A crash game draws one number per round — the bust point — from a distribution calibrated to the published edge. The climbing multiplier is animation; the number was set before the round began, and at provably fair implementations it is derived from a seed chain committed in advance. Understanding that single fact reorganises the whole game.
How the bust point is computed
Typical implementation: the round's hash converts to a number that maps onto a distribution where the probability the game reaches multiplier M is roughly (1 − edge) ÷ M.
| Target multiplier | Reach probability (3% edge example) |
| --- | --- |
| 1.5x | 64.7% |
| 2x | 48.5% |
| 5x | 19.4% |
| 10x | 9.7% |
| 100x | 0.97% |
If you want the math: if P(reach M) = 0.97/M, the expected value of cashing out at M is M × 0.97/M = 0.97 — a 3% edge at every target. That uniformity is the design: no cashout point is smarter than another. Many implementations add an instant-bust floor (e.g. 1 in 33 rounds bust at 1.00x) as part of reaching the same number.
What auto-cashout actually does
Sets your target before the round and executes it without you. It does not change expected value — the table above is flat — but it converts crash into a fixed-odds bet chosen deliberately, removes mid-round decisions (the place where discipline dies), and eliminates the latency risk of manual cashout. Use it always; choose the target by variance appetite, not by strategy claims.
Streaks at your target
At a 2x target you lose just over half of rounds. Per 100 rounds:
| Losing streak at 2x | Probability |
| --- | --- |
| 6 in a row | ~60% |
| 8 in a row | ~22% |
| 10 in a row | ~6% |
Stake so the routine streak cannot end the session: 1–2% units, exactly as in the blackjack bankroll framework. High targets invert the experience — long droughts, rare spikes — at the same expected cost.
Reading a crash game's fairness page
Three things to find: the published edge (1–4% across common implementations — the number varies by game and operator, so check per game), the seed-chain commitment (verifiable history anchor), and the verifier. The provably fair walkthrough covers the checking procedure; a game that hides any of the three has answered your question.
FAQ
Is there a best cashout multiplier?
No — the edge is engineered flat across targets. Low targets grind, high targets spike; expected cost per wagered dollar is identical.
Do crash strategies work?
Target-switching and stake progressions rearrange variance, never the edge. Auto-cashout at a fixed target with flat stakes is the entire defensible playbook.
What edge do crash games charge?
Published per game, commonly 1–4%. The fairness page states it; the seed math enforces it. Compare before playing — the spread between games is real money.
Can the casino make me bust early?
Not at a provably fair implementation without breaking its committed seed chain — the bust point predates your bet. Verify once and the question retires.
Why did I bust at 1.00x?
Most implementations include instant busts as part of the edge calibration (often ~1 in 30–35 rounds). It is in the published distribution, not a malfunction.