Slots Payout Rates: Everything You Need to Know

What a Return to Player Figure Is

The Gambling Commission's consumer guidance on gaming machine payouts, a page last updated on 16 June 2021, describes the figure plainly: “Percentage Return To Player (%RTP) can be thought of as part of the money paid to play the machine that is then given back to the player as prizes.” It is an average and not a session forecast: “The %RTP is an average achieved over a significant number of game plays and not each time the gaming machine is played.” The page works the point through with its own example: “If a gaming machine displays an 85% RTP, you should not expect to win an average of 85 pence for every £1 you stake during a playing session.” It also states the window the average is taken over: “Average %RTP is generally measured over 10,000 or 100,000 games for compensated machines and more games than this for random machines, depending on their category.” That page addresses gaming machines, not online slots. Payout rate, return to player and payback percentage are three names for the same figure: the share of everything wagered that is returned to players as prizes, on average and across all play.

Two further statements on it bear directly on how a machine should be read. “There is no minimum RTP on a gaming machine.” And “Gaming machines must make information available about their category, %RTP and whether they are compensated or random.” The Commission's guidance to licensees puts the display duty in regulation: “Your gaming machines must clearly display the percentage return-to-player figure (% RTP).”, and “You must ensure that all the machines you make available display the correct percentage return to player figure (%RTP), this is required by the Circumstances of Use Regulations.”

What an Online Player Must Be Told

For remote games the duty sits in the Commission's remote gambling and software technical standards, a document last updated on 7 May 2024. The aim of RTS 3 is “To enable customers to make informed decisions about whether to gamble based on their chances of winning, the way the game, lottery or event works, the prizes or payouts on offer and the current state of multi-state games or events.” The rules themselves must be reachable first: “An explanation of the applicable rules must be easily available to the customer before they commit to gamble.” So must the prize information: “For each virtual event, game (including bingo), or lottery, content describing the potential prizes and payouts or the means by which these are calculated or determined must be easily available before the customer commits to gamble.” The guidance names the form that usually takes: “Information should be made available about the amounts that customers may potentially win, for example in the form of pay-tables, or by showing the odds paid for particular outcomes.” Among the acceptable ways of describing the likelihood of winning, the guidance lists “the average theoretical return to player percentage.”

The Commission's live return to player monitoring guidance, last updated on 29 January 2025, defines what that advertised figure is: “This is the designed return to player percentage of the game, it will also be the advertised RTP of the game as displayed in the player facing rules, as per RTS 3C.” Operators are then held to it: “Remote operators are required to monitor the performance of the games they offer.”, and “Licensees must ensure sufficient RTP monitoring is in place for both under and overpayments.” The aim is “to ensure games are operating fairly as designed and advertised.”

A Return Figure Belongs to One Game, and to One Build

The Commission's testing annex states where a game's return comes from: “Collectively the game rules determine the overall game RTP.” Those rules are software, and a change to them is a regulated event: “A major update, which will require external retesting by an approved test house, is any software change which may affect the fairness of a game.”, while “An update that does not impact game fairness is referred to as a minor update and can be released without the need for external retesting.” Builds of the same title are not assumed to behave alike. “Remote games are often released separately on different platforms or channels (for example, mobile, flash, download).” and monitoring has to be able to see them apart: “Monitoring must not be so aggregated that it hides errors at a lower level.”, “For example, errors that only exist in the mobile version of the game might be less visible if monitoring aggregates all markets and channels into one calculation.” A return figure is therefore a property of a particular game, and of the build of it that was released and tested — never of a studio's catalogue or a casino's lobby.

How Much Play the Average Needs

“Measuring a game with only a small amount of play will be pointless as the tolerance will be too large for meaningful results.” What narrows the band is volume: “The tolerance will be wider when only a limited amount of play has been measured, but as the volume of play increases the tolerance will decrease.”, and “After a significant amount of plays the actual RTP should be very close or equal to the theoretical RTP.” Volatility sets the band at any volume: “Volatility is vital to these calculations regardless of volume of play and will be a key parameter to include when establishing the allowable tolerance for each game.” The Commission publishes worked tolerances. For a game designed at 91.68% with a volatility of 5.6: “So if 400,000 games had been played to accrue turnover and win figures of the example then the allowable tolerance will be 1.75 above or below 91.68.” At that volume, “The game could return between 89.93% and 93.43% and still be considered to be performing as expected.” Those are volumes for a game across all of its players, not for one session.

No Game Runs Loose or Tight, and No Game Is Due

RTS 7 forbids the mechanisms that the folklore assumes. “Random number generation and game results must be ‘acceptably random’.” and “Adaptive behaviour (that is, a compensated game) is not permitted.” — which the guidance explains as prohibiting “automatic or manual interventions that change the probabilities of game outcomes occurring during play.” Misleading presentation is also out: “Game designs or features that may reasonably be expected to mislead the customer about the likelihood of particular results occurring are not permitted, including substituting losing events with near-miss losing events and simulations of real devices that do not simulate the real probabilities of the device.”, and “Games may not falsely display near-miss results, that is, the event may not substitute one losing outcome with a different losing outcome.” A game's maths cannot be changed under a player mid-session: “The rules, payouts and outcome probabilities of a virtual event or game may not be changed while it is available for gambling, except as provided for in the rules of the game, lottery or virtual event.” And the Commission's consumer page states the consequence for a player watching a reel: “The odds of achieving a win in the current game remain constant and are not affected by wins or losses in previous games.” Nothing a player can observe in a game's behaviour identifies it as higher-returning; the published figure in the game's own rules does.

What Regulators Publish, and What They Do Not

Published figures from regulators are about operator revenue, not a comparable consumer return rate. The Gambling Commission's industry statistics annual report for April 2025 to March 2026, published on 17 September 2026, “provides an overview of Gross Gambling Yield (GGY) made by licensed gambling operators from GB gambling customers for each customer-facing sector, along with the numbers of licensed operators and premises.” It is “based on data reported to us by the operators we license and regulate.” Within the remote sector, “online casino games dominate the sector, generating £5.7 billion in GGY, £4.8 billion of which was from slots games”. In Nevada, the Gaming Control Board's monthly publication is described as follows: “The Monthly Revenue Report is a summary of revenue information for nonrestricted gaming activity.”, and “Each report reflects 1-month, 3-month, and 12-month data.” In Pennsylvania, “The PA Gaming Control Board is dedicated to ensuring the public is informed about the gross revenue and tax revenue generated across all gaming markets regulated by the Board.” None of these is a payout rate for a player to compare between casinos.

Large Jackpots Skew the Average Upwards

Large jackpots skew an average return upwards. If most people playing a particular game see returns around 88% and one or two players hit six-figure jackpots, the figure reported for that game is pulled above 88%.

Slots are highly volatile games. A typical session is a long run of losing spins interspersed with the occasional small, medium or large win. Imagine watching 10 players each play slots for an hour: nine of them might be down by varying amounts while the tenth hits a large jackpot. Even though nine players lost money, the tenth player’s jackpot raises the average for all of them.

That is worth knowing when choosing which slots to play. A progressive jackpot game may advertise a 90% payout rate, and the jackpot component is part of what that figure describes rather than a return a typical session sees.

None of that is a finding about any one game. It is the arithmetic of a skewed distribution, and how much it matters depends on how the figure is published. The Gambling Commission’s implementation guidance for progressive jackpot systems states that “The jackpot system’s return to player percentage should be displayed as per RTS 3C, this could be one combined game and progressive jackpot RTP figure or broken down into the base game and jackpot component.” Where the components are broken out, the base game can be read on its own; where a single combined figure is published, the jackpot contribution sits inside it.

What a Return Figure Costs per Hour

Consider a player in Las Vegas with $100 budgeted for slots, treated as money that may be lost in full.

Suppose the game chosen has a 90% payout rate and the stake is $1.00 per spin, and suppose the reels can be spun 10 times a minute, or 600 times an hour.

At an average loss of $0.10 per spin and 600 spins an hour, the expected cost of play is $60 an hour. At that rate a $100 bankroll will not last two hours unless a jackpot lands.

Now take the same situation with a game returning 98%. The expected loss is $0.02 of every $1.00 staked, which at 600 spins an hour works out at $12 an hour, or more than eight hours of play from the same bankroll — barring exceptionally poor or exceptionally good luck.

The difference in the average outcome is large even on an 8% gap in payout rate and a small starting bankroll. Extend that across a long run of play and the reason the figure matters becomes clear.

More important than any of this is to plan a budget before playing, and to stick to stakes that fit it.