Random Name Picker Guide

A random name picker takes a list of names and selects one at random, for a raffle, a classroom pick, a giveaway, or deciding who goes first. The idea could not be simpler. Whether a particular tool is actually trustworthy is a much more interesting question, and the answer usually has nothing to do with the quality of its random number generator.

Two different jobs

Before choosing a tool, work out which of these you need, because they have different answers:

A random result. You need one name and nobody is going to question it. Deciding who reviews a document, picking a test case, choosing a lunch spot. Any picker will do. Honestly, a spreadsheet formula will do.
A demonstrably random result. Other people are affected by the outcome and at least one of them may not take your word for it. A prize, a place on a team, an order that confers advantage. Here the picker is the easy part and the evidence is the whole problem.

Almost every complaint about a name picker comes from using a first-job tool for a second-job situation.

What goes wrong with a simple picker

Most name pickers run entirely in the browser: the list of names and the “random” result both live on the visitor’s own screen. For a throwaway pick that is fine. It is what the free wheel on this site does, deliberately, and why it needs no account and sends nothing anywhere.

It becomes a problem the moment a result matters to somebody who was not watching. Three specific weaknesses:

Nothing stops a reload. An organiser can refresh until a preferred name comes up and screenshot only that spin. The generator was flawless; the process was not.
There is no record. When it is over you have whatever was on screen when somebody took a photo. If a result is disputed a week later, there is nothing to appeal to.
Every viewer gets their own result. If a picker generates locally, two people opening the same page see different winners. There is no shared event, only a set of individual ones.

A fourth, more technical weakness is worth knowing about because it is invisible. Many pickers map a random number onto a list with a plain modulo, which makes the first names on the list very slightly more likely to win. The effect is far too small to notice by eye and completely avoidable. The explanation is in how random number generators actually work.

What SpinForFree does differently

The pick happens on the server, using a cryptographic random source, before the result reaches any browser. A viewer reloading their own screen is re-reading a decision that has already been made and recorded.
Every draw is logged with a timestamp, and spinning again does not overwrite it. Redraws stay in the record alongside the original.
Everyone watches the same spin. One link shows the identical result to everyone who opens it, landing at the same moment.
Odds are weighted and visible. A slice drawn three times bigger really is three times more likely, so the picture matches the probability instead of hiding it in a settings panel.
The free wheel corrects modulo bias too. Rejection sampling on crypto.getRandomValues, not Math.random and a remainder.

Adding names without typing them all

Paste a whole list at once, one per line or separated by commas or semicolons, which means a column copied out of a spreadsheet and a comma-separated line from a chat message both work without reformatting. On an account-backed wheel duplicates are skipped automatically, so pasting a messy roster twice will not silently double someone’s odds.

For a giveaway, open public entries instead and let people submit their own name with an optional proof screenshot. You approve each one before it reaches the wheel, so you get a fixed reviewable list rather than a comment thread that changes while you read it, and no transcription mistakes for anyone to attribute to malice.

How many names can it handle?

The draw itself is unaffected by list size; the labels are what run out of room. Past roughly 150 entries the slices are too thin to print text in, so the wheel drops the labels rather than overlapping them. A 400-entry raffle draws correctly but stops working as a picture.

If the audience needs to see individual names, either keep the list under about 100 or draw in rounds: spin to pick a group of ten, then spin again among those ten. Two spins, both visible, and every name legible in the second one.

Avoiding repeat winners

With 30 names and 10 draws, the chance that at least one name comes up twice is better than even. People consistently guess much lower, so a second win reads as suspicious when it is entirely ordinary.

If each name should come up once, do not argue with probability. Turn on elimination so each winner leaves the wheel. Turn it off when everyone should be eligible every round, like cold-calling in a lesson. Either is fine; what causes trouble is deciding which after somebody has already won twice.

Checking a picker before you trust it

Reload a finished result. If it changes, nothing was recorded and there is nothing to appeal to later.
Open the same link on a second device. If the two disagree, there is no shared draw, just two independent ones.
Look for a history. A tool that shows only the current winner cannot help you when somebody disputes the third draw of five.
Check the slices match the stated odds. Weighting you cannot see is weighting nobody can verify.
Run it 300 times with three names and count. Real bias shows up fastest with few options. Expect unevenness; expect 40/30/30 to be suspicious.

Where name pickers get used

Raffles and giveaways, classroom cold-calling and group assignment, meeting and stand-up order, tournament brackets, gift exchanges, and everyday decisions where a coin does not have enough sides. Each has settings that suit it. See the guides for classrooms, teams, giveaways, tournaments, Secret Santa and everyday decisions.

If you are still deciding whether a wheel is even the right instrument, as against a dice, a hat or a shuffled spreadsheet, that comparison is in wheel, dice, hat or shuffle.

Getting started

Try the free wheel, which needs no account and sends nothing anywhere. For a shared link, a server-side draw and a history you can point at, see the how-to guide or create a free account.