Begin with the part that shapes everything below. No staking pattern, cash-out rule or waiting game changes the expected return of a crash round in a model with a fixed edge and un-rounded payouts. The edge is built into the payout, and nothing you do between rounds reaches it. What the popular systems genuinely change is the texture of a session: how often you win something, how big the swings get, and how fast a bankroll can empty when a bad run arrives.
That distinction is worth more than it sounds. Two players can run the same game for an hour with the same average loss and have completely different evenings. Say one of them sets a low target and wins most rounds, so it feels like control. The other sets a high target, wins rarely, and either goes home early or goes home with a profit. Picking between those is a real choice. Picking a strategy that "wins" is not on the menu.
The strategies people actually run
The most common one is the low auto cash-out, somewhere around 1.2× to 1.5×, left running while the player does something else. Most rounds land. The balance climbs in small steps and then surrenders a chunk of it when three or four rounds in a row crash early. Around 48.50% of rounds reach 2× at a 3% edge in the figures below, and the lower targets clear far more often than that, which is exactly why the approach feels productive before it feels expensive.
The 2× flat bet is the other default. Roughly half the rounds win, each win doubles the stake, and the swings sit in the middle of the range. Chasing 10× or higher flips the picture completely: long dry spells, then an occasional payout that may not cover the preceding losses.
Here is what a hundred $1 rounds look like at four targets. Everything in the table comes from the same 3% edge, and the expected return column is the one to read first.
| Target | Chance of reaching it | Expected return per $1 | Chance of profit after 100 rounds | Typical swing (SD) |
|---|---|---|---|---|
| 1.2× | 80.83% | 97.00% | 25.37% | 4.7233 |
| 1.5× | 64.67% | 97.00% | 35.40% | 7.1701 |
| 2× | 48.50% | 97.00% | 34.43% | 9.9955 |
| 10× | 9.70% | 97.00% | 37.74% | 29.5958 |
Figures above are from our model at a 3% edge, not from any casino's game.
The third column stays flat down the table while the others move a long way. That is the whole argument in one view. What a lower target reliably buys is a higher share of winning rounds, with the losses then arriving in lumps. Whether a short session finishes in profit is a separate question with its own answer, and the fourth column gives it for these four cases only; it does not follow from the win rate, and there is no simple rule that ranks targets by it. Our guide to whether there's a best cash-out multiplier works through the trade-off in more depth, including why the equal returns depend on payouts not being rounded.
Doubling after a loss, and its gentler cousin
Martingale is the system everyone tries once. Lose a round, double the next stake, keep doubling until a win clears the sequence and leaves you one unit up. The recovery arithmetic only works cleanly at a 2× payout, where a win returns everything staked so far plus a unit. At 1.5× it does not, and people who run martingale on a low auto cash-out are usually surprised when a win fails to put them back where they started.
The real problem is the size of the run it has to survive. A given stretch of ten consecutive losses at a 2× target comes out near 0.13% in our model, which sounds like something you would never meet. Across a long session, with thousands of other people playing at the same time, runs like that turn up constantly. Each loss doubles the next stake, so the required bet grows far faster than the bankroll behind it, and either the balance or the table's maximum stake ends the sequence before the win arrives.
D'Alembert is the softer version: add one unit after a loss, remove one after a win. It carries the name of the 18th-century French mathematician Jean le Rond d'Alembert, though the system as played today is a later invention. Stakes drift upward during bad runs, slowly enough that the damage takes a while to notice, and the drift does nothing to the chance of the next round.
Placing two bets in the same round is sometimes sold as a hedge, one early target to bank something and one late target to chase. Both bets ride the same crash point. When a round ends below the early target, both lose, so the arrangement is one bigger bet with a split payout rather than two independent chances.
Waiting for a streak to end
The "wait for five low rounds, then bet big" idea is the most intuitive and the least useful. Each round draws its own result from the same process, and the history strip above the game carries nothing about what comes next. That is the point our guide to whether Aviator is rigged works through at the level of the actual calculation, and it is also why the prediction channels covered in the predictor guide have nothing to analyse.
Sitting out rounds does reduce how much you stake, which reduces your expected loss along with it. Raising the stake on the round you decided was due can partly or fully offset that reduction, depending on how many rounds you skipped and how far you raise the bet, and it concentrates the risk into a single outcome either way.
Where bankrolls actually break
Stake size relative to bankroll drives how a session ends as much as the target does. Say you take a $20 bankroll, bet $1 a round at a 2× target, and run 200 rounds without stopping. The chance of going broke before the end comes out at 27.21% in our model. Shrink the bankroll or raise the stake and that number climbs quickly, which is the mechanism behind most short sessions that end badly, with no bad strategy required.
One fixed-seed run of exactly that setup played 200 rounds, ended at 30.00, and dipped 16.00 below its own high point along the way, with a longest losing streak of 11 rounds. That is one sample and a different seed gives a different story, which is rather the point. You can run your own in the crash simulator and watch where the drawdowns land.
What genuinely improves a player's position is unglamorous, and none of it is an edge. A fixed amount you are willing to lose, decided before the first round. A time limit as well as a money limit. A stopping point, win or lose, that you set in advance rather than in the moment, and a firm rule against raising stakes to win back what the last few rounds took. Those do not shift a single probability on this page, and they do not make continued play cheaper. They decide how much of your money the probabilities get to touch.
If you have been looking for a winning system because the last few sessions cost more than you planned, that is the signal worth acting on. The responsible gambling page lists support services that can help.
