Why let a machine pick the teams?
Because humans cannot help themselves. Captains pick their friends, teachers balance the groups they secretly want, and whoever writes the list last tends to end up on the weaker side. A random split removes every one of those thumbs from the scale. Nobody chose the teams, so nobody can be blamed for them, which is worth more to a five-a-side kickabout or a pub quiz than any amount of careful balancing.
It is also faster. Paste the names once, click, and the argument that usually eats the first ten minutes of the evening is over before it starts.
How the shuffle works
The tool uses the Fisher-Yates shuffle, the standard algorithm for putting a list into genuinely random order. It walks the list from the last name to the first, swapping each one with a randomly chosen name at or before it. Done this way, every possible ordering of your list is exactly equally likely, something naive approaches (like sorting by a random number) fail to guarantee.
The randomness itself comes from your browser's cryptographic random number generator, the same source used for encryption keys, not the weaker Math.random. After the shuffle, names are dealt into teams in turn, first name to team 1, second to team 2 and so on, exactly like dealing a pack of cards. When the numbers do not divide evenly, the earlier teams receive one extra player:
Where it earns its keep
- Sport. Five-a-side football, netball training, doubles pairings at the tennis club. Random sides week to week also stop the same rivalry going stale.
- Classrooms. Group work goes smoother when the groups are visibly random. Project the screen so the class watches the draw happen.
- Quizzes and game nights. Mixing colleagues or families across tables is half the point of the evening; a random deal does the mixing for you.
- Work. Splitting a team across workshop breakout groups, code review pairings or a hackathon without anyone reading politics into the result.
Running a draw people trust
Fairness is only half the job; the other half is being seen to be fair. Agree the number of teams before anyone clicks, do the draw in front of the group, and accept the first result. Re-rolling until the teams "look right" quietly reintroduces the human bias the tool exists to remove. If skill balancing genuinely matters, a common compromise is to split the strongest players across teams by hand first, then let the shuffle deal everyone else. And since the whole thing runs in your browser with nothing uploaded, a list of pupils or employees never leaves the room.
Frequently asked questions
How does it split people fairly?
The full list of names is shuffled with the Fisher-Yates algorithm, driven by your browser's cryptographic random number generator, then dealt into teams in turn like a pack of cards. Every possible arrangement of people is equally likely, so no name and no position in your list gets an advantage.
What happens when the names do not divide evenly?
The earlier teams get one extra player each. Eleven names in three teams gives 4, 4 and 3, and because the shuffle is random it is pure chance who lands in the bigger teams. If exact evenness matters, choose a team count that divides your group.
Can I re-run it if the teams look lopsided?
Yes, every click is a completely fresh shuffle. Bear in mind that a fair draw can still produce teams that look uneven in skill, because the tool only knows names. For friendly games many groups agree in advance to accept the first result.
How many names and teams can I use?
Between 2 and 20 teams, and as many names as you can paste, so a whole school year group is no problem. Each name goes on its own line, blank lines are ignored, and you need at least one name per team.
Is my list of names uploaded anywhere?
No. The shuffle happens entirely on your device in JavaScript. Names are never transmitted, logged or stored, and leaving the page clears everything, so it is safe to use with classroom or staff lists.