Return to player, or RTP, is the theoretical percentage of total stakes that a casino game is designed to return as prizes over a large number of plays. It is a long-run property, not a promise for one session.
This explainer is for operators, suppliers, analysts, journalists and players who need to read RTP claims accurately. It uses British regulatory material as a worked example, while recognising that display, testing and minimum-return rules vary by jurisdiction and product. It is not gambling or financial advice.
What RTP means in online casino games
RTP expresses the return side of a game's mathematical model. If a game has a theoretical RTP of 96%, the model allocates 96 units of prizes for every 100 units staked across the relevant long-run calculation. The remaining 4% is the corresponding theoretical house edge, provided the same rules and assumptions are used.
The denominator is turnover, not the player's initial deposit. If the same funds are staked repeatedly, every wager contributes to turnover. This distinction matters because a 100-unit deposit can generate much more than 100 units of total stakes before the session ends. RTP therefore cannot be applied once to the opening balance as a forecast.
The official Gambling Commission player guide describes percentage RTP as an average reached over a significant number of game plays. It explicitly warns that a displayed percentage should not be read as the amount a person will receive back from each unit staked during one session.
Theoretical RTP and actual RTP are not the same measure
| Role | What it means | Where to verify |
|---|---|---|
| Theoretical RTP | The return built into the approved game maths for a defined version, ruleset and strategy assumption. | Game rules, paytable, maths documentation, market configuration and test report. |
| Actual RTP | The return observed from live wins divided by live turnover over a stated measurement period. | Win, turnover, number of plays, period, game version, volatility and acceptable tolerance. |
Theoretical RTP is set by the game design. Actual RTP is an operational result. The Gambling Commission's terminology for live monitoring says theoretical RTP is the designed and advertised figure, while actual RTP is calculated from the win and turnover produced by the operational game.
A gap between them is not automatically evidence of a defect. Random results fluctuate, and a volatile game can require a larger sample before the measured return sits close to the model. A proper review considers the game's standard deviation, play count and the testing tolerance, rather than comparing two percentages without context.
How actual RTP is calculated
The basic operational formula is actual RTP equals total wins divided by total turnover, multiplied by 100. In this calculation, turnover includes every stake made on the game, including stakes funded with previous winnings. Wins are the prizes awarded during the same measurement window.
For example, 950,000 units in recorded wins divided by 1,000,000 units in turnover produces an actual RTP of 95%. That result describes the chosen dataset only. It does not establish the theoretical RTP and cannot, by itself, prove that the game is working correctly or incorrectly.
The Commission's operator calculation guide explains that volatility informs the acceptable distance above or below the designed return. It also notes that the tolerance narrows as the volume of play grows. Sequential failures or repeated out-of-range results deserve investigation, but one measurement failure does not automatically confirm a faulty game or RNG.
RTP and house edge
RTP and house edge usually describe complementary parts of the same model. The conversion is house edge equals 100% minus RTP. A 96% RTP corresponds to a 4% house edge. The reverse calculation is RTP equals 100% minus house edge.
The relationship is useful only when the inputs share the same basis. A table game's edge may depend on the selected bet, paytable, rules or player strategy. The published RTP for a game with a skill element may also rely on a specified strategy. Comparing figures without those assumptions can create a false ranking.
Expected theoretical cost can be illustrated as turnover multiplied by house edge. At 1,000 units of turnover and a 4% edge, the modelled cost is 40 units. This is an expectation across repeated play, not a bill that arrives at the end of every session. Actual results can vary widely on either side.
RTP, volatility and hit frequency answer different questions
RTP describes the share of turnover allocated to prizes in the long run. Volatility describes how widely results can vary around the expected return. Hit frequency describes how often a defined winning event occurs. A game can have a high theoretical RTP and still deliver long losing stretches if its prizes are concentrated in rare outcomes.
Maximum win is separate again. It states an upper prize limit under defined conditions, not the probability of reaching that prize. Game speed also changes how quickly turnover accumulates. Two games with the same RTP may expose a player to very different distributions and amounts of turnover during the same period.
The Commission's live RTP terminology describes volatility using standard deviation and notes that a highly volatile game can contain very large but rare prizes. That is why RTP should be read with the paytable, rules and risk profile, rather than treated as a complete quality score.
RTP, RNG and game mapping
RTP is part of the game maths; an online casino RNG is the mechanism used to produce random values for relevant games. Mapping logic turns those values into game outcomes. These components interact, but they are not interchangeable. An RNG can pass statistical tests while an incorrect paytable or mapping implementation produces the wrong return.
A complete assurance process therefore examines more than the number generator. It links RNG documentation, source code, mapping, game rules, paytable, theoretical calculations and the production build. The British testing strategy sets out pre-release testing and further controls for relevant remote products, while RTS 7 addresses the generation and implementation of random outcomes.
The organisation responsible for that work may be a casino game provider, an approved test house, a platform supplier or the licensed operator. The evidence record should name each party's role rather than assuming that a provider logo proves what was tested or deployed.
What a published RTP figure should identify
A bare percentage is weak evidence. A useful disclosure connects the number to a specific product and makes the rules easy to reach before play. In Great Britain, RTS 3 requires information that enables an informed decision about the chance of winning, which may include house edge, RTP or probability, depending on the product.
- The exact game title and supplier, including a version or game identifier where available.
- The rules, paytable, enabled features and strategy assumption used in the calculation.
- The theoretical RTP shown to the customer in the relevant market and currency context.
- The date, test laboratory, report reference and scope of any supporting assessment.
- Any configuration difference between the tested build and the live production build.
The licence context also matters. A licence does not create one worldwide RTP rule. Check the gambling licence and the market-facing legal entity, then confirm the regulator's own display and testing requirements. Our separate guide explains how to verify an online gambling licence without relying on a footer badge alone.
Common RTP mistakes
- Treating 96% RTP as a promise that every 100-unit session returns 96 units.
- Applying RTP to the opening deposit instead of the total amount staked.
- Assuming a higher RTP guarantees a better short session result.
- Using RTP as a substitute for volatility, hit frequency or maximum-win probability.
- Comparing different rule sets or strategy assumptions as if the percentages were equivalent.
- Citing a test badge without matching its report, version and jurisdiction to the live game.
These mistakes also distort comparisons between an operator and its technology suppliers. The iGaming platform can deliver games and account services, but the exact contractual and technical allocation varies. A sound article separates the operator, provider, platform, laboratory and regulator before assigning responsibility.
Frequently asked questions
Does 96% RTP mean I will get 96 back from 100?
No. A 96% theoretical RTP describes the game's long-run mathematical return across a large amount of play. It does not promise a 96-unit return from one 100-unit session. Your result may be zero, below 96 or above 100 because short samples are dominated by random variation and volatility.
Is a higher RTP always better?
A higher RTP means a lower theoretical house edge when the rules and calculation basis are comparable. It does not make a short session safer or predict a win. Volatility, game speed, bet size, paytable, strategy and total turnover can matter greatly to the experience and the amount exposed over time.
Can the same game have different RTP figures?
A game may exist in different approved versions, configurations or market implementations, so the title alone is not enough to verify a figure. Check the live rules and paytable for the exact game instance. Then match its identifier and version to the relevant laboratory or regulator record where one is available.
What is the difference between RTP and volatility?
RTP describes the long-run share of stakes returned as prizes. Volatility describes the distribution and variability of those returns. Two games can share the same RTP while one pays smaller prizes more regularly and the other concentrates more value in rare outcomes. Neither measure predicts an individual session result.
How can I verify an RTP claim?
Open the game's own rules or help screen, record the displayed RTP and capture the game version. Check the relevant regulator's requirements and any test report for the same build. Confirm that the operator, market and licence match. A generic certificate or provider-wide percentage is not enough evidence for one live game.
The BETTIMES RTP research rule
BETTIMES records RTP as a versioned game fact, not a marketing adjective. The minimum record connects the game, supplier, operator, market, version, rules, displayed percentage, source URL, verification date and any test reference. Theoretical and actual RTP remain separate fields, with sample size and volatility attached to operational measurements.
This approach reflects the wider structure described in our guide to what iGaming is. It prevents a mathematical claim from floating free of the entity and production context that give it meaning. Where the version or test scope cannot be matched, the result is recorded as unresolved rather than inferred.
Why this record matters
BETTIMES links this publication to structured company, market and source records. Material changes can therefore be checked and refreshed without detaching the article from its original evidence.