RTP, house edge, volatility and hit frequency describe different parts of a game's mathematics. Reading them together gives a clearer picture of cost and payout pattern, but none can predict what a short session will produce.
These measures sit on top of the game-family and mechanics map and the guide to RNG, live and hybrid outcomes. The first tells you what kind of game you are looking at, while the second explains where its result comes from. The figures in this guide describe how wagers and prizes are distributed once those rules are fixed.
Four measures with different jobs
RTP is the share of total wagered value that a game is designed to return as prizes over its mathematical life. House edge expresses the expected share retained by the house instead. Volatility describes how widely individual results can spread around the average. Hit frequency describes how often a defined paying result occurs, without saying how large that payment is.
The terms are connected, but they are not substitutes. Two games can publish the same RTP while producing very different sequences of prizes. Two games can also have a similar hit frequency while assigning very different values to those hits. A frequent award that returns less than the stake still counts as a paying result under many game definitions, even though that round leaves the player with a net loss.
RTP is a long-run ratio
The GLI-19 interactive gaming standard defines RTP as total amount won divided by total amount wagered and distinguishes theoretical RTP from actual RTP. Theoretical RTP comes from the complete probability model or a sufficiently deep simulation. Actual RTP is calculated from recorded wagers and prizes after a game has been played.
The UK Gambling Commission's player guide stresses that RTP is an average over a significant number of plays, not a promise for each spin or session. A displayed RTP of 96 percent therefore does not mean every stake returns 96 percent. One round can return nothing, another can return less than the stake, and another can pay several times the stake. Only the combined value across a very large body of play is expected to approach the designed ratio.
Turnover also includes money wagered again after it was won. A player who deposits 20 units and repeatedly restakes awards can generate much more than 20 units of turnover. Session loss should therefore not be estimated by applying the house edge only to the original deposit. The relevant base for the mathematical expectation is total wagering activity.
House edge turns the ratio around
For a simple house-banked game with a fixed paytable, house edge is the complement of RTP. A theoretical RTP of 96.5 percent corresponds to a theoretical house edge of 3.5 percent. The Commission's explanation of house edge describes it as the percentage the casino expects to keep on average from each hand or spin under normal patterns of play.
That subtraction needs context. In a game where strategy changes the probabilities, the advertised RTP may assume a stated strategy or optimal play. A progressive contribution can be included or disclosed separately. Peer-to-peer games usually use rake or commission rather than a house-banked edge. The UK rules on likelihood information accordingly allow the player-facing measure to be RTP, house edge, margin or the probability of winning events, depending on the product.
House edge is an expectation, not a meter that deducts a fixed amount after every round. A 3.5 percent edge does not impose a loss of 3.5 units in every 100 units of short play. It describes the centre of the game's distribution over enough independent trials under the stated rules.
Volatility changes the route to the average
Volatility concerns dispersion. The Commission's RTP monitoring glossary says standard deviation is commonly used to represent it. A high-volatility design can place more of its return in rare, large prizes. A low-volatility design tends to place more of its return in smaller, more frequent prizes.
Imagine two fictional games that both return 96 percent in their complete models. One may award many modest amounts and only a small number of large prizes. The other may produce longer stretches with no payment and concentrate more value in occasional large results. Their expected return can match even though the experience of a few dozen rounds looks nothing alike.
This is why a short actual RTP can sit far above or below the theoretical figure without proving that the design changed. Higher volatility normally creates a wider range of plausible short-run results. As the volume of play grows, the observed ratio is expected to move closer to the theoretical figure, although a player cannot know when or through which sequence that convergence will occur.
Volatility does not create a schedule. A run of losses does not make a large prize due, and a recent prize does not use up the chance of another independent result. It describes the distribution built into the paytable, not a memory of one account's recent play.
Hit frequency counts events rather than value
Hit frequency is the share of rounds that meet the game's definition of a paying hit. GLI treats it as a separate field from theoretical RTP on the game's PAR specification sheet. That separation matters because frequency answers how often, while RTP answers how much value is returned in aggregate.
A game can show a high hit frequency by awarding many small amounts. Some of those awards may be lower than the stake, so the balance still falls after the round. Another game can have fewer hits but assign more of its total return to those rarer results. Neither pattern is automatically more favourable in a short session, and hit frequency alone says nothing about the chance of finishing ahead.
The definition can also vary. A paytable may count any credited award, only base-game combinations, or a result that includes a bonus trigger. Published game rules should say what is included. Without that definition, comparing hit-frequency percentages from two different products can be misleading.
The paytable joins the figures
The paytable is where probabilities and award values become one mathematical model. Spanish technical rules make the same connection by requiring public prize tables and by monitoring expected return for each game, mode or variant. GLI-19 likewise says a paytable reflects all possible awards and its return percentage.
Before staking, check that the RTP belongs to the exact version and wager configuration on screen. Some games offer several mathematical variants under one title. Bet level, line selection, a side wager, bonus purchase, progressive contribution or strategy assumption may change the applicable return. A lobby badge is less useful than the rules and help screen for the selected version.
The same review should identify whether a displayed maximum prize is part of the base game, requires a separate feature or depends on a jackpot. A large headline award can have little influence on ordinary hit frequency while contributing materially to volatility. It can also contribute only a small portion of total RTP if its probability is extremely low.
What the numbers can and cannot tell you
RTP and house edge summarize expected value. Volatility summarizes the spread around that value. Hit frequency summarizes the occurrence of defined paying events. Together they describe a game more usefully than any one figure alone.
They still do not forecast the next result, guarantee a playing duration or turn random play into a winning system. The practical use is comparison and informed budgeting. Verify the exact rules, treat the entire stake as at risk, and do not interpret a long-run average as money owed back to an individual session.
