Instant busts in crash games: how the house edge is built in

An instant bust is a crash round that stops at 1.00x, before anyone can cash out. In the illustrative model below, with a 3% margin, it happens in about 3.96% of rounds, roughly one in 25. It is the visible part of the house edge, which is taken from every cash-out target.

By Nairobi Ultimate Editorial (prepared with AI tools under our editorial policy) · Last updated 2026-10-12

What an instant bust is

In a crash game the multiplier rises from 1.00x and stops at a point fixed by the game's maths. An instant bust is a round whose stopping point is 1.00x itself: the screen shows the round ending before it has risen at all, so no cash-out is possible and every stake in that round is lost.

It is the sharpest form of the pattern described in how crash games work. This guide asks a narrower question: how does a game decide how many rounds end that way, and what does that rate say about the house edge?

All figures below belong to an illustrative model with a stated margin. They are not the rules of any operator's game, and a real game's rules and return to player are the figures that count.

A stated model with a margin parameter

Many published crash descriptions use a rule of this shape. The game draws a number u evenly between 0 and 1, computes a raw multiplier, rounds it down to two decimals, and shows 1.00x if the result is below 1.00.

raw multiplier X = (1 − h)/(1 − u), shown result = floor(100 × X)/100, shown as 1.00x if below 1.00

Here h is the house margin, chosen as 0.03 (3%) in the main example. Because 1 − u is evenly spread between 0 and 1, the chance that X is at least some value m is the chance that 1 − u is at most (1 − h)/m. For any m of 1 or more that chance is simply (1 − h)/m.

Step 1: the chance of an instant bust

A round is shown as 1.00x when the raw multiplier is below 1.01, because anything from 1.00 up to 1.01 rounds down to 1.00x and anything below 1.00 is shown as 1.00x as well. So the chance of an instant bust is the chance that X is below 1.01.

P(instant bust) = 1 − (1 − h)/1.01 = (0.01 + h)/1.01

With h = 0.03 this is 0.04/1.01 = 4/101, which is 3.96%, or one round in 25.25. Two parts make it up: X below 1.00, with probability h = 3.00%, and X from 1.00 up to 1.01, with probability 0.97 − 0.97/1.01 = 0.96%.

Illustrative example, three margins in the same model. A larger margin lifts the instant bust rate, so the 1.00x rate is a rough signal of how much has been shaved from the curve.

Step 2: the edge still sits at every target

Instant busts are not where the whole margin is collected. A player who targets t (a multiple of 0.01, at least 1.01) is paid when the shown result is at least t, which happens exactly when X is at least t.

average return per shilling at target t = t × (1 − h)/t = 1 − h = 0.97

At the lowest target of 1.01x the chance of being paid is 0.97/1.01 = 96.04%, which is 100% minus the 3.96% instant bust rate, and the payout of 1.01 gives the same 0.97. At 2.00x the chance is 48.5%. The 3% cost appears at every target because the whole curve was scaled down by (1 − h), as the guide on the maths of the average multiplier works out.

Step 3: what rounding contributes

Set h to 0 and the game is fair. The instant bust rate is then 0.01/1.01 = 0.99%, yet the average return at every target is still 1 × (1/t) × t = 100%. Those rounds cost the player nothing on average, because a target of 1.01x or more has its chance measured against the same curve.

This shows that the instant bust rate is a by-product of two things: the margin, and the rule that nothing below 1.01x can be cashed out. With a margin, the rate exceeds h by the 0.96% slice above, which is 0.97 × 0.01/1.01 of all rounds. A game could also be built to stop a fixed share of rounds at 1.00x by a separate rule. Designs differ, so the rules of the game in question are the evidence.

How often a short session meets one

The chance of at least one instant bust in n independent rounds is 1 − (1 − 4/101)^n, which is 1 − (97/101)^n.

Illustrative example, 3% margin. After a few dozen rounds an instant bust is more likely than not, so meeting one is routine and carries no message about the game.

Two in a row has chance (4/101)^2 = 16/10,201, about one round pair in 638. That is rare, but each round is independent, so it does not make the next round more or less likely to bust. Reading such runs as a pattern is the gambler's fallacy.

What the displayed rules should tell you

The regulations require a licensee to display "rules of games, odds, house edge and average return to the player" (regulation 20(2)), and to apply this to online gambling with the necessary modification (regulation 20(3)). Regulation 45 adds that the rules must be displayed before any wager, that the random number generator is certified to test randomness and theoretical return to player, and that the theoretical return to player of each game is publicly disclosed.

A return to player of 97% and a house edge of 3% are the same statement. The stated figure is the one to read, because the instant bust rate is only one of the ways the margin appears. For how a random number generator is meant to supply the draw u, and why the return is a long-run average, see the guide on RTP and house edge.

Why a short session can still cost more than expected

Instant busts and quick rounds put many stakes through the margin in little time. The expected cost is the margin multiplied by the total staked, which in the model is KES 3 for every KES 100 passing through the game.

Regulation 87(2) lists the tools that licensees must offer: deposit limits on a daily, weekly and monthly basis, loss, session and expenditure limits, reality checks and self-exclusion of at least twenty-four hours. A limit set before the first round works on the total staked, which is the figure that carries the margin. Gambling is addictive; play responsibly, and treat any limit as a ceiling on what the game is allowed to cost.