Aviator’s 97% RTP, and the money that number describes

Three percent of turnover, not three percent of your deposit

This is where most readers lose the plot, and it is not their fault, because the way the figure is normally quoted invites the mistake. The 3% is charged on every dollar that passes through the game, not on the money you put in once. A deposit that gets staked, won back, and staked again has been charged twice.

Put a hundred dollars in and bet a dollar a round. After 100 rounds you have staked $100, turnover equals your deposit, and the expected loss is $3, which is genuinely three percent of what you deposited. Keep going and the two numbers separate fast. After 500 rounds the turnover is $500 and the expected loss is $15, or fifteen percent of the deposit. After 1,000 rounds it is $30, thirty percent. After 2,000 rounds it is $60, and you are expected to have lost more than half of a deposit you never increased. Nobody topped up, nobody raised a stake, and the only quantity that grew was the number of times the same money went round the loop.

The same deposit, the same stake and the same game, with only the round count changed.

Carry that forward and you get the expected life of a bankroll. At three cents of expected loss per round, a hundred dollars funds roughly 3,333 rounds of one dollar flat betting before the edge alone has consumed it. Nothing about that number promises you will last that long, because variance can end a session in twenty minutes, and nothing promises you will not last longer. It is the centre of a distribution, not a schedule.

What an hour of Aviator costs

Round length does the rest of the work. A round takes between ten and thirty seconds, which is 360 rounds an hour at the fast end and 120 at the slow end.

At a dollar a round that is $360 of turnover in a fast hour and $120 in a slow one, so the expected cost of an hour lands between $3.60 and $10.80. Open the second bet panel that the mechanics page describes and both ends double, to somewhere between $7.20 and $21.60. At the quoted ten cent minimum the same hour costs between 36 cents and $1.08, which is the most useful thing the minimum stake tells you.

An hourly rate is a better way to think about this game than a percentage, because the percentage hides the clock and the clock is the whole difference between Aviator and a game you play twice a minute.

The share of rounds that end below 2×

Here the arithmetic needs a stated assumption. Spribe does not publish the distribution of crash points, and it is not in our record, so what follows uses the standard crash construction, in which the chance of the curve reaching a multiplier of x is the return figure divided by x. Treat it as a model of the game rather than a reading off the game.

On that construction, the curve reaches 2× on 97 divided by 2, which is 48.5% of rounds. So 51.5% of rounds end below 2×, and a player taking a flat double is on the losing side of a slightly unfair coin every single time. At 1.20× the hit rate is about 81%, at 3× about 32%, at 5× about 19%, at 10× about 10%, and at 100× just under 1%.

Losing streaks fall out of the same number. If a 2× target misses 51.5% of the time, three misses in a row happen about 14% of the time, five consecutive misses about 3.6% of the time, seven of them about 1%, and ten in succession roughly once in every 762 attempts. A player who runs 360 rounds in a fast hour will meet a five-round losing run several times before the hour is out. Streaks like that are not evidence that anything is broken and not evidence that anything is due, which is the same conclusion the doubling page reaches from the money side.

Why every cash out target has the same expectation

Multiply the hit rate by the payout and the target cancels itself out. At 1.20× you win 80.83% of the time and collect 1.2 times your stake, and 0.8083 times 1.2 is 0.97. At 10× you win 9.7% of the time and collect ten times your stake, and 0.097 times 10 is 0.97. At 100× it is 0.0097 times 100, which is 0.97 again.

Ninety seven cents back per dollar, whatever you set. The target chooses how the losses arrive, not how many there are.

That is the honest frame for every scheme built on cash out points, and it is why the strategy page talks about the shape of the ride rather than the destination. A low target gives you many small wins and occasional painful gaps. A high target gives you long droughts and rare large returns. Both end up at 97%, and the difference between them is entirely about which of those two experiences you can sit through without changing your stake, which is where most bankrolls are actually lost.

Why comparing this to a slot goes wrong

A 97% slot and a 97% crash game are not the same product with the same label. Round length and staking behaviour differ enough that the identical percentage produces very different bills.

The comparison also breaks on control. In a slot the return is fully determined by the machine, while in Aviator the player chooses the exit multiplier, which changes the distribution of outcomes without moving the average at all. That is an unusual combination, and it is the source of the persistent belief that skill can be applied here. Skill can be applied to when you stop and how much you stake. It cannot be applied to the 3%, because the 3% is charged before your decision is taken and does not consult it.

A last boundary is worth stating plainly. A return figure describes a very long run of rounds and says nothing whatsoever about the next one, which resolves at 0× or at your target and never at 97%. It is not a rebate, not a schedule and not a promise, and none of it is a claim about whether an operator will actually pay a withdrawal, which is a separate question with its own evidence. The check that does cover the individual round is on the fairness page, and the fields the provider leaves blank are collected on the Spribe reference.