What RTP really means

Think of RTP as a property of the maths model, not of any one player. The studio designs the paytable, the reel strips and the features so that across millions of spins the game returns, say, 96.21% of what goes in. Some players walk away up. Most walk away down. The average lands where the model says.

Short sessions are dominated by luck. One free spins round can return 200x your stake; a hundred dead spins can return nothing. Neither tells you the RTP. I'd go further: if you ever catch yourself thinking a slot is "due" because it has paid less than its RTP lately, stop. Spins are independent. The game has no memory of what it owes anyone.

And every slot has a negative expected value. A 98% game still keeps 2% of turnover on average. A higher RTP means a slower loss, not a win.

Most big studios sell their games with a menu of RTP versions. The operator chooses one, often to fit local tax rules or its own margin. Nothing in the game's art changes. Only the maths does.

Here's what the 105 slots in our database look like:

• 65 of 105 list more than one RTP setting in the sources we checked.
• 28 list two settings, 13 list three, 9 list four and 14 list five. Dead or Alive has six, per SlotCatalog.
• 24 have a lowest listed setting under 90%, and 14 go below 85%.
• The lowest setting anywhere in our data is 84.01%, on Nolimit City's Mental 2.
• Across the 65, the median gap between the top and bottom setting is 5.9 percentage points.

Book of Dead is the textbook case. Bigwinboard lists 96.21%, 94.25%, 91.25%, 87.25% and 84.18%. Same book, same explorer, same ten free spins. Pragmatic Play's Gates of Olympus comes in 96.50%, 95.50% and 94.50%. Relax Gaming's own page for Money Train 3 lists a German RTP of 90% next to an alternative 94%.

The habit varies a lot by studio. In our data Play'n GO lists lower settings on 13 of its 14 games, Push Gaming on 10 of 10, Nolimit City on 11 of 12 and Hacksaw Gaming on 10 of 12. NetEnt lists them on 2 of 12, and Red Tiger and ELK Studios on none. Don't read too much into the zeros. Red Tiger publishes one RTP per game, and our notes record that operators may still run lower settings. A single published figure only means a single published figure.

One more wrinkle. Some games change RTP with the mode you play. Bigwinboard gives Hacksaw's Chaos Crew bonus buy a 95.92% RTP, below the 96.30% base game. Nolimit City's Tombstone RIP goes the other way, with its 3,000x Boothill buy at 97.03%. Starburst XXXtreme is 96.45% with XXXtreme Spins and 96.26% without, per SlotCatalog.

This takes thirty seconds and it's the single most useful habit on this page.

1. Load the game at the casino you actually use, not on a review site or the studio's demo.
2. Open the menu or the "i" button and find the game rules or help pages (not just the paytable).
3. Scroll to the bottom. Look for "RTP", "theoretical return to player" or "expected payback".
4. If the figure there is lower than the headline, that's the version you'd be playing.

Push Gaming says outright that its RTP may vary per casino and points players to the loading screen. For Play'n GO, which publishes several settings and lets the casino pick, our records carry the same advice: check the info screen before playing. If a game shows no RTP anywhere, I'd treat that as an answer in itself and pick another game.

Two things the info screen won't tell you: how volatile the game feels, and whether the bonus buy runs at a different RTP. The second one is sometimes in the rules too, so keep reading past the first number you see.

House edge is the flip side of RTP: 100% minus the RTP. Book of Dead at 96.21% has a 3.79% edge. At 84.18% the edge is 15.82%.

The edge applies to turnover, not to your deposit, and that's the bit people miss. Every spin is a new stake, including spins paid for with money you just won. So the useful sum is: total amount staked times the edge.

Take a made-up session of 600 spins at €1. That's €600 staked. At the 96.21% setting the expected cost is €22.74. At the 84.18% setting it's €94.92. Same game, same stake, same number of spins, and the expected cost is more than four times higher. Expected cost is still an average; any real session can land well above or below it.

Put another way: if you stake €1,000 in total over a month, the top setting keeps about €37.90 of it on average, the bottom one about €158.20. The art on the screen is identical.

Check the setting before you play, every time, at every casino. If you find a lower version, play something else or play somewhere else. That's it, really.

Use RTP to compare games, not to predict outcomes. Two slots at 96.5% and 94.5% aren't far apart in a single evening. Over a year of regular play the gap adds up.

And don't let a high RTP talk you into a bigger budget. The highest RTP slots still lose money over time. Our Spin Score gives RTP 40% of the weight and uses the highest setting the studio publishes, which is exactly why we keep telling you to check the one in front of you.