What RTP really means
Return-to-player is a long-run average built into a game’s mathematics — not a promise about your session, your night or your money. What the number is, how it relates to house edge and hold, and the misconceptions that follow.
What the number actually is
The certifier that writes the rules puts it plainly. GLI-11, the technical standard used by many regulators, defines it in its glossary as “a ratio of the ‘total amount won’ to the ‘total amount wagered’ by a player” — and it draws one distinction that resolves most confusion. Such a return, the standard says, may be “theoretical” (based on mathematical calculations or simulations) or “actual” (based on the metering supported by a fielded gaming device).
The RTP printed in a game’s information screen is the theoretical one: a property of the game’s design, computed before anyone plays it. The actual return is whatever the machine happens to have paid so far. The two only converge over a very long run — which is the whole source of the misreadings below. For the reference definition, see RTP in the glossary.
RTP, house edge and hold
House edge is simply the complement of theoretical RTP: house edge = 100% − RTP. A game with a 96.0% RTP has a 4.0% house edge. They are the same fact stated from opposite sides — the fraction the design expects to return, and the fraction it expects to keep. See house edge for the full treatment.
Hold is different again. Hold is an operational number — the share of money actually retained over some real period, the metered “actual” return read backwards. Because it is measured over finite play, hold drifts around the theoretical house edge and rarely equals it exactly. Theoretical RTP is a design constant; hold is a measurement with variance in it.
Why 96% is not “96 back tonight”
A theoretical RTP is an expected value. Expected values are recovered by the law of large numbers: the average of many independent rounds converges to the theoretical mean only as the number of rounds grows large. Over a session of a few hundred spins, the realised return is governed by variance, not by the average — so it swings widely, in both directions, around the printed figure.
The misconceptions that follow
“Hot” and “due” machines. The belief that a game is “due” to pay after a cold streak, or has gone “hot” and will keep paying, assumes rounds depend on history. Certified games are built so they cannot. GLI-11 requires, in as many words, that “a game shall not adapt its theoretical return to the player based on past payouts” — and, separately, that it shall not adjust the likelihood of a bonus based on the history of previous games. Each round is drawn independently. A machine has no memory of what it owes.
Short-session variance mistaken for a broken game. A run of losses is exactly what the variance of an honest game produces some of the time; so is a run of wins. Neither is evidence about the RTP. Distinguishing a genuinely mis-set game from ordinary variance takes far more data than a session provides — the subject of how RTP is verified.
What RTP is not
RTP is not a session promise, not a per-player entitlement, and not a schedule of what any individual will get back. It does not mean a fixed fraction of each deposit is returned. It is one number — a long-run design average — and reading it as anything shorter-run than “millions of rounds” is the mistake behind almost every myth about it.
Common questions
Is RTP guaranteed?
No. RTP is a theoretical long-run average of total returned divided by total wagered, realised only over an enormous number of independent rounds. It is never a guarantee about a single session, a single day or a single player. Over any short stretch the return you actually see can sit far above or far below it.
Why did I lose more than the house edge?
Because the house edge — 4% for a 96% RTP game — is a long-run expectation, not a cap on a session. Real sessions are dominated by variance, not the average. Most sessions on a game with a 4% edge end up losing more than 4% of the amount staked (and a minority win), because the average is only recovered across millions of rounds, not a night of play.
Sources (1)
Education, not advice. This chapter explains how return-to-player is defined, computed and checked so you can read the number honestly. It is not a system, and nothing here treats gambling as a way to make money — over enough play the mathematics favours the house. 18+.
Next in the pathHow RTP is calculated