RTP, or return to player, is the percentage of all money bet on a game that the game is designed to pay back over the long run. The house edge is what is left: a 96% RTP game has a 4% house edge, so on average it keeps £4 of every £100 staked. Volatility describes how that 96% is paid out: often and small, or rarely and big. RTP tells you the price. Volatility tells you how bumpy the ride is.
RTP vs house edge
They are two views of the same number:
RTP = 100% − house edge
Table games are usually described by house edge (European roulette 2.70%, so 97.30% RTP), slots by RTP (96%, so a 4% edge). Both describe what the game keeps on average from every pound you bet, not from your deposit. Bet the same £10 ten times and you have staked £100, and the edge applies to all £100.
What each bet costs, on average
House edge: the share of every bet the casino keeps over the long run.
Video poker: Jacks or Better9/6 pay table, perfect play
0.46%Blackjackbasic strategy, 3:2, 6 decks, dealer stands on soft 17
0.50%Baccarat: Banker8 decks, 5% commission
1.06%Baccarat: Player8 decks
1.24%French roulette: even moneysingle zero, La Partage
1.35%Craps: Pass Line
1.41%European roulettesingle zero, any bet
2.70%Sic Bo: Small or Big
2.78%Dragon Tiger: Dragon or Tigerhalf returned on a tie
3.73%Online slots (typical)94% to 97% RTP; varies by game and version
3–6%American roulettedouble zero, most bets
5.26%Baccarat: Tiepays 8 to 1, 8 decks
14.36%
How RTP is measured and certified
A slot's RTP comes from its maths model: every symbol, reel strip, payline and feature. Game designers calculate the theoretical return exactly and confirm it by simulating many millions or billions of rounds. Independent test labs then check the maths and the random number generator before the game can go live in regulated markets. The best-known labs include:
- eCOGRA, founded in 2003 and based in the UK, which lists RNG testing and RTP testing among its services.
- GLI (Gaming Laboratories International), which says it provides independent testing and inspection of electronic gaming products, including interactive gaming.
- BMM Testlabs, which lists maths and RNG testing and product certification and says it has served the industry for more than 40 years.
In Great Britain, the Gambling Commission's technical standards require operators to make information about the chance of winning available before you bet, including the RTP percentage or the probability of winning. The guidance refers to "the average theoretical return to player percentage". The key word is theoretical: it is the long-run average, not a promise for any session.
Why RTP says little about one session
A slot's RTP is earned across all players and all spins. One player's 200 spins are a tiny sample, and a single feature or jackpot can swing the result by hundreds of times the stake. You can finish a session well ahead or lose everything you set aside on a 97% game. What RTP does predict reliably is your average cost across many sessions.
Volatility and hit frequency
- Volatility (or variance) is how widely results spread around the average. Low-volatility games return their RTP through many small wins. High-volatility games hold most of their RTP back for rare features and big multipliers.
- Hit frequency is the share of spins that win anything. A game with a 25% hit frequency wins on about one spin in four, but many of those wins pay back less than the stake.
Studios often publish a volatility rating. Evolution's games site lists NetEnt's Starburst at 96.09% RTP with low to mid volatility and Dead or Alive 2 at 96.80% with high volatility and wins of up to 100,000x the bet. Similar RTPs, very different experiences.
Worked example: same RTP, different volatility
To show this, we ran a simulation in Python of two simple illustrative slots (made-up games, not real titles). Both have an RTP of exactly 96%. The simulation ran 100,000 sessions of 200 spins at £1 a spin on each, so each session staked £200 and the expected loss was £200 × 4% = £8.
- Low-volatility game: wins on 53.2% of spins. 20% of spins return the stake, 30% pay 2x and 3.2% pay 5x.
- High-volatility game: wins on 23.8% of spins. 20% of spins pay 2x, 3.6% pay 10x and 0.2% pay 100x.
| Result per 200-spin session (our simulation) | Low volatility | High volatility |
|---|---|---|
| Average result | −£8.01 | −£8.12 |
| Median (typical) result | −£8 | −£30 |
| Sessions finishing ahead | 29.4% | 35.4% |
| Middle 80% of sessions | −£28 to +£13 | −£76 to +£80 |
| Lost £50 or more | 0.4% | 33.6% |
| Worst session | −£78 | −£148 |
| Best session | +£64 | +£492 |
The extreme figures (best and worst session) shift a little each time the simulation is run; the averages and the shape do not. Both games cost almost exactly the £8 the maths predicts on average. The difference is the shape. The low-volatility game gave a narrow band of results around the average. The high-volatility game made a typical session much worse (a median loss of £30) while a minority of sessions finished far ahead, which pulls the average back to −£8. A third of high-volatility sessions lost £50 or more of the £200 staked, against fewer than 1 in 200 on the low-volatility game. That is what volatility means in practice: the same price, a much wider spread of outcomes, and a typical session that feels worse than the RTP suggests.
RTP versions: the same slot at different RTPs
Many studios release one game in several RTP versions (for illustration, 96%, 94% and 92%), and each casino chooses which to run. The game looks and plays the same; the maths behind it pays back a different share of stakes. Red Tiger's Gonzo's Quest Megaways, for instance, is listed on Evolution's games site with an RTP of "94.66% / 96.00%".
How to check the version you are playing:
- Open the game and find the info, help or "i" button, or the paytable menu.
- Scroll to the game rules. Look for a line such as "the theoretical RTP of this game is 96.00%".
- If a range is shown, check whether separate figures apply to features such as a bonus buy, where offered.
- Compare with the same game at another site if you want the higher version. Do not rely on the studio's website, which may show a different version from the one the casino runs.
Worked example: what RTP costs you
Expected loss = stake per spin × number of spins × (1 − RTP)
| Stake | Spins | 96% RTP | 94% RTP | 92% RTP (expected loss, our calculation) |
|---|---|---|---|---|
| £0.20 | 500 | £4 | £6 | £8 |
| £1 | 500 | £20 | £30 | £40 |
| £1 | 1,000 | £40 | £60 | £80 |
Moving from a 96% version to a 92% version doubles your average cost for the same play. A fast slot can run several hundred spins an hour, so check the version before you start.
The same formula sets the cost of clearing a bonus. A £20 bonus at 10x wagering means £200 of bets: about £8 on a 96% game and £16 on a 92% version. Try your own numbers:
Wagering calculator
Wagering applies to
You need to bet
£500
About 500 bets at £1
Average cost at 96% RTP
£20
A long-run average, not a prediction
In any one session you can lose more, including your whole deposit.
Mistakes to avoid
- Treating RTP as a session forecast. It is a long-run average across all players.
- Assuming a "due" payout. Each spin is independent. A game that has not paid for a while is not about to.
- Ignoring the version. Two casinos can run the same title at different RTPs.
- Choosing volatility your budget cannot carry. High-volatility games need a larger budget relative to your stake to reach the features. If your budget is small, lower the stake.
- Using betting systems. Raising or lowering stakes after wins or losses does not change the RTP or the house edge.
For game-specific maths, see slots, Megaways slots, bonus buy slots, roulette and blackjack, or read our guide to wagering requirements.
Where to play
We only list licensed casinos. Browse UK casinos, Ontario casinos, or US casinos in New Jersey, Pennsylvania and Michigan. Read how we rate casinos.
Play it safe
Gambling is entertainment that costs money, not a way to make it. Decide your budget before you start, use the deposit limits every licensed site offers, and stop when it stops being fun.
Free, confidential help where you liveQuestions people ask
Does a higher RTP mean I will win more?
It means you will lose less on average over a large number of bets. In a single session, volatility and luck matter far more than a difference of one or two percentage points.
How do I find a slot's RTP?
Open the game's info, help or paytable screen and look for 'theoretical RTP' or 'return to player' in the rules. That figure is the version the casino is running, which can differ from the figure on the studio's website.
Is RTP the same as my chance of winning?
No. RTP is the average share of stakes paid back. Hit frequency is how often a spin wins anything, and many 'wins' pay back less than the stake.
Can a casino change a slot's RTP while I play?
In regulated markets, a casino chooses which certified version of a game to offer. It cannot tweak results in your session; the version's maths is set and tested before the game goes live.
Related reading
Sources
- UK Gambling Commission: RTS 3, rules, game descriptions and the likelihood of winning
- eCOGRA: Testing and certification
- Gaming Laboratories International (GLI)
- BMM Testlabs
- Evolution (Red Tiger): Gonzo's Quest Megaways
- Evolution (NetEnt): Starburst
- Evolution (NetEnt): Dead or Alive 2
Facts checked 14 Oct 2026. Compliance review 14 Oct 2026.
Hub of Luck is an independent guide, not an operator, regulator or law firm. Spotted an error? We correct within 48 hours. Editorial policy · How we test · How we make money