House edge vs RTP vs volatility: what the numbers really mean
Three numbers describe the money side of almost every casino game: RTP, house edge and volatility. The first two are the same idea seen from opposite sides. The third is what you actually feel during a session. Mixing them up is the most common reason players misjudge how long a bankroll will last.
RTP: return to player
RTP is the share of all money staked that a game pays back on average, over a very large number of rounds. A slot with a 96% RTP returns $96 for every $100 wagered in the long run. It is a property of the payout table and the probabilities behind it, computed mathematically or by simulating billions of rounds. It is not a forecast for your session.
Two details matter in practice. First, RTP is measured against turnover (total amount staked), not against your deposit. Re-betting the same $100 fifty times means $5,000 of turnover. Second, many modern slots ship with several RTP settings and the operator chooses one. The same game can be configured differently at two casinos, so always check the info screen of the version you are playing.
House edge
House edge is simply what the RTP leaves for the casino:
house edge = 1 − RTP expected loss = total staked × house edge
Some well-documented reference points:
| Game | RTP | House edge | Note |
|---|---|---|---|
| European roulette (single zero) | 97.3% | 2.7% | 37 pockets, straight-up pays 35:1 |
| American roulette (double zero) | 94.74% | 5.26% | 38 pockets, most bets |
| Stake-style dice / limbo | 99% | 1% | As published by the operator for that algorithm |
| Slots | varies | varies | Depends on game and configuration; check the info screen |
| Blackjack | varies | varies | Depends on rules and on playing correct strategy |
The roulette numbers follow directly from counting pockets: a straight-up bet wins 1 time in 37 and pays 36 units back (35 profit plus the stake), so the return is 36/37 ≈ 97.3%.
A worked example
Suppose you make 1,000 bets of $1 on a game with a 96% RTP. Total staked is $1,000. The expected loss is $1,000 × 4% = $40. That is the centre of the distribution of outcomes, not the outcome you will get. You might finish up $150 or down $300, depending on the game’s volatility.
Now the same player moves to a 99% game with the same 1,000 bets of $1. Expected loss falls to $10. The edge matters a lot over time: across 100,000 bets of $1, the gap between 96% and 99% is $4,000 vs $1,000 of expected loss.
Volatility: how wide the swings are
Volatility (also called variance) describes how spread out results are around the average. Two games can both have a 96% RTP while feeling completely different:
- Low volatility: frequent small wins. An even-money bet on a 99% dice game wins a bit under half the time and roughly doubles the stake when it does.
- High volatility: long runs of losses broken by rare large wins. A slot whose RTP comes mostly from a feature that triggers once in hundreds of spins behaves like this.
The standard way to measure the spread is the standard deviation (SD) of the result of one bet, in units of the stake. Over n independent bets, the SD of the total grows with the square root of n, while the expected loss grows linearly with n:
expected result after n bets = −n × bet × house edge SD after n bets ≈ bet × SD per bet × √n
This is why short sessions are dominated by luck and long sessions are dominated by the edge. With $1 bets, an SD per bet of 1 (roughly an even-money game) and 100 bets, the SD of the total is about $10 while the expected loss at 96% is $4: the edge is barely visible. After 10,000 bets the SD is about $100 but the expected loss is $400: the edge now outweighs the noise. A high-volatility game with an SD per bet of 10 needs a hundred times more bets before the same thing happens.
What this means when you play
- RTP tells you the price, volatility tells you the ride. Over enough turnover, you pay the house edge on everything you stake. Volatility only decides how unevenly that cost arrives.
- Bet size and number of bets drive cost. Expected loss is turnover × edge. Faster games and auto-play increase turnover per hour, so they increase cost per hour too, even at the same RTP.
- Bankroll vs volatility. With a small bankroll relative to the bet, high-volatility games often end the session before any big win arrives. That is not unfairness; it is the shape of the distribution.
- Past results do not change the odds. In independent games, a losing streak does not make a win more likely. Systems that raise stakes after losses (such as Martingale) change the shape of results (many small wins, occasional large losses) but not the expected loss per unit staked.
- Bonus terms interact with RTP. Wagering requirements multiply turnover, so the house edge applies to a much larger amount. See wagering requirements explained.
Checking an RTP claim
You cannot verify a 96% RTP by playing a few hundred rounds: the swings are far larger than the edge, as the numbers above show. What you can check is whether a game result was generated the way the casino says. For games that use a provably fair scheme, our provably fair verifier recomputes individual rounds from the seeds, and the RTP of such games follows from the published formula. For slots, the practical checks are the info screen, the provider’s game sheet and whether the casino discloses which RTP configuration it runs. If it does not, treat that as missing information rather than assuming the highest version.
FAQ
Is a 99% RTP game "almost free" to play?
No. A 99% RTP means you lose about 1% of everything you stake on average, not 1% of your deposit. If you stake $10,000 in total over many bets on a $500 deposit, the expected loss is about $100 - and because turnover adds up fast, a small edge still costs real money.
Does RTP change depending on how much I have won or lost?
Not for independent games such as dice, roulette or most slots. Each round is a fresh draw; there is no memory that makes a win "due". The RTP describes the average of the payout table, not a promise that results will balance out in your session.
Where can I find a game’s real RTP?
Look at the game’s info or help screen, and the provider’s published game sheet. Many slots are released in several RTP configurations and the operator picks one, so the same title can have a different RTP at different casinos. If a casino does not show the figure, treat that as missing information.
Is high volatility better or worse than low volatility?
Neither changes the expected loss for a given RTP and turnover. High volatility makes outcomes more spread out: more sessions end in a large loss, a few end in a large win. Low volatility keeps results closer to the average. Choose based on your bankroll and how much swing you can tolerate.
Related guides
18+. Gambling involves risk of loss. Help and self-exclusion.