Is Aviator rigged? How the lab checks the provider claim

The game is rigged against players beyond its stated return.

Open: measured here

Open question: the lab answers it with counted rounds, game by game.

Reviewed 8 Oct 2026

Short answer

This is what the lab measures. Every game has a built-in house edge; the question is whether actual results match the provider's stated return. We publish measured return with its margin of error so you can see the answer, not take ours.

Why

Stated RTP is a long-run average; measured return converges to it if the game behaves as claimed.

How the lab tests this

Measured RTP with a 95% confidence interval compared with the provider claim.

What the lab measures

Every game here has a stated return, a provider claim. The lab records rounds from the provider's own demo or public result stream and works out the return those rounds paid. For crash games the headline is Observed RTP @ 2x (fixed cash-out), the average return if every round were cashed out at 2x. For games the lab plays itself, it is the return of one fixed way of playing, kept the same for every round.

A figure is published only with its sample size, its as-of time and its 95% interval, and only once the sample passes 10,000 rounds. Until then the game page shows what is still needed, never a placeholder.

What "within the interval" means

No sample lands exactly on the long-run return. A 95% interval is the range of long-run returns consistent with the rounds counted so far. If the provider claim lies inside it, the results are within the interval: consistent with the claim. That is not proof the claim is exact, only that the rounds give no reason to doubt it.

A worked example, under an assumption: suppose Aviator's stated 97%, a provider claim we have not yet sourced first-hand, is exactly true and every round is cashed out at 2x. Each round then returns 2 with probability 0.485 and 0 otherwise, so one round's return has a standard deviation of 2 × √(0.485 × 0.515) = 0.9995. The interval is the measured return ± 1.96 × 0.9995 ÷ √n.

Half-width of the 95% interval at a 2x cash-out, assuming Aviator's stated 97%, a provider claim we have not yet sourced first-hand
Rounds (n)	Half-width, in percentage points
10,000	1.96
40,000	0.98
160,000	0.49

Four times as many rounds halve the width. That is why the lab waits for a large sample before it says anything: with too few rounds, the interval is wide enough to contain almost any claim.

What a result outside the interval would mean

If the claim falls outside the interval, the rounds counted are unlikely if the claim is true. That is a reason to look closer, not a verdict. When a game behaves exactly as claimed, a 95% interval still misses the true value about 1 time in 20, and with many games and many months some misses are expected by chance.

The lab's answer to a miss is to check its own accounting first, keep counting, and publish what it finds with its sample size, time and method; the sealed raw days behind it can be requested and recounted. A game page held for review says so, with the reason.

A return within the interval does not mean a session will return the claim either. A long-run figure describes many rounds; a single session can lose far more than the house edge suggests.

Is Aviator rigged? Is Spaceman rigged? Is Mines rigged?

Is Aviator rigged? Aviator is provably fair, so each revealed round can be checked against the seed hash shown before it. That shows a round was not changed after bets went in. Whether its long-run return matches its stated return, a provider claim, is what the lab measures on the Aviator page.

Is Spaceman rigged? Spaceman is a certified RNG game, with no seed shown to players. A single round cannot be checked; only the pattern over many rounds can, and that is the lab's measurement on the Spaceman page.

Is Mines rigged? Mines is provably fair like Aviator, so a revealed board can be checked against its hash, and the lab measures the return of fixed ways of playing. For games with an Unverifiable badge, such as Chicken Road, we found no public mechanism for checking a single result, pending a first-hand check of the provider's documentation; what can be measured is the pattern of results over many rounds. Each game's lab status is in the list below.