Your bets
Was it a bad run, or a bad game?
Everyone who has had a rough session has wondered. The honest answer is a number: over how many rounds, against what payback, with how much spread. Give it the seeds the casino already shows you and it rebuilds every round you played — then tells you whether the result was remarkable or completely ordinary.
From your seeds
Rebuild a session you actually played
Loading games…
How to read it
Standard errors, not percentage points
Why a percentage gap means nothing on its own
Getting 85% back where the game certifies 96% is unremarkable over two hundred rounds and extraordinary over two hundred thousand. The gap only becomes meaningful once you divide it by how much a result of that length is expected to wander, which depends on the game's own volatility.
What the sigma figure is
Roughly: under two, ordinary. Two to three, unusual but the kind of thing that happens to somebody every day across a platform this size. Above three, worth checking the log itself — a history mixing several games or several modes will read as extreme no matter how fair every one of them was.
It needs a measured game
The comparison uses the game's real volatility, which comes from reading its lookup table. That exists for 263 games so far. Without it your session still rebuilds in full; only the "was this unusual" verdict is withheld, rather than guessed at.
No export, no account access
The input is the three things a provably fair casino already displays: your server seed once rotated, your client seed, and the nonce your bets ran from. Nothing is scraped and no account is touched. The rounds come back from the operator's own public fairness endpoint, and the maths runs on your machine.
A result inside normal variance does not mean a game is generous, and one outside it does not prove anything is wrong — provably fair verification is a separate question, and there is a tool for that. This page describes what happened to you, and how surprising it should be.