Overview
Crash games: the format and the model behind every guide
This overview explains what a crash round is, states the single probability model that every crash guide on this site uses, and points to the guide that covers each topic in depth.
What happens in one round
A crash round shows a number that begins at 1.00x and climbs while the round lasts. At some moment the climb ends and the round is over. Before that moment, a participant can lock in a cash-out at the current multiplier, or can set an automatic target in advance. If the lock-in comes first, the stake is returned multiplied by that value. If the round ends first, the stake is gone.
The important detail is when the ending point is chosen. In the designs this site discusses, the ending point (the crash point) is fixed before the round starts, from a random number. The rising graphic only reveals it slowly. So the participant does not race the game in any real sense; the only choice that affects the outcome is the target.
The model used across this site
Every crash guide here uses one generic teaching model, not the rules of any particular operator. With a house edge e, the chance that a round reaches a multiplier m (for m of at least 1) is (1 - e) / m. At a 1% edge the chance to reach 1.10x is 90.00%, the chance to reach 3.00x is 33.00%, and the chance to reach 1,000x is 0.099%.
This model has a neat property. A target m pays m times the stake, and its chance is (1 - e)/m, so the average return of any target is 1 - e of the stake. Real services may round, cap, or alter the curve, so treat these numbers as a reference shape, not a promise about a specific product.
| Cash-out target | Chance to reach it | As odds | Mean rounds per success | EV per unit |
|---|---|---|---|---|
| 1.10x | 90.00% | 1 in 1.1 | 1.11 | -1.00% |
| 1.50x | 66.00% | 1 in 1.5 | 1.52 | -1.00% |
| 3.00x | 33.00% | 1 in 3.0 | 3.03 | -1.00% |
| 20.00x | 4.95% | 1 in 20.2 | 20.20 | -1.00% |
| 1,000.00x | 0.10% | 1 in 1,010.1 | 1,010.10 | -1.00% |
Read the table row by row. The second column is the chance a single round reaches that target. The fourth column is the average number of rounds per success, which is just one divided by that chance. The last column is the expected result per unit staked, and it is the same in every row because the target cancels out of the calculation.
Where the money goes
The edge has to come from somewhere, and in this model it comes from rounds that end very early. A share of rounds ends at 1.00x, before any target above 1 can be reached. At a 1% edge, with crash points rounded down to whole cents, that share is about 1.98%. Every participant in those rounds loses, whatever target they chose.
What the individual guides cover
Each guide in this group takes one question and works through it with its own numbers. The theme word in each address is just a page label, so pick by topic, not by name.
- Amber: why every target has the same expected loss.
- Aztec: a single 3.00x target examined from every side, at a 2% edge.
- Bamboo: how long you wait for a 5.00x round, using the geometric distribution.
- Chicago: exact chances of long losing runs at a 2.00x target.
- Comet: the crash-point formula and the instant 1.00x crash.
- Crown: a seeded simulation at a 1.50x target, and how to read it.
- Crystal: turning hash bytes into a crash point, step by step.
- Cyber: commit-reveal and pre-computed seed chains.
- Dragon: the doubling system and why it fails, plus common myths.
- Egyptian: bankroll size and the chance of going bust.
Before you go further
Every row of the table above has a negative expected value. No target, timing habit or betting pattern changes that in this model. The guides explain the arithmetic so that the shape of wins and losses is clear; they do not suggest a way to come out ahead. For the general idea of a house edge see the house edge guide, and for hash checks see the hash fairness guide. If gambling stops being fun, the responsible gambling page lists places to get help.