Unclaimed prize: how a re-draw stays as verifiable as the first

The draw ran, the winner was designated, the email went out. And nothing comes back. This is common — abandoned address, message in spam, a winner who loses interest — and it is the moment most setups stop being demonstrable.

Why this is the usual weak point

The first draw benefited from every commitment: sealed seed, announced round, frozen list. Reassignment, however, is decided afterwards, in a context where everything is already known.

Hence the temptation to improvise: "we take the next on the list." But which list, in what order, decided when? If the answer takes shape after the forfeit, it has exactly the shape of a choice — and an entrant is right to point that out.

The principle: nothing new enters the computation

A DrawSeal re-draw introduces no new value. It reuses the revealed seed and the drand value of the initial draw, unchanged. Only the set of excluded entrants changes: those already designated can no longer be.

The consequence is the one we want: the re-draw was already determined at the time of the initial draw. It wasn't computed after the forfeit, it was revealed by it. Nobody chose anything, because there was nothing to choose.

Global ordering, and why per-rank ordering fails

This is the detail that decides reproducibility, and it isn't obvious.

Designations — the initial cascade and every replacement — are ordered in a global chronology, across all ranks. At each step, the excluded set is that of the designations preceding it in that order.

Why not order by rank? Take a two-prize giveaway where ranks 1 and 2 each suffered a forfeit. The pool used for the second re-draw depends on knowing which happened first: if rank 1 was redrawn before rank 2, rank 1's replacement is among the second's exclusions, and vice versa. Rank ordering alone cannot settle it, and two people replaying the computation would get two different winners.

Global ordering removes the ambiguity. It's an invisible design constraint, without which verification would stop being deterministic.

What the rules must announce

The mechanism isn't enough if it isn't announced. Three elements, to write before entries open:

• the claim deadline — after which the prize goes back into play;
• the fact that reassignment is a re-draw, not a discretionary designation;
• the number of re-draws allowed, if capped.

An announced deadline turns an incident into a planned procedure. The same event, unannounced, becomes a decision taken under pressure — and that is how it will be read.

What an entrant can verify

Exactly what they verified for the initial draw, by the same route: with the revealed seed, the drand value and the sealed list, they replay the full cascade — initial designations and replacements — and find the same names, in the same order.

Reassignment is not an add-on at the margin of the proof. It is inside the proof.

FAQ