What it’s for

Public allocations that went wrong, grouped by what went wrong, with the source next to each one. And how a verifiable number is used so it does not happen again.

what this proves     :  that the number was not chosen, and when it came out
what this does NOT   :  who was on the list
prove                   who held the ticket
                        that you get paid what you won

The pattern, in one line

In six of the cases below there was a notary, appointed supervisors, witnesses in the room or a live broadcast.

In five of those six, the control failed or was irrelevant.

The one that worked did so because the error was visible at a glance to people who knew the rules by heart and were watching that exact second. That is not a method: it is luck. And in a housing allocation there is nobody like that in the room.

Where it hurts, and none of this is a game

In all of them the sentence that settles the problem is the same: «I did not choose it, and you can check.»

How it is used, in four steps

The body does not ask us for permission, does not hire us and sends us no data at all. We publish the number; the allocation is done and published by the body itself.

1. the body publishes its list of applicants BEFORE the turn   <- its own commitment
2. it takes the value of the announced turn
3. it uses that as the seed of a shuffle IT publishes
4. anyone reproduces the order from the list + our number

Step 1 is theirs and it is essential: the list of applicants belongs to the body, not to us. Without that prior commitment an honest number does not fix a rigged universe — which is exactly what failed in the first group below.

The cases, grouped by what failed

Grouped on purpose: one by one they look like isolated misfortunes; grouped, you can see they are six failures that repeat across every country and every decade.

1 · The universe was wrong before anything started

The event itself was flawless. The list it ran on was not.

What would have changed: the universe is declared inside the signed file and is the same for everyone, so it cannot be widened or trimmed afterwards. The two rows marked 🔴 are not covered: there what was tampered with is the roll of applicants, and that has to be published and committed by the body before the turn.

2 · The machine or the program was rigged

What would have changed: the value is not produced by the operator. It is derived from a seed published as a commitment before the turn and revealed afterwards, plus a round from an external public network. There is no physical object to swap, no in-house generator to touch, and nobody who gets to choose.

3 · The honest mistake — and it is the most common

None of these is fraud. All of them did the same damage.

This is the whole argument of this site: it takes no bad faith for an allocation to go wrong, and it takes no accusation to fix it. It takes a result that can be recomputed from the outside. Five of these six would have shown on day one, and 1969 is the best example of all: it was caught precisely because the data was public and somebody could do the arithmetic.

4 · The result existed, and nobody outside could check it

This group is the one that best explains why this exists. In Bulgaria, South Africa and the Philippines there probably was no cheating — the investigations found nothing. It makes no difference: with no way to check from the outside, the operator's honesty and the suspicion are worth exactly the same, and the suspicion never dies.

5 · The failure was outside the allocation

Listed because they are real and because they mark the limit: this does not solve them.

6 · They gave up on chance because they could not make it credible

This is the ending this site is trying to avoid: that chance gets discarded because it cannot be demonstrated, and is replaced by something a person decides.

The exception, and it has to be read as one

In December 2021, in the Champions League round-of-16 draw, a ball went into the wrong pot and another was not put back. Club officials present in the room caught it on the spot, and the whole thing was redone three hours later. (ESPN)

It worked because the error was visible at a glance to people who knew the rules by heart and were watching that second. In a housing allocation there is nobody like that in the room, and in the other five cases with human oversight the oversight was no help.

What this does not solve, once more

Repeated on purpose, because it is what separates this from a promise:

what this proves     :  that the number was not chosen, and when it came out
what this does NOT   :  who was on the list
prove                   who held the ticket
                        that you get paid what you won

All of group 5 and the two rows marked 🔴 are here for that reason. The roll of applicants is published and committed by the body; the ticket and the payment are theirs. We publish a number with its proof. That is all.

What was left out, and why

More cases were researched than are listed above. These did not make it in:

A poorly supported case does not weaken only itself: it weakens the other forty. That is why the list above is shorter than the research behind it.

What utc24h is · See the current numbers