What to check
Payout determines the return when it occurs. A useful comparison combines the event, its probability, its return and the cost using the same units.
Read four different measures
Take standard single-zero roulette. A straight-up selection has one winning pocket out of 37. At a total return of 36 units on a one-unit stake for the winning pocket, its expected total return is 36 divided by 37, about 97.3%.

| Measure | What it describes |
|---|---|
| Probability | The chance of a specified event under stated rules |
| Expected return | The average total credit for a stake in the model |
| House edge | The expected share of the stake retained under those same rules |
| Volatility | The pattern of variation around the return |
Read the full explanation
A different wheel or rule changes that calculation.
For an illustrative side selection that pays eight units total once in ten, expected return is 0.8 units per unit staked. A slot with 96% modelled return uses a different distribution and may have longer gaps or fewer large credits. Use the matching rules and stake. A return percentage alone cannot tell you how often a credit appears.
How to use the record
In a toy game, a one-in-four chance of an RM8 total return on an RM2 stake has RM2 expected return before any other cost. An RM8 profit would be different because it excludes the original stake. Real games have additional outcomes and rules, so confirm the edition and keep assumptions alongside any calculation.
