RTP — return-to-player — is the long-run percentage of total wagers a casino game is designed to pay back to players. A slot with 96% RTP is designed to pay back €96 for every €100 wagered, averaged over an extremely large number of spins. RTP is the most commonly misunderstood number in casino content. This page is the working explanation: what RTP is, what it isn't, and how to use it to your advantage.
Author: Marijan Karajanov. Last updated: 20 May 2026. Scoring detail: /methodology#games.
RTP is the inverse of the house edge. A 96% RTP slot has a 4% house edge: over millions of spins, the casino keeps an average of 4% of all stakes. Over your individual session, the actual return can be anywhere from 0% to thousands of percent. RTP is a long-run number, not a session-level prediction.
RTP is calculated and audited by independent labs (eCOGRA, iTech Labs, GLI, BMM Testlabs) for licensed providers. The certificate covers the game's mathematical design as built. Audits do not cover what happens after the game is integrated into an operator's lobby — which is where one of the biggest RTP pitfalls lives.
Many slot games are sold to operators with multiple RTP variants. A single game title might ship with 96.5% / 95.5% / 94.0% / 92.5% / 88.5% / 84% versions, and the operator chooses which variant to install. The version installed at one casino can be substantially less player-friendly than the same title at another casino.
This is documented and well known inside the industry, but it isn't always visible to players. Examples of games known to ship in multiple RTP variants:
The version actually running at the operator is usually disclosed in the game's information panel — open the game, click the menu, look for "RTP" or "Information". If the operator runs an 88.5% version of a game whose default is 96.5%, you're playing a much worse game than the marketing implies.
The single biggest RTP improvement most players can make is to avoid operators that systematically install the lower-RTP variants of popular titles.
RTP tells you how much the game pays back on average. Volatility tells you how the payback is distributed across spins.
Two slots with identical 96% RTP can have wildly different player experiences. A low-volatility slot at 96% RTP might pay you 95% back in a session because the variance is small. A high-volatility slot at 96% RTP can pay you 0% in a session and 800% in the next session — averaged over a million sessions it reverts to 96%.
Players bothered by losses often prefer low-volatility slots even at modestly lower RTP. Players hunting for jackpots prefer high-volatility games even with the trade-off that most sessions return less than 50%. Neither is wrong; it's a preference about which side of the variance you experience.
Table games have published RTPs that reflect the optimal strategy. Specifically:
Realised RTP on table games depends on the strategy you actually play. Hitting blackjack 18 against a dealer 6 (a strategy mistake) costs you several percentage points of expected return that the published RTP assumes you don't make.
RTP is calculated over millions of spins. A single session is a tiny sample of that distribution. Your session outcome is dominated by variance, not RTP. Two players on the same 96% RTP game can finish the same session at +200% and −100% — both outcomes are perfectly consistent with the underlying RTP.
Every spin is independent. A slot that hasn't paid anything for 100 spins is not more likely to pay on spin 101 than on spin 1. The "due to pay" framing is the gambler's fallacy and is the single most expensive misconception players carry into slot play.
RTP doesn't measure fairness, only the long-run mathematical edge. A game with proven provably-fair cryptography and verifiable RNG audit certificates is "fair" in the sense that the outcomes match the published distribution. RTP tells you the expected loss rate; fairness tells you whether the rate is what was promised.
Game library and provider integrity (10% of our overall score) reflects:
The full scoring rubric is at /methodology#games. Operators that run lower-RTP variants score worse, regardless of how many games they list.
The Kansspelautoriteit framework requires operators to publish RTPs and to make the information accessible to players. In practice KSA-licensed operators are less likely to run the lowest-RTP variants of popular slots than offshore-licensed operators — partly because the regulator's audit framework reviews variant selection, partly because Dutch players are reportedly more RTP-sensitive than average European players. The RTP information in the game info panel at a KSA-licensed operator should match the regulator-reviewed setup.
Provably-fair games at crypto-native casinos use cryptographic commitments that let players verify the outcome of each round independently. The RTP of these games is still set by the operator and remains a published probability distribution; the "provably" part means you can verify that the published RTP was respected on the spin you played, not that the RTP itself is somehow higher or lower. Provably-fair is a guarantee of execution integrity, not a guarantee of player advantage.