Illustrative odds — shown to explain formats, not to place bets.
A crash game runs in short rounds shared by everyone playing at once. You place a stake before the round starts, a multiplier begins climbing from 1.00×, and it stops at a point decided before the round began. Cashing out while it is still climbing settles your bet at whatever the multiplier reads at that instant. Doing nothing loses the stake.
The presentation varies — a plane, a rocket, a line on a chart — but underneath, every one of these games is the same four terms.
The crash point of a round is determined before the multiplier starts moving. Nothing you do during the climb influences it, and the number on screen is an animation of a result that already exists. That is the single most useful fact about the format.
Because rounds are shared, the interface usually shows other players’ stakes and cash-outs live. It is entertainment, not information: their decisions have no effect on the crash point, and a visible cash-out at a high multiplier says nothing about where this round ends.
A settled round returns stake × multiplier, stake included. Cash out at 2.00× and you double; at 1.20× you keep a fifth more than you staked. Cashing out low wins small amounts often; waiting for a high multiplier wins rarely and large. Neither approach changes what the game returns over time — it only changes the shape of the swings, exactly as volatility does in a slot.
Auto cash out takes the multiplier for you at a figure set in advance. It is worth using, because it takes the decision out of the two seconds when the number is moving fastest and the temptation to wait is strongest. It does not reduce the chance of a round crashing below your figure.
The result of a round comes from a random number generator, fixed before the betting window closes. Many crash games publish a provably fair scheme: the operator commits to the round’s outcome in advance by publishing a hash of it, reveals the seed afterwards, and anyone can recompute the round and check that the committed value matches. It proves the result was not altered after bets were placed. It does not make the result predictable — that is the whole point of committing to it in advance.
Every round is independent. A run of low multipliers does not make a high one more likely, and a 50× round does not have to be «paid back» with a string of crashes at 1.01×. The history strip along the top of the screen is a record, not a signal.
