Every giveaway ends the same way: a comment appears saying it was rigged and an insider won. After the results are out you cannot prove otherwise — whatever you show is the organiser’s word against theirs. The only way is to prepare in advance.
Por qué una captura no demuestra nada
The usual scenario: the organiser runs a randomiser, screenshots it and posts the image. From a participant’s point of view that proves exactly one thing — that a randomiser exists.
Nobody knows how many times it was spun before. Nobody knows whether the winner was on the list before the draw or added afterwards. A video is slightly better but does not solve the main issue: the participant list was known only to the organiser.
The problem is not the tool but the order of events. As long as the list can be changed after the result is known, fairness is unprovable in principle.
Qué hace verificable un sorteo
Three things are needed, all of them before the winner is announced.
1. The list is committed before the draw. Not “saved on my computer” but published so that it cannot be altered retroactively. In practice you publish a fingerprint of the list — a short string that changes with any edit.
2. The result is drawn once. If a draw can be repeated, every earlier attempt stays invisible. The mechanism must physically forbid a second run.
3. The computation is reproducible. After the draw a random seed and the formula are published. Any participant takes the list, the seed and the formula and gets the same winners — recomputing rather than trusting.
That is exactly how the winner picker here works: the page is created before the contest, gets a permanent address, the draw happens once, and the seed and line numbers are published next to the result.
El orden de actuación
A week before. Collect the participant list — names or profile links, without phone numbers or addresses.
When you publish the rules. Create the draw page and put the link in the contest post. At that moment nobody knows the winner, including you — and the timestamp shows it.
At the announcement. Press the button. The result appears on the same page everyone has already seen.
Right after. Publish the participant list itself. Without it the fingerprint has nothing to be compared against.
Errores que provocan discusiones
The link appeared after the results. The most common one, and it voids the whole construction.
The list was never published. A fingerprint without a list is a checksum without a file.
Rules were added along the way. “The winner must be subscribed”, introduced after the draw, looks like filtering out an inconvenient winner.
No reserves. Winners fail to respond more often than you would think. Without reserves chosen in advance you have to run a second draw — and that one really will look suspicious.
Qué hacer si el ganador no responde
Write the rule into the terms before the draw: how many days to respond and what happens next. The usual wording is 72 hours, after which the prize passes to the first reserve.
Then moving to a reserve is not the organiser’s decision but the execution of a published rule. The difference matters: in one case you are accused of manipulation, in the other there is nothing to accuse.
Si los participantes van numerados
When a contest runs on comment numbers, no list of names is needed at all. The draw by numbers is more convenient: only the range and the result reach the server, and no personal data remains.
For the show, the wheel of fortune remains. It proves nothing, but the audience enjoys it: spin it for the stream and settle the result with a verifiable method.