How to

Check a bet history

Updated 13 Nov 2026 · 5 min read

One verified bet proves one bet. A history of several hundred, checked in a single pass, says far more about how a game is run. To check a bet history well you need to know how many bets are enough, how to gather the lines and how to read a line that does not match. This guide covers all three.

Why a sample beats a single bet

Suppose a game changed the result of 1% of its bets and left the rest alone. Check one bet at random and you would almost surely pick a clean one. Check 50 and the chance that an altered bet is among them is 39.5%. Check 500 and it is 99.3%.

The sum behind this is short. Each checked bet is clean with a chance of 99%. All of them are clean with that chance multiplied by itself once per bet. What is left over is the chance of catching something. To reach 90% you need 230 bets.

So the size of the sample decides what a clean result is worth. A handful of bets shows that the method works. A few hundred start to say something about the game.

Collecting the lines of a history

  • Dice bets: the revealed server seed, your client seed, the nonce and the roll, taken from the bet history of the game. Bets made on a seed that is still live have to wait.
  • Aviator rounds: the server seed and up to 3 client seeds from the window of each finished round, plus the multiplier.
  • One bet or round per line, the values always in the same order.
  • Keep the shown result on the line. Without it the line can be computed, and there is then nothing to hold the answer against.
The bulk verifier takes up to 500 lines in one pass. A longer history goes in as several parts.

Reading the count of matches

Every line ends in one of four states. It matches: the recomputed result equals the shown one to 2 decimal places. It does not match. It was computed with nothing to compare. Or it could not be read, because a value is missing or the nonce is not a whole number.

A history in which every comparable line matches is the plain outcome for an honest game. Save the list together with the date. Seeds do not expire, and the same lines will give the same answer a year from now.

What a mismatch means, from most to least likely

  • A copying slip. One wrong character in a seed changes the whole hash, so paste the values and never retype them.
  • A result written with a decimal comma in a list that is also split by commas. The line then holds one value too many.
  • A nonce shifted by one, when the game counts from one and the list from zero or the other way round. All lines of that seed then fail together.
  • A client seed left out of an Aviator round, or a round from the other game that carries the Aviator name.
  • A dice formula that is not the generic one. Then every dice line fails, and the cause is the method and not the bets.
What remains is one line that fails among many that pass, after it has been copied again with care. That is a fair question to put to the casino, with the seeds and both results attached. Opening the line in the single bet verifier or the Aviator round verifier shows each step of the working.

What a clean history covers and what it leaves open

A clean list covers the bets on it and the seeds behind them. It does not cover bets you left out, other players or other games in the same lobby. For the casinos on the list of casinos the site recorded Aviator rounds and recomputed them, and each casino page gives the count and the date.

Nor does it say anything about price. A game can be exact to the last digit and still keep its house edge on every stake. The method for a single round is laid out in how to check that an Aviator round was fair. For a pair of seeds it is in how to verify a provably fair bet. Nonce, hash and the other terms are in the glossary.

Questions about checking a bet history

How many bets should I check to trust a game?

No count gives certainty, but the arithmetic helps. If 1% of results were altered, 230 random bets would include an altered one with a chance of 90%.

What does a mismatch in a bet history mean?

Most often a copying slip, a shifted nonce or a formula that differs from the one used for the check. A real fault is a line that still fails once all of those are ruled out.

Do all lines have to come from one server seed?

No. Each line carries its own seeds, so a list can mix bets from many seed pairs and rounds.

Can I check bets from a seed that is still in use?

No. The server seed is hidden until the pair is replaced, and without it nothing can be recomputed.

Does a fully matching history mean I will win more?

No. It shows the results were not altered. The share the game keeps is unchanged.

Tools for checking many bets

More guides

Games