Payout percentage is the share of total wagering returned to players as prizes. If players wager $750,000 and the game returns $696,000, the measured payout percentage is 92.8%. The remaining 7.2% is the game’s measured hold for that same pool of play, before any adjustments that the reporting method treats separately.
That definition sounds simple, but the phrase is used for two different numbers: the theoretical payout percentage built into the game and the actual payout percentage produced during a particular period. Confusing them is the source of most arguments about payout.
One phrase, two different measurements
A designed or theoretical payout percentage is a mathematical property of a game, paytable, ruleset, or wager category. It is calculated from every possible outcome and its probability. It answers: What proportion of wagers should this game return over an extremely large number of plays if the approved rules remain unchanged?
An actual payout percentage is calculated from recorded wagering and recorded returns. It answers: What proportion did this game return during the period being measured?
| Number | How it is obtained | What it tells you | What it cannot tell you |
|---|---|---|---|
| Theoretical payout percentage | Probabilities and prize values in the approved game math | Long-run expected return | What one player will receive today |
| Actual payout percentage | Recorded prizes divided by recorded wagering | What happened during a stated period | Whether the game’s design changed |
| Session return | One player’s cash-out or ending value compared with money committed | That player’s result | The long-run cost of the game |
A slot designed to return 94% may record 89%, 98%, or more over a limited sample. That variation does not automatically show a malfunction or a change in settings. It shows that a finite sample can land away from its mathematical expectation, especially when rare prizes account for part of the return.
The calculation needs a defined denominator
The basic formula is:
Payout percentage = Total value returned to players ÷ Total amount wagered × 100%
Where:
- Total value returned to players is the prize value included by the applicable game or reporting definition.
- Total amount wagered is the turnover, handle, or coin-in included in that same definition.
- Both figures must cover the same game population and the same time period.
Suppose a machine records $750,000 in coin-in and $696,000 in included player returns.
$696,000 ÷ $750,000 × 100% = 92.8% payout
The corresponding measured hold is:
100% − 92.8% = 7.2% hold
The two percentages complement each other only when they use the same data and treatment. If a report excludes an externally funded jackpot, promotional credits, taxes, or another adjustment from one side of the calculation, simply subtracting from 100% may produce the wrong comparison.
Payout percentage is not payout odds
The word payout also appears in expressions such as “35 to 1” on roulette or “3 to 2” on blackjack. Those figures describe the prize paid for a winning wager. They are not payout percentages for the whole game.
A bet can advertise a large prize and still have a poor average return because the winning event is rare. Conversely, a wager that pays only even money can have a comparatively high return if it wins often enough.
This is why payout odds, house edge, and return to player should be kept separate:
- Payout odds describe the amount won when a specific bet succeeds.
- Probability describes how often the winning event should occur.
- Payout percentage or RTP combines all outcome probabilities and prizes into an average return.
- House edge describes the average portion of wagering not returned to players in the long run.
A higher percentage can still cost more per hour
A payout percentage is useful for comparing price, but it does not include pace or bet size. A 96% game played rapidly at $5 per round can produce a larger expected hourly loss than a 92% game played slowly at $1 per round.
Expected loss connects payout percentage to actual wagering volume:
Expected loss = Total amount wagered × (1 − payout percentage)
Consider two sessions:
| Session | Wagering volume | Payout percentage | Expected loss |
|---|---|---|---|
| A | $1,000 | 96% | $40 |
| B | $300 | 92% | $24 |
Session A uses the game with the better percentage, yet its greater betting volume creates the larger expected cost. The percentage tells you the average price per dollar wagered; expected loss tells you what that price means for the amount actually put into action.
What players should check before comparing games
A payout percentage is only comparable when the surrounding conditions are comparable. Check:
- Is the number theoretical or actual? A designed return and a one-month result are not substitutes.
- Which paytable or wager category does it cover? Multi-game and multi-denomination products can contain different returns.
- Does correct strategy affect it? Video poker and some decision-based games publish returns that assume a stated strategy.
- Are jackpots included? A return may include a very small probability of a large progressive award.
- Is the figure for the base game or the entire product? Side bets and optional features can have different mathematics.
- What is the betting pace? Game speed changes the amount wagered per hour.
The UK Gambling Commission’s public explanation of return to player on gaming machines makes the central point plainly: the percentage is an average achieved over a significant number of plays, not a result delivered each time the machine is used.
Why short sessions can look nothing like the percentage
Payout percentage is an average. It does not describe the distribution of results around that average. Two games can both have a 95% theoretical return while producing very different experiences.
One may pay small awards frequently. Another may direct more of its return into rare bonuses or jackpots. The second game can produce longer losing stretches even though its long-run return is identical. That difference belongs to volatility and hit frequency, not payout percentage alone.
Consider a game whose rare top prize contributes two percentage points to its 95% theoretical payout. A player who never encounters that prize is not playing a temporary 93% version of the game. The prize remains part of the probability model, but an individual sample may not contain it.
How casinos use the figure
For machine operations, payout and hold are linked to performance review, product configuration, compliance, and accounting. Managers may compare:
- theoretical return for the active paytable;
- actual return over a defined period;
- wagering volume and number of games;
- jackpot and bonus contributions;
- variance from an established range;
- whether meter, system, and accounting data reconcile.
A material difference between theoretical and actual return is a reason to examine sample size and records. It is not, by itself, proof of a problem. The review may identify ordinary variance, a large jackpot, a promotion, incomplete data, a meter issue, or a genuine technical exception.
For table games, operators more often discuss house edge, theoretical win, drop, and hold percentage. The relationship is similar, but the reporting base is different. Table-game drop is money or value entering the table’s drop system; it is not the same as total wagers made as chips circulate through repeated decisions.
The personal-refund mistake
A 95% payout percentage does not mean any of the following:
- a player who wagers $100 must leave with $95;
- a machine that recently retained money is now required to return it;
- a player who is down more than 5% has been treated unfairly;
- a high-return game cannot empty a bankroll;
- past losses improve the next outcome.
If $100 is wagered once, the result may be $0, $20, $100, $500, or another allowed award. If the same $100 is recycled through many bets, total wagering can become several times larger than the original bankroll. Payout percentage applies to that accumulated wagering volume, not merely to the cash first inserted.
A practical way to use payout percentage
Use the figure as a cost comparison, not a forecast of tonight’s cash-out.
When the rules and betting pace are comparable, a higher payout percentage usually means a lower long-run cost per dollar wagered. Then add the factors the percentage does not show: bet size, speed, volatility, jackpot concentration, strategy requirements, and the amount of time you intend to play.
For a direct comparison, the RTP Comparison Tool can translate small percentage differences into expected cost at a chosen wagering volume. The Expected Loss Calculator performs the same step when you already know the house edge.
Payout percentage is valuable because it removes some of the mystery from gambling math. Its limit is equally valuable: it measures an average return across wagering, not a debt the game owes to an individual player.