How random number generation works

Random number generation is the foundation of fair draws, cryptography and simulations. Not all “random” numbers are equal: many systems use pseudorandom generators that only look random if you do not know the secret seed.

A transparent draw should be verifiable afterward and unpredictable beforehand. Unpredictability needs an input nobody can choose at lock time.

Entropy as an anchor

The chaotic image entropy.jpg has a public SHA-256 hash. Anyone can download the file and recompute it. That is a stable domain anchor, not a future secret.

Commit-then-beacon (v2)

EntropyDraw locks the range or name list (plus description), assigns a future League of Entropy drand round, then waits until that public beacon exists. The draw hash mixes the image anchor, locked fields and beacon randomness. Until the round is published, the number cannot be computed.

Rejection sampling

Converting a hash with simple modulo can bias some outcomes. Rejection sampling discards values outside a fair window until a valid number is found — every integer gets equal probability.

Why this matters

Integrity means “these published inputs produce only this number.” Unpredictability means “those inputs were not all known or choosable.” Share the /d/{id} link at lock time, and for name lists publish the list commit hash before you draw.