How Aviator works, from the first tick to the crash

Two bets in one round, and what the second one costs

Aviator lets you hold two stakes in the same round, each with its own cash out button and its own automatic target. The usual use is to take one early for something small and let the other run. The panels settle independently, so you can close the first at 1.30× and still be holding the second when the curve stops.

What the second panel changes is not your chances. It changes turnover, and turnover is what the house edge is charged on. Two stakes of a dollar each commit two dollars per round instead of one, so the expected cost per round doubles from three cents to six. An hour at twenty seconds a round is 180 rounds, $360 of turnover with both panels running and $10.80 of expected loss, against $5.40 for the same session played on one panel. A second bet is not free diversification. It is a second bet.

Auto cash out, and the number behind every target

Automatic cash out closes your position at a multiplier you set in advance. It exists mostly because human reaction time is unreliable at speed, and the target you choose while calm is not the target you press while watching a curve accelerate.

Choosing the target does not change what it is worth. On the standard crash construction a target of x wins on roughly 97% divided by x of rounds, which puts 1.20× in the money about 81% of the time and 10× about 10% of the time. Multiply either frequency by its payout and the same 97 cents comes back per dollar staked. The frequency moves, the expectation does not, and that is the whole reason strategy pages end up discussing variance instead of edge.

Autoplay, the history strip and the player counter

Autoplay repeats your stake and your target for a set number of rounds, with optional stop conditions on total win and total loss. It removes the decision, which helps if the decision was your leak and hurts if the round count was. Two hours of autoplay at a dollar a round at the fast end is 720 rounds and roughly $21.60 of expected loss, accumulated while you are not touching anything.

The history strip shows recent multipliers as coloured tiles.

It feels like data, and it is a record, but a record of independent draws is not a forecast. Read the strip honestly and it stays interesting, just not in the way it is usually read: if about 51.5% of rounds finish below 2×, then five consecutive sub-2× rounds turn up in roughly one five-round window in 28, so across a single hour of 360 rounds you should expect that run about six times. A player treating a run of five as a sign that something big is due is reacting to an event that the construction produces several times an hour. The counter listing how many people are in the round, and the multiplier at which each of them left, is the same sort of object: genuine information about other players, none about your own outcome.

The three things nobody publishes

Three fields a player would reasonably want are absent from every list we compared. The largest possible multiplier is printed nowhere. The largest possible win is printed nowhere. The ceiling on a single stake is printed nowhere either. The floor is quoted at $0.10, and even that has not been confirmed against Spribe’s own material.

Those blanks matter more than they look. A multiplier ceiling caps what a long shot can return, a win ceiling caps what a big round can actually pay you, and a table limit is precisely the wall that any doubling scheme eventually hits. Without those three numbers a reader cannot finish the calculation, and we would rather leave the cell empty than fill it with a figure that merely sounds right.

If you want the controls in your hands before a balance is attached to them, the demo does that job well and one other job badly.