Home / Fairness
Fairness review

A claim is not evidence.
Reproduce the round.

Provably fair should describe a repeatable review, not a badge. An evaluator needs the pre-play commitment, the inputs used by the round, the revealed value and a documented calculation that produces the same result.

Review sequence

Follow one round from commitment to result.

The exact fields depend on the game format, but the reviewer’s sequence should stay understandable and auditable.

01 · Commit

Capture what existed before play.

Record the commitment shown before the player action. The review should make clear which value was fixed and when it became visible.

02 · Inputs

Preserve the complete round context.

Keep the player-controlled input, nonce or round counter, selected game parameters and the identifier that connects the evidence to the round.

03 · Reveal

Resolve the earlier commitment.

After reveal, confirm that the disclosed value matches the pre-play commitment before using it to calculate an outcome.

04 · Recompute

Run the documented calculation.

Use the disclosed inputs and algorithm to reproduce the raw result, then follow the game-specific mapping into the visible outcome.

05 · Compare

Match evidence to settlement.

Check that the recomputed result, displayed result and wallet settlement refer to the same round and configuration.

06 · Retain

Make disputes reviewable later.

Define how long round evidence remains available, how access works and what happens when a record cannot be reconstructed.

Acceptance questions

Test the explanation as well as the math.

A technically correct calculation can still fail an operational review if the evidence cannot be tied to a player-visible round or explained by support staff. Use a sandbox example, save its inputs and ask someone outside the implementation team to reproduce it from the documentation alone.

For each selected game family, document whether the result mapping introduces additional steps such as a multiplier curve, board path, reel evaluation or timed state. These format-specific rules belong beside the common commitment flow; they should not be hidden behind a generic fairness statement.

  • Can the reviewer prove the commitment preceded the player action?
  • Are every required input and transformation documented?
  • Does recomputation reach the same visible and settled outcome?
  • Can support retrieve the evidence using the round identifier?
  • Are missing, expired and mismatched records handled explicitly?

Fairness review is one gate in the broader game acceptance process. It does not replace wallet, retry, limit or operational testing.