Search for a popular slot's RTP and you'll often find two or three different answers. Usually none of them is a mistake. Many studios release the same game in several versions, each with its own return to player figure. The artwork, symbols and features match across versions. Only the maths underneath changes.
This article explains how that works, using studio listings and the rules that apply in Great Britain.
What a studio listing actually shows
Light & Wonder's iGaming site gives a clear example. Its listing for Huff N' Even More Puff Grand has an "Available Versions" section with five entries, each with its own game ID:
| Game ID | RTP |
|---|---|
| 71941 | 96% |
| 71940 | 94% |
| 71939 | 92% |
| 71938 | 90% |
| 71937 | 87% |
These aren't settings the player can change. Each is a separate build of the game. The same site lists only one version for Huff N' Puff CollectR, at 94%, so a studio doesn't always publish multiple versions for a title.
Our catalogue records show the same pattern elsewhere:
- Eye of Horus: 96.31%, 94.83%, 92.44%, 90.25% and 88.26%
- Big Bass Splash: 96.71%, 95.67% and 94.60%
- Pirots 5: 96% down to 87% in the same five steps as the Light & Wonder listing above
Why the screen looks identical
RTP is a long-run average set by the game's maths, such as how often each symbol appears and how often features trigger. Changing it doesn't need a new paytable or new art. It changes how often things happen over a very large number of rounds.
The gap is real over time. For every 100 units staked, the 96% build of Huff N' Even More Puff Grand is designed to return 96 over the long run, and the 87% build 87. Across a few dozen spins, that difference is hidden by normal variation. Over hundreds of thousands, it isn't. Our guide to RTP and volatility explains why short sessions can't show which version you're playing.
Studio listings don't say why each version exists. What they do show is that the choice is made before a game reaches players, by picking a build.
What regulators require
In Great Britain, the Gambling Commission's remote technical standards cover this. RTS 3 requires that information enabling an informed decision about the chances of winning be easily available before the customer commits to gamble, and it names the RTP percentage, house edge or probability of winning as ways to meet that. For games with a skill element, the guidance says RTP should be calculated using either the auto-play strategy or a standard or published strategy.
The Commission's testing strategy covers the numbers themselves. Testing includes checking the game design, meaning the maths, the rules shown to players and the theoretical RTP. It also verifies the actual RTP through simulation, emulation and manual testing. The Commission keeps a list of approved test houses for this work. So each version's figure is meant to be checked against how the game really behaves, not just stated.
Not every second RTP figure is a separate version
Some RTP figures look like separate versions but are something else:
- Feature or mode RTPs. Lucky Pharaoh is listed at 94% for its base game and 96.25% for its Power Spins feature. That is two parts of the same game, not two builds.
- Bet-type RTPs. Table games return different amounts depending on the bet. Our Baccarat record lists 98.94% on Banker, 98.76% on Player, 85.64% on Tie and 89.64% on the Pair side bet.
- Strategy-dependent RTPs. Blackjack returns depend on how hands are played, which is why the regulator's guidance refers to a standard or published strategy.
Mixing these up is the most common reason RTP figures seem to conflict.
How to find the figure for the version you're playing
The game's own information or help screen is the place to check. That's where the standards expect this information to be. On SlotsRanker, the facts panel shows our recorded figure and its source. When we know of lower versions, the game page lists them under "Some sites run this game at a lower RTP setting". If the default can't be confirmed, the card shows a dash instead of the highest figure available.
None of this changes free play here: every version runs on virtual coins. It does explain why the same game can be quoted with different figures, and why the version number matters when comparing them. For more on reading those screens, see how to read a paytable.
