Does RTP Change By Casino? We Found 880 Slots That Ship More Than One Version

Book of Dead is one of the most played slots ever made. Ask what its RTP is and you will get 96.21%, because that is the number on every review page including the ones that never checked. But Play'n GO also ships Book of Dead at 84.18%. Same reels, same art, same paytable, same Rich Wilde walking across the same tomb. Twelve points of return, gone, and nothing on the screen tells you which one you are spinning.

This is not a loophole and it is not a scandal. Providers build slots in multiple RTP configurations and let each operator pick which one to license. It is a documented, regulated, entirely legal part of how the industry works. What it is not, is visible. So we went through our catalogue to find out how widespread it actually is.

Across 36,575 active games we hold RTP data on, 880 of them ship in more than one RTP configuration, spanning 245 different providers. The average gap between a game's lowest and highest version is 3.44 points. For 110 of those games, the gap is 8 points or more.

What We Checked, And What Counts As A Multi-RTP Game

We store a minimum and maximum published RTP per game, separately from the headline figure. A game counts as multi-RTP here when its published low and high configurations differ by more than 0.01 points. That gives us 886 games before cleaning.

Six of those we threw out. Three records carried a minimum RTP of 0, which produced nonsense spreads of 96 to 100 points, and the rest sat outside a credible band. No licensed slot ships a 0% configuration, so those are bad records on our side rather than findings about the games. We set a floor of 80% minimum and a ceiling of 99% maximum and excluded anything outside it. That leaves 880 games, and every number in this article comes from that cleaned set.

We are naming the exclusions rather than quietly dropping them, for the same reason we did it when we published our recorded session data. A dataset is only worth reading if you know what got left out of it.

One more caveat before the tables. These are published configurations from provider documentation, not measured returns. We are not claiming any specific casino is running any specific version. We are saying the versions exist and that a player has no reliable way to see which one they loaded.

The Games You Have Actually Played

Aggregate numbers are easy to skim past, so here are the games most people reading this will recognise. Every row is a real published low and high configuration for that title.

Game	Provider	Lowest Version	Highest Version	Gap
Book of Dead	Play'n GO	84.18%	96.21%	12.03
Legacy of Dead	Play'n GO	84.55%	96.58%	12.03
Rise of Olympus 1000	Play'n GO	84.24%	96.24%	12.00
Starburst Galaxy	NetEnt	88.00%	96.00%	8.00
Chaos Crew 2	Hacksaw Gaming	88.28%	96.27%	7.99
Le Bandit	Hacksaw Gaming	88.36%	96.34%	7.98
Wanted Dead or a Wild	Hacksaw Gaming	88.42%	96.38%	7.96
Gates of Olympus	Pragmatic Play	94.50%	96.50%	2.00
The Big Dog House	Pragmatic Play	94.48%	96.56%	2.08
Big Bass Football Bonanza	Pragmatic Play	94.50%	96.50%	2.00
Sweet Bonanza	Pragmatic Play	96.48%	96.51%	0.03

Sweet Bonanza is in that table on purpose. Its two configurations differ by 0.03 points, which is nothing, and that is worth seeing next to Book of Dead's twelve. "This game has multiple RTP versions" is not automatically a warning. The size of the gap is the whole story, and it varies enormously by studio.

Which Studios Do This Most

Restricting to providers with at least eight multi-RTP games in our data, the pattern splits cleanly into two different philosophies.

Provider	Multi-RTP Games	Share Of Their Games With RTP Ranges	Average Gap	Lowest Version Seen
Play'n GO	12	92%	12.01	84.18%
Hacksaw Gaming	42	100%	7.74	86.10%
SYNOT Games	9	82%	7.36	85.08%
Swintt	10	77%	7.32	85.41%
Kalamba Games	12	100%	7.16	86.48%
Spinomenal	14	67%	6.48	88.70%
NetEnt	22	81%	6.23	88.00%
Nolimit City	8	89%	4.04	92.01%
4ThePlayer	40	70%	3.21	86.00%
Rival	62	39%	1.53	92.20%
Pragmatic Play	47	85%	1.39	94.08%
1spin4win	18	11%	1.66	91.10%

Pragmatic Play offers alternative configurations on 85% of the games where we hold a range, but the average gap is 1.39 points and the floor is 94.08%. Play'n GO offers them on a similar share and the average gap is 12.01 points with a floor of 84.18%. Both studios are doing the same thing structurally. Only one of them has a version that changes what the game costs to play in any meaningful way.

Hacksaw is the one to watch if you play high volatility. Every Hacksaw title in our data where we hold a range has more than one configuration, averaging a 7.74 point gap, and their low versions sit around 88%. Those are games people already play for the max win rather than the return, which makes the downgrade easier to miss.

The Downgrade Ladder

Low versions are not scattered randomly. They cluster hard on a few standard rungs, which is what you would expect from a configuration menu rather than per-title tuning.

Low Version Band	Games
94%	323
96%	131
92%	110
95%	72
88%	56
90%	45
86%	38
84%	17

The bulk sit at 94%, one standard rung below the familiar 96%. That is the version most players are most likely to encounter without noticing, because 94% still looks like a normal slot RTP if you happen to check. Of the 880 games, 225 have a low configuration under 92% and 77 have one under 88%.

What The Gap Actually Costs

RTP is a rate applied to everything you wager, not to what you deposit, so the cost of a downgrade scales with how long you play rather than how much you bring.

Take the average gap across the whole set, 3.44 points. On £500 of total wagering, which is a thousand spins at 50p, the lower version costs you about £17 in expected return against the higher one. Same game, same session length, same everything on screen.

Now take Book of Dead's twelve points. That same £500 of wagering costs about £60 more on the 84.18% build than on the 96.21% build. To be clear about what that number is and is not: it is expected value across a large number of spins, not a prediction about your session. As our own recorded sessions showed, two runs on an identical slot came in 24.6 points apart on observed return, which is far wider than most of the gaps in this article. Over one session, variance will drown the RTP difference completely. Over a year of playing the same game, it will not.

That is the honest framing. Playing a low RTP version will not make you lose tonight, and playing the high one will not make you win. It changes the long run cost of a habit, which is exactly the sort of thing that is worth ten seconds of checking and never worth chasing.

The Advertised Number Is Usually The Best One

Here is the part that makes this hard to check. Of the 867 multi-RTP games where we also hold a headline figure, the advertised RTP matches the highest available configuration in 508 cases. That is 59%. In another 259 cases it sits somewhere in the middle of the range, and in 122 it matches the lowest.

So the industry default, ours included until we started tracking ranges separately, is to publish the best number the game is capable of. Nobody is lying. 96.21% is a true statement about Book of Dead. It is just not necessarily a true statement about the Book of Dead sitting in the lobby you are looking at.

How To Check Which Version You Are Playing

There is no reliable trick and no tell in the gameplay. What there is:

• Open the game's own info panel. The paytable or rules screen inside the slot itself is the only source that describes the build you actually loaded. Most reputable providers state the RTP there. If the number in the game does not match the number on the casino's page, trust the one inside the game.
• Check the operator's game info page, not the review site. Some regulated operators, particularly under UK and Nordic licences, publish the configured RTP per title. That is the operator committing to a number rather than a provider stating a maximum.
• Treat a missing RTP as a data point. If a casino runs games from studios on the list above and does not state RTP anywhere in the game or the lobby, you have no way to know which rung you are on.
• Check per operator, not once. The configuration is chosen by the operator, so the same title can genuinely differ between two casinos you hold accounts with. Checking Book of Dead at one casino tells you nothing about Book of Dead at another.
• Weight your check toward the studios with real gaps. A 0.03 point difference on Sweet Bonanza is not worth your time. A 12 point difference on a Play'n GO book slot you play every week is.

What We Are Changing On Our Side

Publishing a single RTP number for a game that ships in five configurations is not wrong, but it is incomplete in a way that favours the game. We hold minimum and maximum figures on 1,986 titles and until now we mostly surfaced the headline. Where a game has a meaningful spread, showing the range is the more honest presentation, and that is the direction we are moving.

None of this is a reason to play more, to switch casinos chasing a percentage point, or to treat a high RTP build as an edge. It is not one. Every configuration in every table above still returns less than it takes over time, which is the entire point of the business. Set what you are willing to spend before you open the lobby, and treat RTP as information about cost rather than a route to a win.

More on how we source and verify what we publish is on our E-E-A-T and methodology page, and the long form version of the return maths is in our RTP guide. Catalogue figures in this article were pulled on 8 August 2026 and reflect published provider configurations at that date.