Big Multipliers in Crash: Frequency vs Waiting Time

Short answer

Under a narrow illustrative model only, if a chosen target cashout has the same theoretical return r and a hit pays exactly x times stake including the returned stake, then the target-specific hit probability is p = r / x. Using 97% RTP only as an illustration, not verified Aviator math, a 10× target gives p = 0.97 / 10 = 9.7%, so the mean waiting time is about 10.31 rounds.

That does not mean 10× is “due” after ten misses. The mean wait is a long-run average. In any one session, the hit can come earlier or much later.

What are we actually comparing?

Three different things get mixed together: a large multiplier, a real promo bonus, and a game's RTP. A large multiplier is just a round outcome. A promo bonus is an operator action such as Free Bets or Rain. RTP is a long-run return measure, not a signal for when the next 10× will appear. Some casinos run their own crash game, for example Crash in the BC.Game Originals; what its game info shows is covered in our BC.Game review.

When does p = r / x apply?

It applies only in a specific binary model. If success at target x pays exactly x times stake gross, and failure pays 0, then expected gross payout per $1 stake is x · p. If the theoretical return for that target is r, then:

• x · p = r
• p = r / x

Here, gross includes the returned stake on a win. It is not net profit.

Using r = 0.97 as an illustrative input only:

The 50% and 95% columns mean the number of rounds needed so that the chance of at least one hit reaches about 50% or 95%.

The mean is higher than the median because rare long waits pull the average up. So “on average” does not mean “most sessions finish there.”

When should you not generalize that formula?

You should not treat p = r / x as a universal crash distribution, universal named-game odds, or proven Aviator math. It depends on the target-specific model above, plus any real game limits such as allowed cashout targets, caps, ties, rounding, or rule details.

JetX is a useful reminder that RTP alone does not identify a full distribution: the provider page lists an RNG multiplier game with RTP in a 96.2% to 98.9% range, but that range alone does not reveal one unique distribution, one operator configuration, or one target-frequency formula for every build (JetX provider page).

Does a dry window make the next big hit more likely?

No. Under an independent-round model, a dry window does not change the next-round probability. That is the gambler's fallacy.

For a fixed window, use no hit = (1-p)^n and at least one hit = 1-(1-p)^n. These formulas assume iid rounds with constant p.

For example, with illustrative p = 9.7% for a 10× target, 0.903^20 ≈ 13%, so the chance of no 10× at all in the next 20 rounds is about 13%.

• chance of no 10× in the next 10 rounds: 36.05%
• chance of no 10× in the next 20 rounds: 12.99%
• chance of no 10× in the next 50 rounds: 0.61%

And for an illustrative 100× target with p = 0.97%:

• chance of no 100× in the next 20 rounds: 82.29%
• chance of no 100× in the next 100 rounds: 37.73%
• chance of no 100× in the next 300 rounds: 5.37%

Those are probabilities for a fixed forward window from a chosen start point. They are not counts of “how many streaks happen” somewhere in history.

What does this mean in money over 100 rounds?

The direct answer is that expected loss and session outcome are different things.

If you stake $1 for 100 rounds, turnover is $100. At 97% RTP, expected gross return is $97, so expected net result is -$3.

For a binary 10× target model, the expected number of hits in 100 rounds is 9.7, but that is only an expectation. If you get zero 10× hits across those 100 rounds, then your result is a conditional loss of $100 for that session.

So both statements can be true at once:

• Unconditional expectation: 100 rounds at $1 and 97% RTP implies expected loss $3.
• Conditional result if zero 10× hits occur: loss $100.

Confusing those two is one of the easiest ways to misread crash-game variance.

What can an RTP calculator show here?

It can show only expected-value arithmetic, not crash-event frequency or next-round predictions. Our RTP calculator with inputs 97% RTP, $1 stake, 100 rounds shows turnover $100, expected payout $97, and expected loss $3.

The interface label says “spins,” but here it is just a reused generic input for 100 equal bets. It does not turn a crash game into a slot model, and it does not estimate the next multiplier.

English RTP calculator screenshot showing 97% RTP, $1 stake, 100 rounds, $100 turnover, $97 expected payout, and $3 expected loss

Local development interface screenshot from localhost. It demonstrates RTP arithmetic only: turnover $100, expected gross payout $97, expected loss $3. It does not calculate crash-event probabilities or predict the next round.

Are Free Bets or Rain the same as hitting a big multiplier?

No. The available Aviator rules copy describes Free Bets and Rain as promotional mechanics handled by the operator or participating players, not as the same thing as a round reaching 10×. The available rules copy is undated, so check the current game conditions (Aviator rules copy, PDF).

Excerpt from an operator-hosted Aviator rules PDF showing references to Free Bets and Rain as separate promotional mechanics

Crop from page 2 of a public English PDF rules copy retrieved on 2026-10-02. The source document is undated.

Can observed frequency prove fairness or fraud?

No. Observed frequency can only tell you whether a sample looks broadly compatible with a chosen model. For a practical check, prechoose an event such as 10× or higher, use all completed rounds as the denominator, record the exact game, version or rules copy, and stated RTP, and keep conditions comparable. Do not cherry-pick only winning stretches: count k/N across the whole chosen window.

Illustration: 970 hits in 10,000 rounds = 9.7%. That is hypothetical, not our measurement. A sample like that does not by itself prove independence, fairness, or future outcomes. There is no universal minimum N that guarantees convergence.

Observed frequency is also different from algorithm verification. If a game supports Provably Fair, any real verification depends on the provider's published seeds, nonces, rounding, and exact algorithm. A general explainer or tool can help you understand the idea, but it is not guaranteed to be compatible with Aviator. If you are comparing that with “predictor” claims, see our short piece on the Aviator predictor myth.

Does a more frequent target mean a profitable one?

No. Frequency and profitability are different. A lower target like 2× can happen much more often than 100×, but if RTP is below 100%, the long-run expectation is still negative.

That is why crash gambling math and risk shape matter more than isolated screenshots of big hits. Large multipliers change the distribution of results, not the sign of the expectation. If you want the variance comparison versus other casino formats, see our note on crash-game volatility vs slots.

Practical checklist

If you want the practical takeaway, use this page for one thing at a time: estimate target frequency in the simple model, or estimate expected money loss from RTP arithmetic.

1. Start with a clearly stated target, not a story that a multiplier is due.
2. Separate mean wait from 50%/95% at-least-one-hit windows.
3. Do not confuse a large multiplier with operator promos such as Free Bets or Rain.
4. Use RTP tools for expected-value arithmetic only.
5. Follow your local laws and treat real-money play as adult entertainment.

Toolsgambling.com earns through affiliate links to operators. That partnership does not affect these numbers.