Slot Volatility: Why Two 96% Games Feel Nothing Alike
RTP tells you how much a slot returns over its full cycle. Volatility tells you how that return arrives: in a steady drip of small wins or in rare, heavy hits with long droughts in between. Of the two numbers, volatility is the one you actually feel at the reels.
What volatility is
Volatility (the studios' sheets sometimes print "variance") describes the spread of a game's payouts. Picture two slots, both with a 96% RTP. The first pays on one spin in four, almost always under 5x the stake. The second pays on one spin in ten, but its bonus can hit 500x. Over millions of spins they return the same share of money. Over your evening they are different animals: the first bleeds slowly and rarely spikes, the second swallows long stretches of your balance and then pays it back in lumps, or does not.
Low, medium, high: what the labels cost you
| Label | Typical pattern | What it demands |
|---|---|---|
| Low | Frequent small wins, top prizes modest | Small bankroll lasts; ceiling is low |
| Medium | Regular base wins, features carry the real money | A buffer for feature droughts |
| High | Long droughts, most of the return sits in rare big hits | Deep bankroll and patience, or a short, strict session |
One caution: the label is the studio's own wording, and there is no industry standard behind it. One provider's "medium" is another's "high". The label is a starting point, not a measurement, which is why I log sessions instead of just quoting it.
The numbers that reveal volatility
Four figures from a logged session tell you more than any label:
- Hit frequency. The share of spins that pay anything at all.
- Longest no-win run. The worst stretch with no meaningful win.
- Feature frequency. How often the bonus actually triggers.
- Best single win. The top payout the session produced, in multiples of the stake.
A worked example from the logs: Coin Princess x1000 by 3 Oaks is labelled high, and 6,135 logged spins agreed with the label while complicating it. About 30% of spins paid something, the bonus landed anywhere from once in 167 spins to once in 347 across three sessions, and one log held a 1,163-spin gap between features. Single bonus rounds ranged from 16.2x to 519.55x. A high contact frequency with the money parked in rare heavy rounds: in a logged sample, "high" looks exactly like that.
What this means for your bankroll
Match the game to the money, not the other way round. On a high-volatility slot, a bankroll of 100 bets is not conservative: at a bonus rate near one in 245 the logs still include a 1,163-spin gap, and you have to be solvent when the feature finally arrives. On a low-volatility game the same bankroll can run all evening, with the trade-off that the top end stays low.
The practical rule I use: decide how many spins you want your money to survive, divide your bankroll by that number, and that is your stake. It will not beat the house edge, nothing does, but it keeps volatility from ending your session in the first cold stretch. RTP context for that decision is in our RTP guide, and the gap between stated and session figures is covered in observed vs stated RTP.
Quick answers
Which is more important, RTP or volatility?
They answer different questions. RTP sets the long-run cost of play; volatility sets the shape of your session. For one evening with a fixed budget, volatility usually decides how the night feels. Over thousands of sessions, RTP decides what it all cost.
Can I measure volatility myself?
You can read its fingerprints in a demo session: hit frequency, no-win runs, feature gaps and best hits. A few hundred free spins will not give you the true distribution, but they will tell you whether the game is a drip-feeder or a hit-chaser before money is involved.
Do provider volatility labels lie?
They are honest but vague. There is no shared scale between studios, so treat the label as a rough class, then look at session data: a logged hit frequency and a longest-no-win-run figure say more than the word "high" ever will.
Former studio-side game mathematician: paytables and RTP models, now checked from the player's side. Full profile and method →