Aviator: 97% RTP, and why no cash out strategy changes it

Aviator is the game that created a genre. A plane takes off, a multiplier climbs, and at some unannounced moment the plane leaves and everything still riding on it is gone. Cash out before that and you keep the multiplier you cashed at.

Spribe publishes a return to player of 97%, the highest of any crash game we have read and beaten on this site only by Plinko‘s 99%. It is also the figure most likely to be misunderstood, because almost everyone who plays this game believes their cash out strategy is doing something to it.

No cash out strategy changes the RTP

This is the whole article, so it is going first.

Ask around and you will be told that auto cashing at 1.2x is the safe way to play Aviator, that waiting for 10x is the greedy way, and that somewhere between the two is an optimal number. The first two are true statements about risk. The third is not a statement about return.

The intuition that fights this is a good one, and it is worth taking seriously. Cashing at 1.2x wins most rounds. Cashing at 50x wins almost none. Surely the first is better?

It is better if what you want is to win often. But the payouts scale exactly against the frequencies. You win at 1.2x about eighty times as often as you win at 50x, and when you do you are paid about a fortieth as much. The two effects cancel, and what is left is the 3% the game keeps either way.

What your strategy genuinely controls is variance. Low targets give you a long grind with small edges against you. High targets give you long droughts and rare large results. Both bleed at the same rate. Choose on temperament, not on arithmetic, and be suspicious of anyone selling you a number.

97%, and why crash games can afford it

Three percent is a thin house edge by casino standards. Every slot on this site keeps between 3.29% and 3.94%.

Aviator can run thinner than a slot because it costs almost nothing to run. There are no reels to license, no symbol art, no bonus round to build, no thirty second animation when something good happens. It is a rising number and a plane. The studio does not need a wide margin to make it pay, and a headline 97% is a genuinely effective way to get players to try it.

The catch is what that thin edge is attached to. A slot spin takes a few seconds and resolves once. An Aviator round takes about the same time, but the format invites you to bet twice at once, and it puts a running feed of other people’s winnings next to the button. Three percent of a stake you place twice as often, twice per round, is not three percent of your evening.

Two bets, and the thing they are for

Aviator lets you place two independent bets on the same round, each with its own cash out. That is not a gimmick, it is the game’s actual strategic content, and it is the only decision in it that changes anything structural.

The common use is to split: one bet on a low automatic cash out to cover the stake, one left running for something larger. It feels like hedging. It is not, in the sense that matters, because both bets carry the same edge and the combination carries it twice. What it genuinely does is smooth the experience, letting you bank something on most rounds while keeping a lottery ticket alive.

That is a real benefit if it stops you chasing. It is a real cost if it means you are staking twice as much per round as you planned to.

Provably fair, and what it does and does not prove

Spribe labels Aviator provably fair, and it is one of the few games in a mainstream casino lobby to carry the badge. The mechanism generates each round’s crash point from a seed that combines server and player inputs, published in a way that lets you verify afterwards that the result was fixed before the round began and not adjusted while it ran.

That is a real guarantee and worth having. Be clear about its scope though. Provably fair proves the round was not tampered with. It does not prove the game is generous, it does not change the 3% edge, and it does not mean the outcome was in any way in your favour. A verifiably fair coin still takes your money if you are paid less than even odds.

What the game itself says, and what it does not

The demo above runs Spribe’s own build. It takes around twenty five seconds to appear, which is long enough that we previously wrote this game off as broken and said so on this page. That was wrong, and the correction is the reason the rest of this section exists: we have now read the client.

Three places in it can hold a figure. Game Rules, read end to end. How To Play. And Game Limits. Here is what they contain.

No return to player figure appears anywhere in the client. Not in Game Rules, which runs to six headed sections. Not in How To Play. Not in Game Limits, which states the stake range and the win cap and stops there. The 97% at the top of this page exists on Spribe’s website and nowhere a player sitting in front of the game can reach it.

That is worth putting beside JetX, which we spent a long time believing was the secretive one. JetX prints its return inside the game. Aviator, the game with the better number and the cleaner disclosure on its own website, prints nothing.

The multiplier is decided before you touch anything

Two passages sit about four hundred words apart in the same Game Rules panel.

Closing the three step explainer: “Aviator is pure excitement! Risk and win. It’s all in your hands!”

And under “More details”, the third bullet: “Before the start of each round, our provably fair random number generator generates the multiplier at which the Lucky Plane will fly away.”

The second sentence is the honest one and it makes the first untrue. The number is fixed before the plane moves. Nothing in your hands changes where it stops. What is in your hands is whether you get out before it, which is a real decision about a fixed outcome rather than any influence over it.

This matters more here than it would in a slot, because Aviator’s whole feel is one of agency. You are watching a number climb and deciding. The studio has documented that the number was already written down.

“100% fair”, with no stated return

Under “Randomisation”: “The multiplier for each round is generated by a ‘Provably Fair’ algorithm and is completely transparent, and 100% fair.”

Read carefully, “100% fair” is a claim about the draw, not about the price. It says the multiplier was not adjusted to suit the house. It is entirely compatible with a game that keeps 3% of everything staked, forever, and the panel that makes the claim never mentions the 3%.

A player who reads “100% fair” and takes it to mean “you get back what you put in” has made a reasonable mistake with an unreasonable amount of help.

If your connection drops

One line, in the last section of the rules: “If the internet connection is interrupted when the bet is active, the game will auto cash out with the current multiplier, and the winning amount will be added to your balance.”

That is a good rule and it is also refreshingly short. JetX answers the same question with five separate cases depending on which phase of the round you dropped in, two of which treat an identical dropout very differently. Aviator has one case and it favours the player.

The panel adds that in the event of a hardware or software malfunction, all affected bets and payouts are void and the bets refunded, which is the standard clause and worth knowing is there.

What a demo cannot settle

One thing the free play build removes. The entire skill of Aviator, such as it is, consists of cashing out when you said you would. That is a question about you rather than about the game, and a play balance cannot ask it, because a play balance takes away the only thing that makes the decision hard.

Use it to read the panels above and to see how fast a round is. Do not use it to find out what kind of player you are.

Where these numbers come from

The 97% return, the provably fair designation and the game type were read from Spribe’s own product page for Aviator on 20 August 2026. That figure is the studio’s and appears nowhere in the game.

Everything quoted from Game Rules, How To Play and Game Limits was read off the running client on 24 August 2026, and the screenshots show the panels it came from.

Spribe’s public page states no maximum multiplier, so this page does not either. The stake range and win cap above are the demo build’s, in the demo’s currency. Real limits are set by the operator rather than the studio and vary between casinos, which is the same reason our Crazy Time page leaves them out. Check them in the client you are playing.

An earlier version of this page said Aviator’s demo would not run and that every other game page here carried one. Neither was true. The demo needed longer than we gave it, and Crazy Time is the one page with no demo, because Evolution do not offer a framed free play version of a live game show.