What RTP tells you. And what it cannot.

A model average, not a personal balance

Theoretical RTP describes the expected proportion of total stakes returned as awards under a defined game model. “Total stakes” means turnover: money can be wagered repeatedly, so this is different from an initial deposit.

The UK Gambling Commission distinguishes the long-run concept from a short playing session. Its material is used here to explain a concept, not to imply that UK technical requirements apply to every product or to Cambodia.

A hypothetical 95% example

Suppose an invented game model has a theoretical RTP of 95%. Applied to 100,000 units of turnover, the model’s expected awards are 95,000 units. That arithmetic does not say a player depositing 100 units will leave with 95.

For that simplified fixed-stake model, 100% − 95% = 5% is the complementary theoretical house edge. It is not a fee deducted from each individual round. Actual awards can be above or below the model average.

Observed return answers a different question

An observed payout ratio is total recorded awards divided by total recorded stakes, multiplied by 100. It describes the sample measured. Without enough context, it does not establish the theoretical configuration.

Volatility and sample size affect how widely measured results can vary. A short run with a low return does not mean the next round is “due” to compensate. For independent random rounds, previous wins or losses do not change the next round’s probabilities.

Attach a figure to its version

A provider cannot be assigned one RTP across its whole catalogue. We list no real game percentage here because a numerical claim needs a specific title, configuration and source.

• Record the exact title and edition.
• Read the published help screen or technical document for that configuration.
• Check whether a stated value includes the relevant features or jackpot component.
• Leave the figure unconfirmed if the source or version is missing.