RTP & Expected Loss Calculator

Calculate a stated scenario

The expected figures use the theoretical percentage you supply. Optional actual returns add a measurement of entered results; they do not replace the theoretical assumption or identify an underlying game setting.

Enter a complete stake, paid-round count and relevant percentage

1. Total stake per spin: enter the full paid stake in AUD for the scenario, not a line component, denomination or starting deposit. The simple model assumes the same stake on every counted paid spin.
2. Number of spins: enter a positive whole number of paid rounds using that stake and return assumption. Do not add every cascade animation as another charged round.
3. Theoretical RTP: enter the percentage relevant to the model, between zero and one hundred. The default example is not evidence that a particular game uses that percentage.
4. Actual total returns, optional: enter the gross game prizes for those same stakes. Leave this blank when there is no observed result; enter zero only when the results really contain zero gross prizes.

Use the whole stake to which the supplied percentage applies. If an option, paid feature or different mode changes the relevant return model, do not assume its price can be folded into ordinary spins under the same percentage. The help and applicable configuration must support that comparison.

RTS 2 addresses clear stake values and currency-to-credit conversions within Great Britain’s standards. For this calculation, a credit count needs its monetary value before it can serve as an AUD stake.

For a fictional conversion, twenty credits at A$0.025 each cost A$0.50. Entering twenty as the monetary stake would instead model A$20 per spin. The multiplication could be correct while describing the wrong amount.

Enter stake and actual-return money amounts with no more than two decimal places. The RTP percentage can carry more precision; the displayed money and percentage results round to two decimal places.

For another fictional input check, sixteen lines at A$0.03 each make a complete A$0.48 stake when there are no other charges. At two hundred and fifty paid spins, turnover is A$120. With an assumed 95% RTP, expected prizes are A$114 and expected gaming loss is A$6. Entering the A$0.03 line component instead would model just A$7.50 turnover, sixteen times smaller. The arithmetic would describe that smaller input correctly while answering the wrong financial question. This example does not assume that every actual game permits line selection; read its complete-price definition.

All money inputs must use the same unit. The calculator does not obtain an exchange quote or convert a foreign-currency history into AUD. Convert consistently using the applicable information before comparing amounts, rather than treating different dollar symbols as the same currency.

The expected calculation starts with turnover

Let the full stake per paid spin be S, the paid-spin count N and the supplied theoretical percentage R. Turnover is S × N. Expected gross prizes are turnover × R ÷ 100. Expected gaming loss is turnover × (100 − R) ÷ 100.

The displayed theoretical house edge is 100 − R, in percentage points. It is a rate in the entered model, not a fixed fee charged separately or a prediction that each round loses exactly that fraction.

The prefilled example uses A$0.50 per spin, one hundred spins and 96% theoretical RTP:

The theoretical edge is 4%. A$48 is the expected gross total returned as prizes across the defined workload, not A$48 profit on top of the A$50 staked. Subtracting turnover gives an expected net result of minus A$2.

The Commission’s consumer RTP explanation describes an average over many plays rather than a session refund. Our arithmetic applies an entered mean to a fixed workload. It does not establish that the next hundred rounds will return the average.

Keep the workload fixed when comparing assumptions. At A$200 turnover, 94% implies A$188 expected prizes and A$12 expected loss; 96% implies A$192 and A$8. The two-percentage-point difference changes expected prizes by A$4 in that model, not by two dollars regardless of turnover.

Actual returns add an observed calculation

The Commission’s monitoring guidance calculates observed return by dividing gross prizes by turnover. The calculator applies that ratio to your optional results: actual gross returns ÷ turnover × 100. Actual net result is gross returns minus turnover.

Keep the prefilled A$50 turnover and enter A$75 gross returns. Observed return is 75 ÷ 50 × 100 = 150%, and the actual net result is A$25 profit. An observed percentage above one hundred is valid arithmetic, not evidence that the game’s designed RTP is 150%.

Entering A$35 instead gives 70% observed return and a loss of A$15. Entering zero gives 0% and a loss of the full A$50 turnover. Leaving the optional field blank is different: there is no observed percentage or actual net result to calculate.

For a separate pooled record, suppose fifty A$0.20 spins have A$10 turnover and return A$15, while another four hundred and fifty at the same price have A$90 turnover and return A$45. The individual observed percentages are 150% and 50%. Combined, five hundred spins cost A$100 and return A$60, giving 60% observed return and a loss of A$40. Averaging the two percentages to get 100% would wrongly suggest break-even because it ignores the unequal turnover. Use the combined paid-spin count and gross prizes when entering this consistent record.

Gross prizes include an amount returned on a break-even action. If an action costs A$0.50 and returns A$0.50, its gross prize is A$0.50 and its net result is zero. Entering the zero net figure as a gross return would wrongly record a full-stake loss.

Check what a history column means before summing it. A label such as win can describe a component, a running total or a net amount depending on the record. Sum the appropriate gross prizes once, including the completed stages belonging to the same paid rounds, without adding their final summaries again.

These figures measure the entered sequence. They do not, by themselves, prove a fault or identify a hidden setting. For a result discrepancy, use the complete round history and the responsible service’s verified support route.

A deposit and a starting balance are not turnover

Turnover adds the stakes actually made. Prizes can be used again, so a smaller starting balance can support a larger cumulative total in a completed record. The deposit is not the denominator of observed return unless it happens to equal the relevant turnover.

Consider a fictional account starting with A$20. A first block of rounds costs A$10 and returns A$8, leaving A$18. A second costs A$10 and returns A$7, leaving A$15. Combined turnover is A$20, prizes are A$15 and game-only loss is A$5.

A further block costing A$10 and returning A$9 would leave A$14. Now total turnover is A$30 and gross prizes A$24, even though starting funds were A$20. Observed return is 80% and the game-only loss is A$6, assuming no other movements.

The tool’s count should also reflect paid actions, not every visible reel movement. If one hundred paid rounds include twenty granted feature actions with no separate stake, do not automatically enter one hundred and twenty paid spins. Check whether the entered theoretical model already includes those feature outcomes.

Likewise, add all prizes belonging to the counted paid rounds under the same definition. A free-action counter and a cash prize are different units. How Slots Work explains the relationship between a paid round, its stages and final accounting.

The percentage must match the scenario

A calculator cannot verify a game’s RTP from a title or a screenshot. The Commission’s RTS 3 addresses accessible game rules and return information. Use the relevant help and configuration rather than assigning the prefilled example to an actual release.

A shared name does not establish that editions, modes or stake options use identical mathematics. If a percentage is uncertain, the expected output is conditional on that uncertain input. More decimal places do not turn it into verified information.

The tool accepts theoretical percentages from zero to one hundred. At zero, the formula returns zero expected prizes and an expected loss equal to turnover. At one hundred, expected prizes equal turnover and expected loss is zero. Those boundary calculations describe the entered model, not available games.

Zero expected loss also does not necessarily mean zero risk. Define an entirely fictional one-unit action that returns zero half the time and two units half the time. Its expected gross return is one unit, yet a single action still loses its stake on half the outcomes.

Observed return uses a different range because realised prizes can exceed stakes. Do not replace the theoretical input with a short observed percentage merely to make the expected result match history. That would mix the assumption with the measurement being compared.

The same RTP can describe different prize distributions

RTP supplies a mean, not a complete probability table. To see the missing information, define two separate toy models, each costing one unit per action, with the following full outcome distributions:

Both have 96% theoretical return and a 0.04-unit expected loss per action. Their single-action profit probabilities are 48% and 4.8%. The common return percentage therefore cannot identify even that simple probability without the distribution.

A session-profit calculation needs more: the sequence length or stopping rule, relevant dependencies, stake pattern and prize distribution. A bankroll-survival calculation also needs the starting funds and rules governing continuation. This calculator does not estimate either.

Monitoring guidance considers volatility when assessing variation from theoretical return. We do not insert an assumed volatility, confidence interval or normal-distribution band into the tool. A numerical range around the mean would require justified statistical assumptions beyond RTP alone.

For the concepts, read RTP Explained and Volatility Explained. A profit probability, maximum loss, affordable budget and expected loss are different answers.

Handle changing stakes or return models in separate blocks

The widget has one stake and one theoretical percentage, so it describes a constant-input scenario. When those values change, calculate consistent blocks separately and add the expected money amounts. It has no automatic variable-stake or mixed-mode feature.

For a fictional first block, one hundred paid spins at A$0.20 create A$20 turnover. At an assumed 96%, expected prizes are A$19.20 and expected loss is A$0.80. A second block of fifty spins at A$0.80 creates A$40 turnover; at 94%, expected prizes are A$37.60 and expected loss is A$2.40.

The combined workload therefore has A$60 turnover, A$56.80 expected gross prizes and A$3.20 expected loss. Its stake-weighted return is 56.80 ÷ 60, about 94.67%. Simply averaging 96% and 94% to get 95% would assign the wrong weight to the larger-turnover block.

If every action instead used the same theoretical percentage, an appropriate average stake could preserve total turnover: total stakes divided by the paid-round count. That is a per-round weighted calculation, not the average of the smallest and largest prices. It still would not reproduce the entire distribution of results.

An average stake may contain fractions of a cent. Because this calculator accepts money inputs to two decimal places, do not round an unrepresentable average and treat turnover as unchanged. Calculate consistent blocks separately when the exact average cannot be entered.

For observed return across blocks, sum actual gross prizes and total stakes, then divide those totals. Do not average observed percentages from unequal turnover records. A record’s weight comes from its relevant stakes, not merely how many rows appear in a spreadsheet.

Expected gaming loss excludes other account effects

The formulas concern stakes and game prizes under the supplied model. They do not calculate payment fees, currency conversion, promotional-credit removal, cashout caps or taxes. An account balance can change for reasons outside the game record.

If the prefilled model’s expected gaming loss is A$2 and a separate fixed service fee is A$3, the combined expected expense under those expressly stated assumptions is A$5. Entering the fee as though it were more game stake would incorrectly apply RTP to it.

Promotional rules need their own calculation. A bonus balance is not automatically withdrawable cash, and a wagering target is qualifying turnover rather than an expected loss. The Bonus Wagering Calculator handles the target arithmetic; it does not estimate completion probability either.

Starting funds and stopping decisions also matter to a real sequence. The entered spin count is a fixed workload for the expectation example, not a prediction that a particular balance can fund it. A player can run out before a desired count, and RTP alone does not tell you when that happens.

Use Bonus Terms Explained, Withdrawal Limits and Withdrawals for those separate conditions. Do not read a calculator’s expected gross prize as approval of a future withdrawal.

Read the rounded output with the input assumptions beside it

Money outputs round to two decimal places. A displayed A$0.00 expected loss can therefore represent a small positive amount rather than mathematical zero. For example, A$0.01 turnover at 99.99% gives an unrounded expected loss of A$0.000001.

Keeping precision until the end matters when amounts accumulate. Under a separate fixed model of ten thousand A$0.01 paid spins at 99.99%, turnover is A$100 and expected loss is A$0.01. Rounding each spin’s A$0.000001 expected loss to zero first, then adding those zeros, would miss that cent. Expected values can contain fractions of a cent even when actual monetary entries use whole cents. The formula is an average calculation, not an assertion that each round is charged a millionth of a dollar as a separate fee.

Rounding also affects comparisons. Use the underlying formulas when examining tiny differences, then round the final result rather than repeatedly rounding intermediate products. The extra digits describe arithmetic precision, not certainty about an actual game’s configuration.

If a value is rejected, check its meaning as well as its format: a monetary stake needs the full paid amount, the count must be a whole positive number and actual returns cannot be negative. Blank actual returns and zero are deliberately different inputs.

For an observed record, reconcile starting balance plus incoming transfers minus outgoing transfers, stakes and charges plus gross prizes and adjustments. Include each category once. If the ending balance does not fit, investigate the missing category rather than changing RTP to force a match.

The Slot Glossary and Frequently Asked Questions explain the units. Responsible Gambling provides support and spending-control information. Do not turn an expected amount into a recovery target; this tool performs arithmetic, not a plan for recouping losses.

RTP Calculator FAQs

References

Primary sources checked 7 November 2026. GB explanations and Australian service context retain their scope. All worked money and probability examples are independently defined, not verified game statistics.

• UKGC: consumer RTP explanation.
• UKGC: observed RTP and volatility.
• UKGC: RTS 3, game and return information.
• UKGC: RTS 2, transaction values.
• ACMA: protect yourself from illegal gambling operators.