Slot RTP: what does it mean, and can the number be trusted?
RTP is the share of stakes a game pays back over the very long run. We measured it ourselves: for 49 of 54 games, the official figure matches our measurement. But RTP does not apply to one session: of the real 500-spin sessions of Gates of Olympus we played, 67% ended in a loss.
By the OddsLab lab · data as of 2026-10-08 04:07 UTC
- Games whose official RTP matches our measurement
- 49/ 54
- 500-spin sessions of Gates of Olympus that ended in a loss
- 67%
- How blurry a one-hour RTP is
- ±32points
The official figure lies inside our 95% margin of error. Excludes two providers whose demos run far looser.
From 316 real sessions in our lab. The typical session returned 79%.
The Dog House Megaways, 4,896 spins an hour in our lab. An hourly figure misses by this much.
What is RTP?
RTP (return to player) is the share of all stakes a game pays back to players, counted over the very long run. An RTP of 96.5% means: of every 100 staked by all players, 96.5 comes back as wins on average and 3.5 goes to the casino.
The key words are "on average" and "very long run". RTP says nothing about what will happen to you today, and it does not change from hour to hour. It is a property of the game set when the game is built, like the number of cards in a deck.
Is the official RTP true?
Most sites just copy the RTP from the provider. We measured it: our lab spins the official demos around the clock and has recorded 6,966,088 spins. For 93 games that have each passed 20,000 spins, we compared the measurement with the official figure. The chart below pools every game by provider.
Our measurement minus the official RTP, by provider (percentage points)
See the numbers behind this chart
| Provider | Games | Spins | Official RTP | Our measurement | Difference (points) |
|---|---|---|---|---|---|
| NextSpin | 13 | 1,520,123 | 96.94% | 96.89% ±1.00 | −0.05 |
| Spadegaming | 14 | 397,568 | 96.83% | 97.15% ±2.15 | +0.32 |
| Pragmatic Play | 14 | 3,160,920 | 96.53% | 97.03% ±1.54 | +0.50 |
| Funky Games | 13 | 372,680 | 96.00% | 96.78% ±2.69 | +0.78 |
| FA CHAI | 13 | 324,000 | 96.40% | 134.09% ±3.31 | +37.69 |
| PG Soft | 14 | 457,008 | 96.77% | 142.38% ±2.81 | +45.60 |
Game by game, the official figure is inside our margin for 49 of 54 games (91%). If every official figure were right, about 95% should be, because a 95% margin misses now and then by design. So for these providers: the official RTP matches what we measure.
FA CHAI and PG Soft are a big exception: their demos pay back far more than 100%. That does not prove their official RTP is wrong (we cannot measure the real-money version), but it does show their demo is not the same machine. We cover that in a separate guide. Demo slot gacor guide →
One honest note: for Joker King, Gems of Zeus, our measurement is far below the official figure, too far to be chance. Most likely we are not yet capturing every kind of win in that game, so it is left out here.
Every game's official RTP and our measurement, with its margin, are in the RTP table →
Why doesn't your session feel like 96%?
Gates of Olympus's RTP is 96.5%. But how much actually comes back in one session? We did not guess with a formula. We took the spins our lab really played, in order, and cut them into "sessions" of 100, 500, 1,000 and 2,000 spins, exactly like a player sitting down and playing that many.
Real sessions: share of stakes returned, by session length
See the numbers behind this chart
| Session | Sessions | Bottom 5% | Typical session | Top 5% | Ended in a loss |
|---|---|---|---|---|---|
| Gates of Olympus · 100 spins | 1,584 | 28.1% | 60.5% | 265.0% | 74.1% |
| Gates of Olympus · 500 spins | 316 | 46.0% | 79.0% | 186.5% | 67.1% |
| Gates of Olympus · 1,000 spins | 158 | 56.1% | 84.2% | 156.2% | 67.1% |
| Gates of Olympus · 2,000 spins | 79 | 63.9% | 94.8% | 160.0% | 65.8% |
| Sweet Bonanza · 500 spins | 250 | 64.5% | 86.3% | 164.6% | 69.2% |
| Big Bass Bonanza · 500 spins | 461 | 46.6% | 91.1% | 155.4% | 61.2% |
| Sugar Rush · 500 spins | 235 | 51.2% | 81.2% | 173.0% | 66.4% |
Take 500-spin sessions: 67% ended in a loss, the typical session returned only 79%, and 90% ended between 46% and 187%. Yet the average over all those sessions still lands near the RTP, because the few sessions that win big cover the losses of many others. That is why "96% RTP" feels like a lie to most players: the average is carried by a handful of people. Only after about a million spins does the result settle near the RTP (±3.6 points). Nobody spins a million times.
Wilder (high-volatility) games have even wider ranges. Every game's volatility, as we measured it, is in the Volatility ranking →
Can a "live RTP" site know this hour's RTP?
Many sites show "live RTP" or "RTP today": hundreds of games with percentages that change every few minutes, 97.3% now and 84.1% an hour later. Figures like that cannot be measurements. We can show why with our own lab data.
If we worked out RTP from just one hour of lab spins
See the numbers behind this chart
| Game | Lab spins/hour | Official RTP | ± (points) |
|---|---|---|---|
| The Dog House Megaways | 4,896 | 96.55% | ±32.1 |
| Gates of Olympus Super Scatter | 3,672 | 96.50% | ±74.7 |
| Big Bass Bonanza | 3,672 | 96.71% | ±25.6 |
| Starlight Princess | 3,672 | 96.50% | ±49.1 |
| Starlight Princess 1000 | 3,672 | 96.50% | ±43.7 |
| Sugar Rush 1000 | 3,672 | 96.53% | ±55.1 |
| Wild West Gold | 3,672 | 96.51% | ±40.8 |
| Prosperity Dragon | 3,672 | 96.96% | ±34.5 |
Our lab spins The Dog House Megaways about 4,896 times an hour, day and night. If we computed "this hour's RTP" from those spins, it would miss by up to ±32 points through chance alone. "Live RTP" sites spin nothing and have no access to casino data. A one-decimal figure that changes every hour is not measured from anything.
And even if it could be measured, it would be useless: last hour's RTP does not affect the next spin. A game is not "hot" because it paid a lot in the last hour.
We tested the "hot game" idea over millions of spins: Is a slot ever due? →
Is a game's RTP the same at every casino?
Not always. Some providers build several RTP versions of the same game, and the casino chooses which to install. They look identical; only the number in the game info differs.
Habanero, for example: for the 19 Habanero games we measure, the provider offers versions from 86% to 98%. That is why we show Habanero's RTP as a range, not one number.
Pragmatic Play also offers several versions of many games. To know which one you are playing, open the game's info or rules at the casino you play at and read the RTP there. A figure you read on another site may not be the same version.
Does a higher RTP mean winning more often?
No. A higher RTP means the casino takes slightly less over the very long run; the difference between 96% and 97% is 1 of every 100 staked. Within one session, that difference drowns in luck.
What shapes your experience far more is volatility: how often a game pays small, and how far away the big wins are. Two games with the same RTP can feel completely different.
How can I check this myself?
- See every game's official RTP and our measurement side by side, with the margin of error. RTP table →
- Download the raw record of every spin and compute RTP yourself: total won divided by total staked. Verify the data →
- Read how we measure, including how the balance is reconciled every spin. Full methodology →
Frequently asked questions
What does RTP mean on a slot?
The share of stakes paid back to players over the very long run. An RTP of 96.5% means that of every 100 staked, 96.5 comes back on average.
Can the official RTP be trusted?
From our measurements, generally yes: for 49 of 54 games the official figure is inside our margin. The exception is two providers whose demos pay far more loosely.
Is "live RTP" or "RTP today" real?
No. RTP is a game setting and does not change by the hour. Even our lab, spinning thousands of times an hour, cannot read one hour's RTP accurately.
What is a good slot RTP?
Most modern slots are around 96%. But within one session a 1-point RTP difference drowns in luck; volatility shapes your experience more.
Why did I lose when the RTP is 96%?
Because RTP is the average over all players in the very long run, and that average is carried by a few sessions that win big. Of the real 500-spin sessions in our lab, 67% ended in a loss; 90% ended between 46% and 187%.
Is a game's RTP the same at every casino?
Not always. Some providers build several RTP versions and the casino chooses. Check the RTP in the game info at the casino you play at.
How do I find a game's RTP?
Open the info or rules menu inside the game. On this site every game has a page with its official RTP and our measurement.
The data behind this guide
18+ only. OddsLab is a measurement site: we take no bets, have no casino partners and link to no casino. There is no way to beat a slot machine over the long run.