Explainer

Slot RTP explained: theoretical vs. actual return

By Slots USA editorial desk Published Updated

Three illuminated slot machines with empty chairs
A context photograph of slot machines. This image does not document the probabilities or performance discussed in the article. Photo: Darya Sannikova / Pexels

The short answer

Theoretical RTP describes a probability model over repeated outcomes. It does not specify what any one session will return. Observed return is a measurement of the outcomes that actually occurred.

In this article

A percentage looks precise even when its meaning is unclear. A theoretical return of 90%, for example, is a statement about a model. It is not an instruction to return 90 units after every 100 one-unit events. To see the difference, start with a deliberately small model whose arithmetic is visible.

Build the average from outcomes

Illustrative example

A hypothetical distribution
Return per one-unit eventProbabilityContribution to the mean
0 units70%0.00 units
1 unit20%0.20 units
7 units10%0.70 units
Original teaching example, not the probability table of a real game. Probabilities sum to 100%; the expected return is 0.90 units.

Multiply each possible return by its probability and add the results: 0 × 0.70 + 1 × 0.20 + 7 × 0.10 = 0.90. This is the expected-value calculation described by OpenStax. In this example, dividing the expected return by the one-unit input gives a theoretical return of 90%.

The observed number answers another question

Suppose a ten-event sample contains seven zeros, two one-unit returns, and one seven-unit return. It returns nine units from ten units of input, so the observed return happens to be 90%. Another ten-event sample could contain ten zeros. The second sample is different; the model does not adjust the next event to make the first ten look closer to the average.

For an observed figure, the denominator matters. Total amounts returned divided by total amounts staked is a return ratio; final account balance divided by an opening balance is something else. Deposits, withdrawals, and repeated use of previously returned amounts make those two calculations diverge.

What to check beside a percentage

  • Whether the figure is theoretical or observed.
  • Which version and rules the figure describes.
  • For observed data, the measurement period and total input.
  • Whether included features and payouts match the stated scope.

Sources and further reading