A slot can return small amounts in many rounds while the overall balance falls. A claim such as “a payout every third spin” says little about the chance of making a profit unless the word payout is defined.
Hit Rate, or Hit Frequency, is commonly used to describe how often a particular paying outcome occurs. First establish what the source counted: any positive payout, a return greater than the full cost of a round, or the trigger of a specific feature. Two percentages are comparable only when they count the same event: triggering a bonus differs from receiving part of a stake back.
What counts as a payout?
It helps to distinguish three situations. A zero payout returns none of the round’s stake. A payout above zero but below the full stake returns some money, yet the round’s net result remains negative. A payout equal to the stake breaks even; a round is profitable only when its return exceeds its full price.
For example, a payout of 0.40 on a stake of 1 leaves a net result of minus 0.60. An animation, a sound or a Win label does not change that arithmetic. Under a definition counting every positive payout, this round is a hit. It is still a losing round in net terms.
Use the full participation cost. If an option increases the paid stake, do not compare its return only with the earlier base price. For a cascading game, compare the combined payout of the cascades with the single full price of that round when paid rounds are your chosen observation unit.
An example: a 50% hit rate and an overall loss
Consider ten invented rounds at a stake of 1 each. Their payouts in order are: 0; 0.20; 0; 0.50; 1; 0; 2; 0; 3; 0. This dataset is an educational illustration and does not describe an actual game.
Five of the ten rounds have positive payouts, giving an observed hit frequency of 50% under this definition. Only two payouts exceed the stake: 2 and 3. The proportion of profitable rounds is therefore 20%. One round returns the stake exactly, and two return only part of it.
Total wagers are 10 and total payouts are 6.70, leaving a net loss of 3.30. Payouts divided by wagers equal 67% in this small dataset. That is the return of our invented example, not a real slot’s theoretical RTP. Frequent positive payouts are entirely compatible with an overall loss.
Why hit frequency does not determine RTP
Frequency asks how often an event occurred. Return asks how much of the amount wagered came back. Payout sizes are essential to the second calculation; hit frequency alone cannot recover RTP.
Two imaginary datasets can each contain ten stakes of 1 and five positive payouts. In the first, every payout is 0.20, returning 1 in total. In the second, every payout is 2, returning 10. Both have a 50% hit rate, yet their returned amounts differ by a factor of ten. This is arithmetic, not a comparison of two commercial slots.
The UK Gambling Commission explains measured RTP as payouts divided by turnover. When recording observations, keep the number of paid rounds, total wagers and total payouts separately. A single “number of wins” column cannot replace those figures.
Bonus frequency measures a different event
Triggering free spins or another feature is a distinct event. Its frequency need not equal the frequency of any payout in the base game. A hit rate does not supply the probability of free spins, a maximum win or a profitable session.
Also distinguish a naturally triggered feature from a purchased feature where the rules offer a purchase. Paying to enter a bonus does not reveal how frequently it triggers in normal play. A valid comparison needs matching modes and the same event definition.
Check the units and counting method
A percentage, “one in N rounds” and an unexplained number are different forms of information. If a source explicitly means one occurrence in four rounds, that corresponds to 25% within its stated definition. A bare number of 4 cannot automatically be converted to 25%.
Identify the denominator: paid rounds, individual cascades, free spins or all displayed events. Several payouts within a single purchased round should not silently become several paid wagers. If the source does not publish a definition, mark the metric as unspecified.
What a short observation cannot establish
Twenty rounds with frequent payouts describe those twenty rounds. They do not prove a stable theoretical frequency or predict what follows. An observation cannot support a promise that a bonus “must be due soon”.
The length of a current sequence without a payout does not, by itself, identify the wait until the next one. A probability calculation requires a model and its assumptions, not just a counter of empty rounds. There is no need to continue gambling to obtain “proper statistics”.
How to read a game information card
Find the metric’s original source. Read the event definition and units. Confirm the mode and full price of a round. Consider frequency separately from RTP, payout sizes and volatility. When information is missing, leave the uncertainty visible instead of guessing.
Why cascades are not separate paid spins
A single paid spin can produce several paying cascades. A statistic counting each cascade does not give the proportion of spins that return money. Consider four illustrative spins costing 1 each, with no extra charge for the cascades within them.
| Paid spin | Individual payouts | Total for the spin |
|---|---|---|
| 1 | 0.20 and 0.30 | 0.50 |
| 2 | No payouts | 0 |
| 3 | 1.40 | 1.40 |
| 4 | 0.10; 2; 0.50 | 2.60 |
Three of the four spins pay, giving a paid-spin hit frequency of 75%. There are six individual paying combinations: two in the first spin, one in the third and three in the fourth. Six payouts divided by four spins equals 1.5 payouts per spin. Calling that “150% winning spins” would be wrong: a proportion of spins cannot exceed 100%.
Another possible measure is the proportion of paying cascades among all cascades, including those ending without a combination. It requires the complete cascade count, which this payout log does not contain. A list of awards alone is therefore insufficient to reconstruct that frequency.
A statistic’s name needs to identify what counts as one attempt. In game specifications, the observation unit matters as much as the stake unit. Payout variability does not remove this distinction: count matching events before comparing their sizes.