Slot hit frequency describes how often a defined event happens, usually a spin returning a payout. RTP describes the proportion of stakes a game is designed to return over extensive play. Neither tells you whether your next spin will pay or whether a session will finish in profit.
The distinction matters because a game can return small amounts regularly while your balance keeps falling. Another can pay less often but distribute a larger share of its theoretical return through rare outcomes. Calling the first game “more generous” from its frequency alone leaves out the size of the payments.
This guide uses invented examples to explain the arithmetic. They are not measurements of a named slot, recommendations to gamble or forecasts of what you will receive. The regulatory references concern remote gambling licensed in Great Britain.
What does hit frequency mean on a slot?
In its simplest form, hit frequency is the number of spins with a qualifying outcome divided by the total number of spins. If 24 out of 100 recorded spins produce a non-zero return, the observed frequency of any return is 24%. That is a description of those results, not proof of the game's underlying probability.
The important word is “qualifying”. A published statistic might count any award, only a return above the stake, or a particular feature trigger. Those are different events. Before comparing two percentages, find out what each publisher counted and whether the figure relates to the base game or the whole game cycle.
A £1 spin returning 20p can count as a hit under an any-return definition. Your net result on that spin is still a loss of 80p. “Winning spin” is therefore an unhelpful label unless the writer explains whether they mean a payout or an actual gain.
Hit frequency, RTP and volatility compared
| Measure | Question it answers | What it cannot promise |
|---|---|---|
| Hit frequency | How often does the specified event occur? | A profit whenever that event occurs |
| Theoretical RTP | What proportion of total stakes does the game model return on average? | That proportion of your deposit back |
| Volatility | How widely can outcomes vary around their average? | An exact session length or loss limit |
A useful comparison needs more than one of these measures. Frequency describes occurrence; RTP weights outcomes by their value. Volatility describes variability. None is a personal budget control, and a low-volatility label is not a statement that a game is harmless.
The Gambling Commission distinguishes designed RTP from the return actually measured during play. Its RTP monitoring guidance explains why both sample size and volatility matter when interpreting a measured result. A short personal session is not equivalent to a technical audit.
Two games can have the same RTP and different hit frequencies
Consider two deliberately simplified games. Every play costs £1, every play is independent, and each game has just two possible results. A stated payout includes the amount returned to the player; it is not additional profit on top of a returned stake.
| Model | Paying outcome | Otherwise | Expected return per £1 |
|---|---|---|---|
| A | 40% chance of £2.40 | 60% chance of £0 | 0.40 × £2.40 = £0.96 |
| B | 10% chance of £9.60 | 90% chance of £0 | 0.10 × £9.60 = £0.96 |
Both models have a theoretical RTP of 96%, with an expected loss of 4p per £1 play. Model A pays four times as often, but Model B's paying result is four times as large. In these examples Model B also has greater variance. Real slots normally have many more possible payouts, so you cannot recover their full distribution from one frequency figure.
After ten plays, neither model owes you a particular outcome. You might receive no return, one payment or several. Multiplying a probability by the number of plays gives an expected count across repeated experiments; it does not create a schedule for your next ten spins.
Why frequent payouts can still leave a loss
Here is a separate, invented record of 20 spins at £1 each. Twelve return nothing, six return 50p, one returns £2 and one returns £5. There are eight spins with a non-zero payout, so the observed any-return frequency is 8 ÷ 20 = 40%.
The money calculation tells a different part of the story. The six small returns total £3. Add £2 and £5 and the total returned is £10 against £20 staked. The session result is a £10 loss, with only two spins returning more than their individual stake.
This example does not imply that a real game with a 40% hit rate has a 50% RTP. It shows why counting payments cannot substitute for counting money. The observed return in this made-up sample is 50%; its underlying theoretical RTP has not been specified.
For Great Britain-regulated casino games, RTS 14F restricts celebratory effects for returns at or below the total stake. A displayed award can still be shown to explain the outcome. Its appearance on screen does not, by itself, establish that you made a profit.
Does a dry run mean a payout is due?
No, not in the independent fixed-probability model used here. With a 25% chance of any payout, the chance of no payout on one play is 75%. The chance of no payout over ten independent plays is 0.75 raised to the power of ten, approximately 5.63%.
That calculation answers a question about a run of ten plays considered in advance. After nine unsuccessful plays, the probability of a payout on the next play remains 25% in this model. The game does not need to repair your personal sequence.
Actual slots may contain persistent features or different game states, so a published base-game percentage should not be applied blindly to every situation. Read the rules for the specific mode. A feature meter does not justify assuming a guaranteed profit or increasing spending to reach it.
If a run of losses makes you feel that you must keep playing, stop the session. The mathematical example explains uncertainty; it is not a reason to fund more trials.
Where should you look for reliable game information?
Start with the help menu inside the exact game you have opened. Note the game title, provider, rules, stated RTP and any explanation of features. Record which version or mode the information describes. A review of a similarly named game is not necessarily evidence about the version in front of you.
The Gambling Commission's RTS 3 sets requirements for accessible rules and information about the likelihood of winning. Do not interpret that as a promise that every operator will publish a statistic called “hit frequency”. If the definition or value is absent, treat it as unknown rather than replacing it with an affiliate table's estimate.
Our practical reading sequence is:
- Identify whether the number refers to any payout, a profitable spin or a feature.
- Check the mode, version and source date.
- Look for the game's own explanation rather than relying on a headline ranking.
- Keep frequency separate from RTP and any maximum-win figure.
- Leave the game unplayed if information you need is unclear.
Can you calculate it from your own spin history?
You can calculate an observed frequency if your history includes every completed spin and enough detail to identify the event consistently. Divide the qualifying count by the total count and multiply by 100. Do not exclude losing spins, abandoned sessions or small returns merely because they make the result less interesting.
Suppose you see 27 paying spins in a record of 100. The observed figure is 27%. It does not establish that the true probability is precisely 27%, and it says nothing by itself about how much the game returned. A different group of 100 plays may look quite different.
Avoid collecting extra gambling results solely to make your estimate “more accurate”. More play costs money and still cannot replace access to the game's mathematical model. Existing records are useful for understanding spending, not for proving that a payout is approaching.
Is bonus frequency the same thing?
No. A bonus trigger is one specified event; an ordinary paying spin is another. A slot could return small amounts quite often while its main feature remains rare. Nor does the feature's appearance guarantee that its payout exceeds the money spent before reaching it.
Ask what the denominator is. “One feature per 150 spins on average” does not mean that the 150th spin must trigger it. A figure calculated from base-game paid spins also cannot automatically be compared with one that includes free spins or retriggers.
Maximum-win advertising answers yet another question: what is the largest permitted payout under stated conditions? It does not tell you its probability. Frequency, maximum payout and typical experience should not be treated as interchangeable descriptions.
Common questions
Does a higher hit frequency mean a better slot?
It means the defined event occurs more frequently in the stated model or dataset. It does not establish higher RTP, lower overall losses or suitability for a particular person. Without a definition and payout distribution, the comparison is incomplete.
Is a 96% RTP a 96% chance of winning?
No. RTP is a value-weighted return measure. The two fictional games above both return 96% theoretically but have very different probabilities of a paying play.
Can changing the stake make a payout due?
No. Increasing a stake is not a way to recover a shortfall in previous results. Where a game offers different configurations, consult the rules rather than assuming the odds stay the same or improve.
18+ only. Gambling can cause financial harm. For practical limits and support options, visit our responsible gambling page. The safest way to avoid gambling losses is not to gamble.