RTP Explained for Australian Readers

Return to player, usually shortened to RTP, describes a game’s theoretical return on money staked under a specified set of rules. It does not promise that a particular deposit, session or player will receive that percentage back. Understanding its calculation starts with separating total stakes from deposits and long-run mathematics from the results you actually record.

This guide explains those distinctions through hypothetical examples. It also covers house edge, game configurations, random outcomes and the questions to ask when a feature or progressive jackpot is involved. Use the Gambling Guides hub for other reading routes and the homepage RTP overview for a shorter introduction.

For adults aged 18 and over. Gambling involves the risk of losing money. Where commercial referral links are provided, the website may earn commission; commercial relationships can affect inclusion and placement. The Affiliate Disclosure explains the funding arrangements.

Separate three questions before using a percentage

These questions connect, but they need separate labels. A game description supplies theoretical information; a transaction or game log supplies observations. A deposit receipt alone supplies neither total turnover nor a measurement of the game’s return.

What return to player means

Theoretical RTP is the proportion of stakes that a game’s mathematical model returns as prizes on average over the long run, under the model’s rules. A hypothetical 96% RTP means a theoretical return of 96 units for every 100 units staked. The remaining 4 units represent the corresponding theoretical loss.

The return includes all qualifying game payouts counted by that model. It is not an extra profit paid on top of the original stake. Read the paytable’s payout convention so you know whether a displayed amount includes any returned stake.

Actual or observed RTP uses a different calculation: game returns divided by stakes over a selected past period. The period and included activity matter. A sample can be above or below the theoretical figure without changing what the theoretical model describes.

The UK Gambling Commission’s RTP definitions distinguish designed return, measured return and turnover. They provide a technical vocabulary for these calculations. The source’s regulatory context is the UK; it does not establish Australian authorisation or verify a particular online service.

A percentage without its context is incomplete. Record the game, version or configuration, the source and which activity the figure covers. Do not borrow an RTP from another installation merely because the title and artwork look familiar. The relevant question is which mathematical version applies to the game being described.

Turnover counts every stake, not just the deposit

Suppose someone makes a hypothetical $50 deposit and stakes $2 on each of 40 rounds. Turnover is $80 because $2 multiplied by 40 equals $80. That is possible when some earlier rounds return money that is then staked again. It does not imply an additional $30 deposit.

Each repeated stake exposes money again. A return remaining in a balance is still money that can be lost on a later round. Describing later play as free because it uses earlier returns hides that exposure.

Figures to keep separate

Figure	What it records
Deposits	Money added to an account, rather than every subsequent game stake.
Turnover	The sum of stakes on the activity being measured.
Game returns	Amounts returned by those game outcomes, using a consistent payout definition.
Withdrawals	Money taken out of the account, rather than a measure of game turnover.
Remaining balance	Money still held, with unsettled or pending items identified separately.

If stakes vary, add the actual amounts rather than multiplying by an assumed average. Twenty rounds at $1 and ten rounds at $3 produce $50 turnover: $20 plus $30. The number of rounds alone is insufficient. Keep promotional credit, fees and unrelated account adjustments separate if they would affect the record.

A theoretical-return example needs a turnover base

Assume a hypothetical game has a stated 96% theoretical RTP for the configuration concerned. For $1,000 of total stakes, the model’s average return is $1,000 multiplied by 0.96, or $960. The corresponding theoretical loss is $40.

This calculation does not say that depositing $1,000 produces a $960 withdrawal. It also does not say that every block of $1,000 in stakes loses exactly $40. Actual results can vary, and the same starting money can contribute to several stakes before it is lost or withdrawn.

“Long run” describes the mathematical average, rather than a personal waiting period after which a player is owed the expected amount. A session ending below that average does not create a debt that later rounds must repay. Extending play also exposes more money to the game’s rules.

The formula is useful for understanding a model, provided the inputs match. If the source percentage excludes a separate side bet or applies to a particular feature mode, do not apply it to a combined stake without accounting for those differences. An accurate multiplication cannot repair a mismatched base.

Read an observed result as a record of that sample

Consider an illustrative account starting at zero, with no fees, bonuses, outstanding bets or other adjustments. It receives $100 in deposits. Across the measured activity, stakes total $500 and gross game returns total $470. The account pays $60 in withdrawals and finishes with $10.

The balance reconciles as $100 plus $470, minus $500 and minus $60, leaving $10. The net game loss is $30 because stakes exceed game returns by that amount. The cash view agrees: deposits of $100, less withdrawals of $60 and the remaining $10, leave a $30 loss.

Observed RTP for that game log is $470 divided by $500, or 94%. The $100 deposit is not the denominator. Nor should the $60 withdrawal be substituted for $470 of game returns: a withdrawal is an account transfer, not the total returned by the measured rounds.

That sample does not prove the game is designed at 94%. Equally, a different illustrative log with $200 of stakes and $230 of game returns has an observed return of 115%, without proving a designed RTP above 100%. The dates, sample size, game mix and scope all matter when interpreting a recorded percentage.

The Commission’s RTP calculation guidance explains that play volume and volatility affect statistical monitoring. There is no fixed number of rounds that guarantees an individual’s sample will equal the theoretical percentage.

House edge is the complement of the stated return

For a complete theoretical model using the same stake base, house edge is 100% minus RTP. A hypothetical 96% RTP therefore corresponds to a 4% theoretical house edge. A 94% RTP corresponds to 6% under the same definition.

Keep the unit straight. Four per cent of $1,000 turnover is $40. It is not automatically 4% of a deposit, account balance or withdrawal. Likewise, house edge is not a fee deducted as a separate, visible charge from every round.

A quoted edge can also depend on the rules and assumptions used in its calculation. For a decision-based game, the figure may assume specified decisions. For separate side bets, each paytable can have its own mathematics. Our casino-rules overview explains why the complete rule set matters. Do not transfer one bet’s figure to another simply because they appear on the same screen.

Compare percentages on a consistent basis

Two RTP figures can be compared mathematically only after identifying what each covers. Check whether they concern the same stake type, configuration and feature treatment. A base-game percentage and a percentage including a jackpot return component answer different questions.

For a purely hypothetical comparison on $1,000 turnover, 94% produces $940 theoretical return and 96% produces $960. The $20 difference describes the models under those assumptions. Neither amount is a predicted session result, and neither percentage changes a losing session into dependable income.

When activity is split between games, a combined theoretical average needs turnover weights. Suppose $200 is staked under a 94% model and $800 under a 96% model. Expected returns are $188 and $768, totalling $956 on $1,000 turnover. The weighted theoretical return is 95.6%, rather than the simple midpoint of 95%.

That weighted example assumes the quoted models apply to all the stated stakes and uses no additional charges or adjustments. It is an explanation of averages, rather than a method for controlling actual results. Keep observed game returns in a different column if you also hold a real activity log.

The exact configuration matters more than the title

A supplier may make more than one mathematical configuration available for a game. Where that is the case, the available versions and the operator’s permitted choices depend on the product and applicable requirements. The same game name does not establish a single universal percentage.

For a supplier example, Hacksaw Gaming’s Le Clubber information lists multiple RTP versions for one title. That establishes the existence of supplier variants. It does not identify the version used by any particular operator or establish availability to a reader.

Selecting an available tested configuration is different from inventing a personal RTP or deciding whether the next round wins. A statement about an approved configuration needs evidence identifying the version and the scope of its approval. A losing streak alone does not establish that the RTP configuration has changed.

Look for the information supplied within the exact game’s rules or help screens and any relevant version details. If a supplier lists several possible returns, that list does not by itself identify which one a particular installation uses. Record the installation’s own stated figure where available and leave an unresolved configuration unresolved.

Under the UK framework, the Commission’s testing strategy describes approved test-house work on random-number generators, mathematics, software and gameplay. A testing or certification statement has a scope: it can relate to a specified game, version or system under stated standards. It does not establish service authorisation in Australia or promise an individual’s result. Our Australian services overview separates authorisation checks from game mechanics.

Scope claims about random outcomes carefully

A random-number generator supplies random values that the game’s rules use to determine outcomes. Random does not mean every possible prize has the same probability. The prize probabilities and payout structure together determine the mathematical return; the word random alone does not provide an RTP figure.

For a hypothetical game made up of independent rounds, previous results do not change the next round’s probabilities. A sequence of losses does not make a win due. A recent large prize does not, by itself, prove that the next independent round has a lower chance of paying.

Do not turn that example into a claim that every event within every gambling product is independent. A finite deck without replacement changes as cards are removed. A feature can carry state between stages. A jackpot can have rules involving a displayed meter or qualifying condition. Read the actual design before applying an independent-round explanation.

The UK Gambling Commission’s remote technical standard on random outcomes sets requirements for the products within its scope, including rules on adaptive behaviour and changes to outcome probabilities. It also recognises rules-based bonus and special features. This is a scoped technical reference, rather than proof that an unidentified website complies with it.

The useful distinction is between a rule that is part of the game’s disclosed design and an unsupported explanation invented after a result. A losing sequence alone cannot identify a hidden setting or establish that the next result must compensate for earlier play.

RTP and variation answer different questions

RTP describes an average. Volatility describes variability in outcomes. Two hypothetical games can share a 96% theoretical return while distributing prizes differently, such as through more frequent smaller amounts or less frequent larger awards.

That means the percentage alone cannot tell you how often a round returns money, the likely length of a session or the chance of receiving a headline prize. A maximum win is a ceiling under stated conditions. A hit-frequency figure, if supplied, needs its own definition of what counts as a hit.

A return can also be smaller than the stake. If a $2 round pays $0.50, the gross return is $0.50 and the net loss is $1.50. Counting it as an event that paid something does not make it a profitable round. Keep prize frequency, net outcomes and theoretical return as separate measures.

The Betting Odds Explained guide covers a related distinction for betting: a price-implied percentage is not proof of an event’s true chance. For games, avoid turning a return percentage into a claim about the probability of a particular prize.

Read feature percentages and paytables in context

A feature can change the relevant rules, stake or payout structure. Check whether the quoted RTP covers the whole game cycle, a particular mode or a conditional stage. Do not assume that a percentage calculated after entering a feature applies to every ordinary stake used to reach it.

Component percentages can be added only when the documentation establishes compatible definitions. If an illustrative model explicitly assigns 90 percentage points of total-stake return to ordinary outcomes and 6 to features, the combined return is 96%. That does not justify adding an unrelated feature’s advertised percentage to a base-game RTP.

Likewise, a payout multiple needs its stated stake base. A 30-times prize based on a qualifying $0.50 amount is $15. Multiplying by a separate $2 figure would give $60, but that is the wrong calculation unless the rules designate $2 as the base. Our paytables overview explains why labels belong beside prize figures.

Use the actual game’s information for stake-dependent qualifications, purchase options or special modes. This guide does not assume that every mode is offered, that every option uses the same RTP or that increasing a stake improves the percentage.

Progressive jackpots need their own definitions

For a progressive game, establish what the quoted return includes. A source may distinguish ordinary game return from a jackpot component, or provide a combined figure under stated assumptions. Do not assume the word progressive tells you whether the headline RTP includes every prize.

A contribution to a prize pool is also different from an individual’s chance of winning it. If a hypothetical rule allocates $0.02 from each eligible $1 stake to a pool, that describes a 2% contribution. It does not mean every player receives 2% back as a jackpot or has a 2% chance of winning the pool. Nor does it by itself establish a two-percentage-point component of RTP: funding and payout accounting need the actual rules.

Check the qualifying stake, pool arrangement, current meter, reset or starting amount and any published trigger conditions. A fixed maximum win, a progressive meter and a must-hit-by limit are different kinds of information. A stated cap needs its own scope, currency and conditions.

These details depend on the particular game and installation. They do not establish that every operator can choose every contribution rate or cap. If the source does not explain how a jackpot component enters the quoted return, keep that question open rather than adding a guessed percentage.

The UK Commission’s progressive-jackpot standard illustrates the detail needed within its scope, including funding, eligibility and ceilings. For a capped jackpot, check what the rules say happens to later contributions or overflow rather than assuming the meter explains the entire arrangement.

Use money and time records without treating averages as limits

A theoretical loss is not the maximum that can be lost. The $40 theoretical-loss example on $1,000 turnover does not limit an individual’s loss to $40. Actual outcomes can exhaust the money available for a session before an intended number of rounds is completed.

Increasing rounds while keeping the same stake increases total exposure. One hundred $1 stakes create $100 turnover; five hundred create $500. Any returned money staked again is part of that larger total. A small per-round stake can therefore sit beside substantial overall activity.

Record time separately from money. A deposit limit, a turnover total and a time limit each describe something different. A limit is a stopping point, rather than an amount or duration that must be used. No return percentage creates an obligation to continue until the balance approaches an average.

The money and time overview provides further context. If gambling is affecting essential spending, relationships or wellbeing, the Responsible Gambling page links to independent support. Understanding a percentage does not remove the risk of losing money.

Keep a record that preserves the distinctions

For a published figure, note the game, exact configuration where identified, quoted percentage, included modes and current source. Keep its date. For an observed sample, record the measured period, stakes, gross game returns and any exclusions or adjustments.

For an account reconciliation, add opening balance, deposits, withdrawals and closing balance as separate fields. Do not silently merge several games into a measurement described as one game’s actual RTP. If the available statement lacks the detail needed, explain the limitation rather than reconstructing missing stakes from deposits.

A useful correction report identifies the particular sentence or calculation and the public source that differs. Use our Contact page for editorial enquiries and remove private account information. The operator concerned holds its own game and transaction records; this website cannot alter a game setting or settle an account dispute.

Sources and editorial context

Georgia Hitch is the website’s named author and editorial contact. Our Editorial Policy explains sources, worked examples and corrections. Numerical examples here are hypothetical illustrations with stated assumptions, rather than observed tests of an operator.

Use the current game’s own rules alongside technical definitions. A general mathematical explanation cannot identify a private account’s settings or establish that a provider is authorised to serve Australian customers. The About Us page explains the website’s coverage and role.

Related guides and tools

Use the RTP and Expected Loss Calculator to check turnover, theoretical figures and entered gross returns. The Volatility guide explains outcome variability, while the Pokies guide explains the stake and feature bases needed for a matching model.

RTP FAQs

What does a stated RTP of 96% mean?

It describes a theoretical long-run return of 96 units for every 100 units staked under the specified game's mathematical model. It is not a promise to return 96% of an individual's deposit or session stakes. The game configuration, rules and included features matter.

Why is turnover different from my deposit?

Turnover adds every stake, including money wagered again after earlier returns. A deposit records money added to the account. A hypothetical $50 deposit can support $80 of total stakes if some returned money is reused, but that money is exposed to loss again each time it is staked.

How is actual or observed RTP calculated?

Divide gross game returns by total stakes for the measured sample. If stakes total $500 and game returns total $470, observed RTP is 94%. Use a consistent payout definition and identify the period and activity included. The result does not by itself establish the designed percentage.

Is a game's house edge the amount I will lose?

For a complete model on the same stake base, house edge is 100% minus RTP. A 96% return corresponds to a 4% theoretical edge. That describes an average, rather than an individual's exact loss or maximum loss. It is not automatically a percentage of the deposit.

Does every installation of a game have the same RTP?

A supplier may provide different mathematical configurations. The available versions and permitted operator choices depend on the product and applicable requirements. Check the exact installation's published information. A shared title does not prove a shared percentage, and a list of available versions does not identify the selected one.

Do earlier losses make an independent round due to pay?

No. In the independent-round example, previous outcomes do not change the next round's probabilities. Do not extend that statement to every stateful feature, finite deck or progressive design. Those products need their own rules. A losing sequence alone does not establish a compensating future outcome.

Can two games with the same RTP distribute prizes differently?

Yes. The average return can be the same while the variability and distribution of outcomes differ. RTP alone does not state prize frequency or the chance of a headline award. A round returning less than its stake still produces a net loss even though it paid something.

Does a quoted RTP always include progressive jackpots?

Check the particular game's definition. A source may distinguish ordinary return from a jackpot component or give a combined figure under stated assumptions. Pool contribution, winning chance and maximum prize are separate concepts. Do not add a guessed jackpot percentage to an unexplained headline return.

Can an observed return above 100% prove a profitable game?

No. A hypothetical sample of $200 in stakes and $230 in game returns has an observed return of 115%. It describes that sample, rather than proving a theoretical return above 100% or a dependable future result. The measured scope and outstanding account items also need identifying.

What information should I keep beside an RTP figure?

Record the game, configuration where identified, source, date, stake basis and included modes or features. For an observed sample, retain stakes and gross returns separately from deposits and withdrawals. Mark unresolved details clearly and keep private account credentials out of editorial enquiries.