RTP, or return to player, is the percentage of total wagering that a casino game is designed to return as prizes over a very large amount of play. A 96% RTP describes long-run game math. It does not promise that a player who wagers $100 will finish with $96.
That distinction matters because RTP is often read as a session forecast when it is really a pricing measure.
The percentage answers one question
RTP answers: What share of total action is expected to flow back to players over the long run?
If a game has a 96% theoretical RTP, the complementary house edge is 4%, assuming the usual casino-game relationship where player return and house advantage account for the full wager.
| Measure | What it describes | What it does not describe |
|---|---|---|
| Theoretical RTP | The return built into the approved rules, paytable, or game model | What one player will receive tonight |
| Actual RTP | Prizes paid divided by turnover during a measured period | Proof of a defect from a small sample |
| Session return | One player's cash result relative to that session's wagering | The game's true long-run percentage |
| House edge | The casino's long-run share of action | The percentage of a starting bankroll guaranteed to disappear |
The terms sound similar but should not be mixed. A player can have a 160% session return on a 96% RTP game by winning and leaving. Another player can have a 0% session return after losing the session bankroll. Neither result changes the game's theoretical RTP.
The two RTP calculations
Theoretical RTP comes from the probability model and payout schedule:
Theoretical RTP = Sum of (Probability of each outcome × Total return for that outcome)
Each probability is the chance of an outcome. Total return includes the winning profit plus the returned stake where the calculation uses total return rather than profit alone.
Actual RTP is measured from live game records:
Actual RTP = Total prizes returned ÷ Total amount wagered
Suppose a game records $500,000 in turnover and pays $472,500 in prizes during the period:
Actual RTP = $472,500 ÷ $500,000 = 94.5%
That does not automatically mean a game designed for 96% is malfunctioning. The amount of play and the game's volatility determine how much measured results can move around the theoretical figure.
A player-cost example
For a game with 96% RTP:
House edge = 100% − 96% = 4%
If a player produces $2,000 in total action:
Expected loss = $2,000 × 4% = $80
The $2,000 is not necessarily the player's starting cash. A $200 bankroll can create much more than $200 of action because wins are replayed. That is why coin-in and total action matter when converting RTP into expected dollars.
The $80 is an average over repeated comparable action, not a bill due at the end of that session. Actual results can be a large win, a small loss, or the loss of the entire session bankroll.
The advertised number may contain assumptions
RTP is not always a single unconditional figure.
In slots, the percentage belongs to a particular approved game configuration or paytable. Two cabinets displaying the same game theme may use different approved return settings. A progressive game may include jackpot contribution in its stated return, meaning part of the percentage sits in a rare prize most players will never hit.
In video poker, the quoted maximum RTP normally assumes the listed paytable and mathematically correct decisions. Strategy errors reduce the player's achieved return. Read video poker RTP for that skill-dependent context.
In blackjack, the return depends on the rules and the player's decisions. A figure calculated for basic strategy should not be presented as the result of arbitrary play. In roulette and baccarat, the percentage is better attached to a specific wager than to the whole game because different bets can have different house edges.
Why equal RTP games can feel unrelated
RTP does not tell you how prizes are distributed. That is the job of volatility, variance, and hit frequency.
Consider two hypothetical games with the same 95% RTP:
- One returns much of its value through frequent small wins.
- The other places more of its return in rare bonuses or a jackpot.
Their long-run price is equal, but the second game can produce longer losing stretches and a greater chance that a limited bankroll expires before a major prize appears. A high RTP therefore does not make a game low-risk in a short session.
Game speed matters too. A 2% edge applied to 40 decisions per hour creates less expected hourly cost than the same edge, same bet, and 400 decisions per hour. RTP is one input; wager size and pace determine how quickly action accumulates.
How operators use actual RTP
Casinos and game suppliers do not judge performance from one player's result. They compare actual return with the theoretical target over suitable play volume, accounting for volatility and approved tolerances. The UK Gambling Commission's guidance on calculating actual RTP from wins and turnover also explains why wider variation is expected when the sample is smaller.
Operationally, a result outside an expected range is a signal to review data, configuration, metering, or game performance. One unusual day is not enough to prove a fault. Repeated or statistically meaningful exceptions deserve investigation.
This is also why actual hold and theoretical hold should remain separate in casino reporting. Actual results measure what happened. Theory estimates what comparable action is expected to produce over time.
How to use RTP without fooling yourself
Use RTP to compare the long-run price of genuinely comparable choices. Confirm that the percentages refer to the same rules, paytable, strategy assumptions, jackpot treatment, and game version.
Then add the missing questions:
- What is the total wager per decision?
- How quickly will decisions occur?
- How volatile is the game?
- Does the stated return assume correct strategy?
- Is a large part of the return tied to an extremely rare prize?
The RTP Comparison Tool can compare percentages, while the Slot Loss Estimator translates return, bet size, and pace into expected cost. For the casino-side counterpart, read house edge and theoretical loss.
RTP is valuable because it exposes the price built into the game. It becomes misleading only when a long-run percentage is treated as a personal refund, a short-session guarantee, or a complete description of risk.