How crash games work, and why no cash-out strategy wins
Crash is the simplest casino game ever designed and the one that produces the most confident strategies. A multiplier climbs from 1×. At an unknown moment it dies. Cash out first and you are paid your bet times whatever the multiplier read. Wait too long and you get nothing.
One decision, one number. That simplicity is also why crash is the easiest place to see, exactly and provably, that no exit strategy is better than any other.
The mechanic
You place a bet before the round starts. A multiplier begins at 1.00× and rises. You can cash out at any point while it is climbing; the moment you do, the round is settled at the multiplier displayed. If the round crashes before you act, the bet is lost.
The crash point is drawn before the curve is drawn. The animation is presentation — a way of turning a number that was already decided into a few seconds of tension. Climb speed changes how long the reveal takes and nothing about the outcome.
The distribution behind the curve
Crash points follow a distribution with a very particular property: the chance of reaching multiplier t is inversely proportional to t. Getting to 2× is about half as likely as getting to 1×. Getting to 10× is about a tenth as likely. Getting to 100× is about a hundredth as likely.
That inverse relationship is not decoration. It is the shape that makes every cash-out target pay the same, which is what lets the game print a single house edge instead of a table of them.
Why every target has the same expected value
Take a target multiplier t. On a game running house edge e, the probability of surviving to t is (1 − e) / t, and cashing out there pays t times your stake.
Expected return is probability × payout: ((1 − e) / t) × t = 1 − e. The t cancels. Whatever target you pick, the expected return is 1 − e — the same number, every time.
| Target | Win chance at 1% edge | Pays | Expected return |
|---|---|---|---|
| 1.10× | 90.0% | 1.10× | 0.99 |
| 1.50× | 66.0% | 1.50× | 0.99 |
| 2.00× | 49.5% | 2.00× | 0.99 |
| 5.00× | 19.8% | 5.00× | 0.99 |
| 10.00× | 9.9% | 10.00× | 0.99 |
| 100.00× | 0.99% | 100.00× | 0.99 |
The right-hand column is the whole game. Cashing out at 1.10× nine times out of ten and cashing out at 100× once in a hundred are the same bet, wearing different clothes. One is a slow grind, the other is a lottery ticket, and both return 99 credits per 100 wagered.
The strategies, and what each one really does
Auto cash-out at a low multiplier
Setting 1.05× or 1.10× and letting it run feels close to free money — you win almost every round. What you have built is a game that pays a fraction of a percent per win and occasionally takes the entire stake. The losses are rare and exactly large enough to offset the wins.
Doubling after a loss
Martingale on crash converts a smooth stream of small wins into an eventual very large loss. Expected value per round is unchanged, so the total is unchanged. All that has moved is the shape: the distribution now has a long thin tail where the whole bankroll lives.
Waiting for a long streak of low crashes
Rounds are independent. A run of six crashes below 1.2× has no bearing on the seventh, in the same way a coin that has landed heads six times is still a fair coin. The pattern is real; the prediction is not.
Two bets, one low target and one high
Splitting a stake across two targets gives a blend of two bets with identical expected value, which produces a bet with that same expected value. It is a genuinely useful way to tune how a session feels. It is not an edge.
Crash in LarpCasino
LarpCasino ships Crash as one of six games, in a simulated casino with no real money anywhere in it. Credits are invented, nothing can be deposited or withdrawn, and everything runs on the device.
The difference from a real crash game is that the parameters are yours. The house edge is a slider from −100% to +100%, defaulting to 1%. Instant-bust chance, maximum multiplier and climb speed are all tunable — and the game states plainly that speed is cosmetic, because it is.
That makes the cancellation above something you can check rather than accept. Set the edge to 0%, run the RTP simulator over a million rounds with an auto cash-out at 1.5×, and note the empirical return. Change the target to 20× and run it again. Same number, different journey. Then set the edge to 10% and watch both drop to the same lower number.
The short version
- The crash point is decided before the curve is drawn. The climb is presentation.
- Survival probability is inversely proportional to the multiplier, which makes payout and risk move together exactly.
- Expected return at any target
tis((1 − e) / t) × t = 1 − e. The target cancels out. - Cash-out choice sets volatility, not expected value.
- Rounds are independent, so streak reading and loss-chasing change the shape of the risk and nothing else.
Frequently asked questions
- What is the best cash-out multiplier in a crash game?
- There isn't one, in expected-value terms. The win chance falls in exact proportion to the payout, so every target returns the same fraction of what you wager. Pick the target whose win frequency you prefer.
- Can you predict when a crash game will crash?
- No. The crash point is drawn independently each round, so previous results carry no information about the next one.
- Does auto cash-out improve your odds?
- It removes reaction time and emotion from the exit, which makes the game behave more consistently. It does not change the expected return of any target.
- Is crash gambling in LarpCasino?
- No. LarpCasino is a simulator rated 18+ for simulated gambling. Credits are invented and cannot be bought, won, cashed out or transferred, and there is no real-money wagering.
None of this is real.
LarpCasino is a simulated casino for iPhone. Credits are invented, nothing can be deposited, won, cashed out or transferred, and the house edge is a slider you set. 18+: simulated gambling, no real-money wagering.