Return to Player (RTP)

RTP is the number you will see quoted on every online slot, usually somewhere between 94% and 97%. It is the single most cited statistic in online gambling and also the most widely misunderstood one, because the word “return” strongly implies something that it does not actually promise.

What the number actually measures

RTP is a theoretical figure calculated from the game’s maths model. The designer knows every symbol combination, every payout, and the exact probability of each one. Multiply each payout by its probability, add them all up, and you get the proportion of total stakes the game is designed to pay back.

A 96% RTP means that for every £100 wagered on the game, £96 is returned to players collectively and £4 is retained. That retained portion is the house edge. RTP and house edge are the same fact stated from opposite ends, and always sum to 100%.

The critical word is collectively. RTP describes the behaviour of the game across its entire lifetime, pooled across every player who has ever spun it. It says almost nothing about what happens to you tonight.

Why your session never looks like 96%

Here is the part that trips people up. If RTP were a reliable per-session figure, a £100 evening on a 96% slot would end with roughly £96 in your account, every time. That would be an extraordinarily boring product, and nobody would play it.

What actually happens is that the 96% is delivered through a wildly uneven distribution. Most spins return nothing. Many return less than the stake. A small number return a great deal. The long-run average only emerges after a sample size far larger than any individual will ever produce, typically millions of spins.

Over a realistic session of a few hundred spins, your actual return might be 0% or it might be 4,000%. Both are entirely consistent with a 96% RTP game working exactly as designed. The spread of those outcomes is governed by volatility, which is a completely separate property from RTP and arguably the more useful one to know.

Two games can both have 96% RTP and feel nothing alike. One might pay small amounts constantly; the other might pay nothing for two hundred spins and then deliver the lot in a single bonus round. RTP cannot tell those apart.

The compounding problem

There is a second reason your money disappears faster than 4% per £100 might suggest: you do not stake your money once.

Start with £100 and stake it all on a 96% game. On average you have £96. Stake that, and you have about £92. Stake that, and £88. The house edge applies to turnover, not to your deposit, and every time you recycle winnings back into the game it gets applied again.

This is why the practical question is never “what is the RTP” but “how many times will I put my bankroll through this game”. A hundred spins at £1 is £100 of turnover regardless of whether you started with £20 or £200. Play long enough and the theoretical edge stops being theoretical.

Where RTP varies

A few things worth knowing when you see a figure quoted:

• The same game can ship at different RTPs. Many studios produce several configurations of one title, 96.5%, 95.0%, 94.0%, and the operator chooses which to deploy. The game looks identical. Regulated markets generally require the active figure to be visible in the game’s info panel, so check there rather than trusting a review site.
• Table games depend on the rules and on you. Blackjack’s RTP is only ~99.5% if you play basic strategy correctly; play by instinct and the real figure drops several percent. Roulette’s is fixed by the wheel, 97.3% single-zero, 94.7% double-zero, and no decision you make changes it.
• Progressive jackpots distort the headline. A portion of the quoted RTP is locked inside a jackpot that almost nobody wins. The base game RTP that applies to your actual session is meaningfully lower than the advertised total.
• Side bets are consistently worse. The main game might run at 99%; the tempting side bet next to it often sits between 85% and 94%.

What to do with the number

Use RTP for exactly one thing: comparing two games you are already choosing between. Higher is better, and the difference is real over time. 97% versus 94% is a threefold difference in expected cost per unit staked.

Do not use it to predict a session, to estimate what you will walk away with, or to reason about whether a game is “due”. Each spin is independent; the random number generator has no memory of what the game owes anyone.