Random Order Generator

Paste a list, get it back in an order nobody chose.

What a running order is really deciding

Presentation order is not neutral. Going first in a pitch means being the benchmark everyone else is compared against. Going last means being freshest in memory. In the middle is where things get forgotten. When somebody has to assign that order by hand, whatever they choose looks like a judgement, because from the outside it is indistinguishable from one.

Randomising is the cheapest way to make the order defensible. It is used for exactly this reason in tournament draws, in the order candidates speak, in who demos first at a showcase, and in the order names appear on a ballot — several jurisdictions require the ballot order to be randomised precisely because first place on the paper is worth measurable votes.

It is also the right tool for stripping accidental structure out of a list. A list exported from a spreadsheet arrives sorted by something — signup date, surname, department — and that hidden ordering leaks into whatever you do with it next.

The shuffle underneath

This uses a Fisher-Yates shuffle: walk the list from the end, and for each position swap it with a randomly chosen earlier position. Done properly, every one of the possible orderings is exactly as likely as every other, and it takes one pass.

The tempting alternative is to sort the list with a comparison function that returns a random result. It is one line, it looks clever, and it is wrong. Sort algorithms assume the comparison is consistent; feeding them a random one produces orderings that are measurably lopsided, and the specific way it is lopsided depends on which sort algorithm your browser happens to use. It is a good example of code that passes every casual inspection and is still not doing what it claims.

The random values themselves come from your browser's cryptographic generator rather than the ordinary one, and values landing in the small leftover range are discarded and redrawn so no position is even slightly favoured. That precision is unnecessary for shuffling a playlist and important for shuffling a draw.

Using it for draws and picking winners

To pick several winners without repeats, shuffle the full list of entrants and take the top few. That is better than picking one at a time and removing the winner, because it is a single operation with a single result you can screenshot, and there is no opportunity to redo an individual pick you did not like.

It also gives you an ordered reserve list for free. If the first name declines, the next in the shuffled order takes their place, and the succession was fixed at the same moment as the win rather than decided afterwards.

If you need to prove the draw was fair to other people, record the shuffled list in one go, in front of them or on video. The weak point in a public draw is almost never the randomness — it is the gap between generating the result and announcing it.

Things randomness does that surprise people

A genuinely random order will contain runs and clusters. Two names that were next to each other before will sometimes still be next to each other after; a shuffled playlist will play two songs by the same artist back to back. That looks broken and is not. Streaming services eventually stopped using true shuffle for this reason and now deliberately spread artists out, which listeners describe as feeling more random.

The number of possible orderings also grows faster than almost anyone expects. Ten items have over three and a half million arrangements; twenty have more than two billion billion. So if you shuffle a list of any length and get the same order twice, that is not luck, it is a bug — and it is worth checking.

Nothing here is stored or uploaded. Each shuffle is generated fresh in your browser, so there is no history and no way to recover a previous order. If a result matters, copy it somewhere before you press the button again.

Questions