What does live RTP mean?
The term usually refers to an observed payout ratio over a recent window of completed bets. Observed RTP = total payouts ÷ total wagers × 100. A dashboard may update that ratio frequently, but its update speed does not change the definition or make it a forecast.
Theoretical RTP is the expectation implied by a game’s model and rules. A recent sample can sit above or below it because outcomes vary. If a dashboard uses “live RTP” for a different measure, it should explain the inputs and calculation. Without that definition, apparently comparable percentages may describe different things.
A measure limited to one operator, game mode or set of reporting players also has limited coverage. It is not automatically the return of the whole market or the version available in your account.
Reference: UKGC: monitoring theoretical and actual RTP. The calculations in this section are illustrative unless a named product is explicitly identified.
Calculate a recent sample without averaging percentages
The correct denominator is money wagered in the included completed bets. Deposits, withdrawal amounts or a count of spins cannot replace it. For different currencies, a consistent documented unit is needed. See weighted RTP calculations for the general method.
Does hot or cold RTP predict the next result?
For independent rounds from an unchanged model, previous outcomes do not alter the next round’s probabilities. A hot display can be caused by a rare award in a small sample. A cold display can be caused by that award not appearing. Neither establishes a repayment schedule.
The probability question changes only if there is a relevant disclosed state or rule. A progressive jackpot amount or a collection feature may require state-specific analysis; a colour assigned to a recent payout sample does not provide that analysis.
The UKGC’s random-outcome standard addresses adaptive behaviour within its scope. That is a requirement for covered products, not evidence that every unverified dashboard accurately describes a game. Check the rules and version separately.
Reference: UKGC RTS 7: random outcomes and adaptive behaviour. The calculations in this section are illustrative unless a named product is explicitly identified.
For the reasoning behind hot and cold claims, read why recent winners do not identify an opportunity.
What data must a credible RTP dashboard disclose?
- The exact game, version, selected mode and included operators.
- Matching monetary stakes and total payouts for completed rounds.
- The sample window, time zone, end date and treatment of unsettled bets.
- Sample size and enough payout information to discuss uncertainty.
- Currency units, conversion method and whether promotional credits are excluded.
- Whether the dataset is complete or selected from volunteers, winners or public clips.
Turnover, player accounts, bet counts and wallet flows can be useful activity measures. Without matching payout totals they do not produce observed game RTP. A large dataset can still be biased if it excludes losing sessions or mixes incompatible modes.
What if a dashboard shows 110%?
That can be a valid description of a completed sample: $11,000 paid out on $10,000 wagered gives 110% observed RTP. It is not proof that theoretical RTP has become 110%, nor that the next customer has a positive expectation.
Two samples with the same observed percentage can carry very different uncertainty. A hundred high-variance bets and a million well-documented bets are not interchangeable evidence. There is also a difference between a random sample and one deliberately selected after a large win.
Use the sample-size guide to understand the assumptions behind a confidence interval. A precise-looking percentage with several decimal places does not make the underlying evidence precise.
A worked calculation shows how a confidence interval depends on the payout model.
How should readers use recent activity figures?
Use them for the claim they actually support: the reported historical activity of a defined dataset. Keep a studio’s published theoretical return, a casino’s active setting and a dashboard’s measured result in separate fields. If one is missing, label it unknown instead of deriving it from another.
Our activity rankings show interest through the published metrics they identify. Popularity is not a payout forecast. The practical next step for a particular game is checking its active rules, not increasing a stake because a recent percentage is unusually high.
Common questions
Can live RTP tell me which slot will pay next?
No. A historical payout sample does not establish better odds for the next independent round. A disclosed stateful rule needs its own analysis.
See the explanation: Does hot or cold RTP predict the next result? →
Explore the explanation: Why hot and cold results do not identify the next winner →
Can I calculate RTP from bet volume and player counts?
No. Observed RTP also requires matching total payouts. Bet counts and turnover alone are activity data.
See the explanation: What data must a credible RTP dashboard disclose? →
Is 110% live RTP evidence of a profitable game?
It can describe a winning historical sample, but it does not establish a theoretical return above 100% or a positive expectation for future play.
How do I combine live RTP from different samples?
Add compatible monetary payouts and stakes, then divide. Do not take an unweighted average of percentages from different turnover amounts.
See the explanation: Calculate a recent sample without averaging percentages →
Sources & further reading
Primary sources checked on 17 November 2026. The examples above use the stated rules and assumptions. Product pages describe published versions; they do not verify every operator’s configuration. Regulatory guidance applies within its stated scope.