Bonus Frequency Calculator

Defaults are Spins per Bonus 200, Bet Size 1 dollar, Avg Bonus Win 50 dollars. Submit. Cost per Bonus is 200.00, Bonus EV is −150.00, and 1 in N Spins reprints 1:200. Cost is frequency times stake: 200 × 1 = 200 spent on average to see one feature. EV is the average feature payout minus that cost: 50 − 200 = −150. The feature, taken alone, is a 150-dollar expected hole at these numbers. Base-game line hits are not included, so this is not the game’s RTP.

Keep bet at 1 and raise Avg Bonus Win to 200. EV becomes 0: the feature pays back exactly the average spin cost to trigger it. Raise it to 250: EV +50. Drop Spins per Bonus to 100 with the original 50 average: cost 100, EV −50. Double the bet to 2 with 200 spins and 50 average: cost 400, EV −350, because the average win field is in dollars, not multiples. If the help file says “average bonus 50×,” type 50 × your stake, not 50.

The frequency field is a mean waiting time in spins, the reciprocal of bonus probability if triggers are independent. A published 1 in 200 is an average; clusters and droughts still happen. Do not type the last drought you suffered (480 spins) as if it were the parameter. Do not type the last two-spin miracle either.

A second worked case: frequency 180, bet 0.80, average win 90. Cost = 144, EV = 90 − 144 = −54. Same structure. If you buy the bonus for a fixed 100× and the average feature is 80×, this tool is the wrong shape; you would enter frequency 1, bet equal to the buy price, and average equal to the feature mean, which is a different product. The defaults model natural triggers during base play.

Reload 200, 1, 50 whenever you want the canonical −150 illustration. It exists to kill the intuition that “the bonus is where you win.” At these defaults the bonus is where you recoup 50 of the 200 you spent waiting. The other 150 has to come from base-game pays if the certified RTP is near 96%. This calculator will not add those pays in.

Bonus frequency is the average number of base-game spins between feature triggers. Combined with stake and with the mean feature payout, it yields a cost to see a bonus and an isolated expected value for that bonus cycle. The calculator’s purpose is to separate the feature from the base game, because streamers show the former and the PAR sheet is built from both.

Video-slot design since the late 1990s parked a large share of RTP inside free-spin rounds, pick-and-click games, and hold-and-win respins. Hit frequency in the base game can look friendly while the feature, the part players wait for, is a 1-in-200 event with a mediocre average. Marketing still talks as if the bonus were a gift. Arithmetic treats it as a random prize whose trigger you paid for in advance, spin by spin.

Limits: the EV here is bonusAvg − freq × bet. It ignores base-game wins during the wait, retriggers already baked into “avg bonus win,” stacked features, and ante-bet modes that raise both trigger rate and stake. It ignores the distribution: a 50-dollar average can be 90% 20-dollar features and 10% 320-dollar features. Variance of the wait is also ignored; a geometric 1/200 process has standard deviation near 200 spins.

History of the number: studios publish “bonus frequency” on sell sheets for operators, sometimes as a range at different stakes. Regulators care about overall RTP, not about whether the bonus EV is −150. Players encounter the number in YouTube thumbnails (“bonus every 80 spins!”) which are sample paths. This tool only accepts a mean.

Do not read a negative Bonus EV as a broken game. Default −150 on a 200-dollar wait with a 50-dollar feature simply states that 75% of that cycle’s handle is still unaccounted for and must live in small base pays or in a fatter tail than 50 dollars. If the certified RTP is 96% on 200 dollars of handle, expected total return is 192, so the base game plus any residual would need to supply 142 on top of the 50. Whether a given title does that is not computed here.

Feature cost is a cycle identity, not a PAR sheet. If a bonus is advertised as arriving every 120 spins in expectation, and the stake is one unit per spin, then the mean amount spent waiting for one feature is 120. If the feature itself pays 80 on average, the isolated ledger is 80 minus 120, which is minus 40. That minus 40 is the feature's own hole. It is not the game's return to player, because the base game still dribbles small pays during the wait. A PAR sheet would split those two streams and add them. This page never asks for return to player and never performs that split.

Derivation when triggers are independent. Waiting time until the first success in Bernoulli trials is geometric. If the trigger probability per spin is p, the mean wait is one over p. Typing 120 as the frequency is typing that mean, not typing p; equivalently p is 1/120. Cost per bonus in expectation is mean wait times stake, hence 120 when the stake is one. Isolated expected value is mean feature minus that cost. Linearity still holds if the wait is not geometric, provided 120 is the true mean wait and 80 is the true mean feature: expected cost is still expected wait times stake. Meters, mystery stacks, and guaranteed-feature counters make the wait non-geometric in many modern titles; the ledger identity survives, the geometric variance formula does not.

Worked defaults. Frequency 120, stake 1, bonus average 80: cost 120, isolated expected value minus 40. If return to player were 96 percent, two hundred spins are not the cycle here; one hundred twenty spins at one dollar would return 115.20 in a full-game mean, which would equal the isolated 80 plus 35.20 of other pays in expectation. That split is illustrative only. Raising stake to five without converting an 80-dollar average into a 400-dollar average is a unit mismatch, not a new theorem. If the help file quotes 80 times rather than 80 dollars, convert first: at a one-dollar stake the numbers coincide; at twenty cents they do not.

Sign of the isolated mean is educational. A positive isolated mean would mean the feature alone more than refunds its trigger cost, which would imply either a very high game return, a tiny base-game share, or optimistic typing. Minus 40 on the shipped numbers is the intended hole. Buy-bonus prices are a different product: you pay a multiple to skip the wait, and the relevant comparison is purchase price versus mean feature, not 120 times stake. Retriggers belong inside the 80 if you typed them that way; they are not added afterwards.

Assumptions. Constant stake. One feature type. Average already including retriggers if entered that way. No laboratory reel-strip weights. No contribution to a progressive. The page reports a feature ledger so that "every 120 spins" stops sounding like a gift. It is a mean wait whose cost is 120 units when the stake is one, against an 80-unit prize, leaving a 40-unit leak before any base-game crumbs are counted.