuations tend to be narrower. | A win on every few spins, a stable balance or protection from a complete loss. |
| Medium | A mix of smaller base-game payouts and less frequent larger outcomes. | An evenly spaced sequence of wins or a universally defined middle point. |
| High | Greater dispersion, with a larger share of returns potentially attached to uncommon features or substantial prizes. | A large prize during a particular session, or a higher RTP. |
| Very high / extreme | An especially uneven distribution that may include rare, large multipliers and long stretches with little returned. | A standardised numerical risk rating or a jackpot that becomes more likely after losses. |
Volatility vs RTP, hit frequency and maximum win
| Measure | What it measures | How to read it |
|---|---|---|
| Volatility / variance | Dispersion of payouts around the expected return. | A category, provider-specific index or statistical value; not a winning percentage. |
| RTP | Expected gross payout divided by total stakes over a very large sample. | 96% RTP corresponds to £0.96 returned per £1 wagered on average, not in each session. |
| House edge | The mathematical share of turnover not returned as prizes. | With a fixed 96% RTP, the theoretical house edge is 4%. |
| Hit frequency / hit rate | How often a defined qualifying payout occurs. | A 30% rate does not guarantee three hits in the next ten spins; definitions of a hit must be checked. |
| Profit-making win rate | How often a payout exceeds the stake on that round. | Different from a hit that merely returns part of the wager. |
| Bonus frequency | How often a free-spins round or another specified feature triggers. | Not the average size or profitability of that feature. |
| Maximum win | The largest payout permitted by the game's specified rules or configuration. | A ceiling, not a probability or a typical result. |
A £1 spin paying £0.40 is a hit under some definitions, but still a £0.60 net loss. Winning paylines, lights and sound effects do not change that arithmetic. A game's hit rate can be much higher than its rate of profitable spins, and neither figure gives the probability of finishing a session ahead.
Theoretical RTP comes from the designed probability model. Actual RTP is the prizes paid divided by turnover in a particular sample. Actual results can stay above or below the theoretical rate for a long time when rare prizes account for a substantial part of the expected return.
Same RTP, same hit frequency — different volatility
Compare two hypothetical slot machines charging £1 per spin. Each spin is independent. The percentages describe the chances of receiving the stated gross payout, not net profit. These are invented distributions, not commercial games.
| Payout on a £1 spin | Game A | Game B |
|---|---|---|
| £0 (0×) | 60% | 60% |
| £0.40 (0.4×) | 0% | 39% |
| £2.40 (2.4×) | 40% | 0% |
| £80.40 (80.4×) | 0% | 1% |
Game A: 0.40 × 2.4 = 0.96. Game B: (0.39 × 0.4) + (0.01 × 80.4) = 0.96. Each has 96% theoretical RTP, a 4% house edge and a 40% hit frequency when every non-zero payout counts as a hit.
Game A's standard deviation is about £1.18 per spin; Game B's is about £7.99. Game A pays more than the stake on 40% of spins. In Game B, 39% return just £0.40, and only 1% pay more than the stake. Identical RTP and hit frequency hide a very different distribution of profitable outcomes.
Game B's 80.4× prize has a 1% chance on each independent spin. Over 100 spins, the chance of missing it altogether is 0.99100 ≈ 36.6%. Both models have an expected loss of £4 per 100 £1 wagers, but individual results can be far apart. Expected loss is not a spending cap.
How a game's design creates volatility
A slot machine's random number generator (RNG) produces values that the game maps to reel positions or other outcomes. Reel strips, virtual stops and symbol weighting influence the chance of particular combinations; the paytable sets their prizes. That combined distribution, not the player's recent results, determines volatility. Paylines, ways-to-win layouts and cluster-paying systems define which combinations qualify. Wilds, scatters and expanding wilds affect prizes or trigger features. Cascading reels, respins, free spins, retriggers and multipliers change the distribution of total payouts per round. Maximum-win caps belong in that calculation too.
A slot machine can pay frequent small base-game prizes and still have high overall volatility if rare free-spin features or a progressive jackpot carry much of its RTP. Bonus frequency is not the same as average bonus value. Very large, infrequent jackpot payments may also require separate monitoring.
Optional modes or paid bonus entry can change the stake, payout distribution and sometimes RTP. The relevant figures are those for the mode in use. Near misses, dramatic animations and sound effects offer no evidence about the odds of the next spin.
What volatility changes during a playing session
Bankroll is the money available for play. Turnover counts all wagers, including those funded by earlier payouts. A £50 starting balance can generate more than £50 in turnover without another deposit; the deposit and the total staked are different amounts. At £0.50 a spin, 300 spins create £150 of turnover. At a hypothetical 96% RTP, the expected loss on those wagers is £6. The session might finish ahead or lose much more. Volatility affects the range of results, while additional spins increase the amount wagered.

Raising a stake from £0.50 to £1 doubles the dollar amounts of expected losses and payout swings if the distribution stays unchanged. It does not change volatility measured in stake multiples. Bet-dependent features, alternative modes and progressive-jackpot eligibility can break that assumption, so the game rules matter.
Risk of ruin is the probability of exhausting a bankroll before a chosen stopping point. Calculating it needs the starting balance, stake, complete payout distribution, spin count and stopping rules. A low-volatility label alone cannot say how long £50 will last.
Across independent spins with a fixed distribution, the standard deviation of the average payout falls in proportion to 1/√n, with n spins. A single session still need not resemble the expected RTP. Rare prizes may remain absent for hundreds of spins; a poor run alone does not establish that the RNG is faulty.
How to verify a slot machine's volatility rating
The game's own information screen and the developer's specification are more useful than a third-party volatility label. For the exact version being offered:
- Match the exact title and version. Sequels and similarly branded games may use very different payout models.
- Read the displayed RTP. A supplier may offer several RTP configurations, so the figure on the version being used matters more than a generic review.
- Look for the provider's volatility or variance field. Record the original low/medium/high category or numerical scale without treating it as a universal measurement.
- Check the paytable and feature rules. Identify maximum win, paylines or ways, bonus triggers, multipliers and any separate modes.
- Check what counts as a hit. An advertised hit frequency may include prizes worth less than the original wager and may not include the same events across different games.
NetEnt/Evolution lists the original Starburst as Low / Mid, with 96.09% RTP. Starburst XXXtreme is rated High, with published RTP figures of 96.26% and 96.45%. Shared branding does not mean the same mathematics, and available configurations still need checking.
A 4/5 volatility rating is not an 80% win probability. A maximum prize of 10,000× tells you the ceiling, not the odds of reaching it. A demo or a few hundred observed spins cannot reliably measure a distribution with very rare prizes.
UK slots: statistics and player protections are separate
In Great Britain, check the operator against the Gambling Commission public register and read the specific game rules. Published return figures, payout schedules, jackpot eligibility and player controls can differ between games and versions. A mathematical model or laboratory test does not establish that an operator is licensed.
Mathematical probabilities and expected returns cannot predict an individual session or eliminate the risk of losing money.
Common questions about slot volatility
Does high volatility mean a higher RTP?
No. RTP is the average return; volatility is the spread around it. Either volatility category can appear alongside different RTP percentages.
Does a high hit frequency mean I am more likely to make a profit?
Not on its own. A counted hit may pay less than the stake. Profit depends on the amounts returned, not just how often a payout occurs.
Can a slot machine become due for a big payout after a losing streak?
Not in an independent-spin game. Previous losses do not change the next spin's probability. Features that carry state between rounds have their own rules; a growing jackpot changes the prize amount, not automatically its trigger probability.
Do free spins and large multipliers always mean extreme volatility?
No. Free spins, expanding wilds and multipliers affect volatility through their probabilities and payout sizes. A frequent small bonus can contribute less variance than a rare base-game prize.
Can a demo or a few hundred spins reveal a game's volatility?
Not reliably. Rare awards may be absent from a short sample. A verified paytable or provider classification carries more weight than a run of personal spins.
Is low volatility a safer way to win money?
No. Narrower fluctuations do not create positive expected value. A low-volatility slot machine can still exhaust the player's starting balance.
