Casino basics
What Is RTP (Return to Player) and Why It Matters
RTP is the single number that tells you how much of every £100 wagered on a game is expected to be paid back to players over millions of spins. A 96% RTP means the game is mathematically built to return £96 for every £100 staked across its full payout cycle — and to keep £4 as the house edge. That is the floor of what you need to know before you spin a single reel.
The catch is that RTP is a long-run figure, not a session promise. You can wager £100 on a 96% RTP slot and walk away with £0 or £500 in a single sitting. This guide explains the math, where to find verified RTP, why variance changes how the number feels in practice, and how to use RTP to compare games honestly.
RTP defined in one paragraph
Return to Player is the percentage of all wagered money that a game returns to players as winnings over its theoretical payout cycle. Cycle length is usually measured in tens of millions of spins. The figure is set by the game’s math model, certified by an independent test lab (typically eCOGRA, GLI, or iTech Labs), and published by the provider on a per-game datasheet. The remaining percentage — 100% − RTP — is the house edge, the operator’s mathematical margin.
A 96% RTP slot does not mean you will get £96 back per £100 in your session. It means that across the full lifetime of the game, summed across every player and every spin, the payout will converge towards 96%. Your individual session is a tiny sample of that distribution.
The math, with actual numbers
Imagine you wager £1 per spin on a slot with a stated 96.7% RTP — for example Pragmatic Play’s Big Bass Splash — for 10,000 spins. The expected return is:
- Total wagered: £10,000
- Expected return: £10,000 × 0.967 = £9,670
- Expected loss: £330
That £330 is the casino’s mathematical edge over 10,000 spins at £1 stake. The actual result for any individual playing 10,000 spins will sit somewhere on a wide distribution curve around that £9,670 expected value — usually within a few percent, but occasionally far outside it.
Now compare that to a 94% RTP slot, also at £1 × 10,000 spins:
- Total wagered: £10,000
- Expected return: £9,400
- Expected loss: £600
The 2.7 percentage point difference in RTP costs you nearly twice as much in expected loss for the same play volume. Over a year of regular play, RTP differences of two or three points compound into hundreds of pounds.
RTP vs variance — the part casinos do not advertise
Two slots can both have a 96% RTP and feel completely different to play. The reason is variance (also called volatility) — how spread out the wins are around the expected return.
| Variance | Hit frequency | Win size | Bankroll feel |
|---|---|---|---|
| Low | High (1 in 3-4 spins) | Small (1-10× stake) | Slow, steady drain |
| Medium | Moderate (1 in 5-7) | Mid (10-50× stake) | Mixed sessions |
| High | Low (1 in 10-20) | Large (50-5000× stake) | Long dry spells, occasional big hit |
A 96% RTP, low-variance slot will feel like you are “almost winning” most of the time — small payouts that net to a slow loss. A 96% RTP, high-variance slot like Hacksaw Gaming’s Dork Unit (96.42% RTP, max 10,000× stake) will pay you nothing for 200 spins, then hit a 500× win, then nothing for 300 more.
Both games have the same expected return. The session experience is wildly different. High-variance games chew through bankrolls faster in the short term, even if the math is identical over a million spins. If you only have £20 to play, a high-variance game can wipe it out in 15 minutes; a low-variance game might keep you spinning for an hour at the same stake.
Where to find verified RTP
You should never trust the RTP number a casino prints on its in-game info panel without cross-checking. Some games ship with multiple RTP versions — for example, a “high” version at 96.5% and a “low” version at 94% — and operators are free to choose which version they license. The lower version still carries the provider’s brand, so you might play a “96.5% RTP slot” while the casino has actually licensed the 94% build.
The safe sources are:
- Provider websites — Pragmatic Play, NetEnt, Play’n GO, Hacksaw Gaming, Nolimit City, and the other major studios publish per-game RTP datasheets. Pragmatic’s game library lists RTP under each title.
- Independent databases — SlotCatalog tracks RTP, variance, and which operators are running which version. BigWinBoard and Casino Guru also list verified figures.
- Test lab certificates — eCOGRA and GLI publish operator-level seal verification. Click the eCOGRA badge in a casino’s footer and you can read the live audit report.
We have seen casinos quietly downgrade RTP versions when traffic spikes. The UK Gambling Commission’s LCCP requires UK-licensed operators to disclose the RTP of every game they offer; in practice this is buried in a help-centre article.
RTP across game categories — what to expect
Different game types sit in different RTP bands. The numbers below are the typical industry range, not absolutes.
| Game type | Typical RTP range | House edge |
|---|---|---|
| Video slots | 94% - 97% | 3% - 6% |
| Blackjack (basic strategy) | 99.4% - 99.6% | 0.4% - 0.6% |
| European roulette | 97.3% | 2.7% |
| American roulette | 94.74% | 5.26% |
| Baccarat (banker bet) | 98.94% | 1.06% |
| Video poker (Jacks or Better, optimal play) | 99.54% | 0.46% |
| Crash games (Aviator, etc.) | 97% | 3% |
| Live game shows (Crazy Time, etc.) | 94% - 96.5% | 3.5% - 6% |
| Scratch cards (online) | 85% - 95% | 5% - 15% |
A few takeaways:
- Table games dominate RTP rankings, but only with correct play. Blackjack at 99.5% assumes you play perfect basic strategy. A casual player who hits and stands by feel can give back 2-3% of that edge.
- Slots are the worst-RTP house product by category. They make up for it with variance (the chance of a big hit) and visual engagement.
- Live game shows are slots in disguise from a math perspective — Crazy Time runs at around 95.4% RTP and feels like a casino floor, but the edge is closer to a slot than to roulette.
How to use RTP without fooling yourself
A handful of practical rules drawn from years of game testing:
“RTP tells you the long-run drain rate. It does not tell you whether you will win tonight.” — that is a common framing from gambling-math analysts, and it is the right one. RTP is the most useful single comparator between games, but it does nothing to predict a single session.
- Use RTP to rank games, not to predict results. A 97% RTP slot is genuinely better for your bankroll than a 94% slot of the same variance — but neither will reliably profit you.
- Pair RTP with variance. A 96% low-variance slot at £0.20 stake is the longest-playing combination for a small bankroll. A 96% high-variance slot at the same stake might be over in 50 spins.
- Cross-check the version before you play. If the in-game info panel says 94%, that is the version this operator licensed — not the headline 96.5% number you may have seen on a review site.
- Ignore “hot” and “cold” labels. Slots use certified random number generators. Past spins do not influence future spins. Any UI element implying a “hot streak” is design, not math.
- Set a session budget independent of RTP. A higher RTP does not change the fact that the expected outcome of any session is a loss. Slots are entertainment with a price, not a return-generating activity.
Frequently asked questions
Is a higher RTP always better?
For the same game category and variance level, yes — a 97% RTP slot is mathematically better than a 95% RTP slot. But variance and stake matter as much for the session experience. Two 96% RTP slots can play very differently.
Why do casinos run multiple RTP versions of the same game?
Providers ship games in 2-5 RTP tiers (e.g. 96.5% / 95% / 94% / 92%). Operators choose which build to license, usually based on their margin model. This is allowed in most jurisdictions but the active version must be disclosed to players on request.
Does the casino set the RTP?
No — the casino picks a version, but the math is built into the game by the provider and certified by an independent test lab. Casinos cannot tweak RTP on a running game. They can only switch to a different licensed version.
How long do I have to play before RTP “kicks in”?
The honest answer is: longer than you will ever play. Convergence to stated RTP requires tens of millions of spins. In any practical bankroll (£20 to £2,000), your outcome is dominated by variance, not by RTP. RTP is the average across all players combined, not a guarantee for your session.
What is the lowest RTP I should accept?
For slots, anything below 96% is poor value when 96.5-97% options exist on the same casino. For table games, accept nothing below the standard published RTP for that variant — if a blackjack table is paying 6-to-5 on naturals instead of 3-to-2, the RTP drops by around 1.4 percentage points and the game is no longer worth playing.
A note on bankroll and risk
RTP math assumes you keep playing. The biggest leak in most players’ bankrolls is not the 4% house edge — it is chasing losses, raising stakes after a downswing, or playing past the budget set at the start of the session. If you find yourself doing any of those, the RTP of the game becomes irrelevant; the player error swamps the math.
Set a session deposit cap, set a session time cap, and walk away when either is reached. If you are unsure whether your play is staying recreational, the PGSI screening on BeGambleAware takes three minutes and the result is private.
Affiliate disclosure
GambleDragon earns commission when readers sign up to operators we review. We never receive payment for higher RTP claims, and we cross-check every RTP figure on this site against provider datasheets and independent databases before publishing. If you spot an RTP we have listed incorrectly, tell us and we will fix it within 24 hours.