Random Team Generator

How the teams are balanced

However the numbers fall, team sizes never differ by more than one. Twenty-two people in four teams become two teams of six and two of five; ten people in three teams become 4, 3 and 3 — never 4, 4 and 2, because a team two players short is a worse game than two teams one short. The arithmetic is simple: divide the number of people by the number of teams, and the remainder is how many teams get one extra person. Those teams are listed first.

Who goes where is completely random. The list is shuffled with a Fisher–Yates shuffle, in which every possible ordering is equally likely, driven by your browser's cryptographic random generator. The shuffled names are then dealt like cards, each to whichever team is currently smallest.

Number of teams, or people per team?

Split by number of teams when the limit is the space: four courts, three breakout rooms, two ends of a pitch. Split by people per team when the activity has a natural size — pairs for a worksheet, groups of four at a lab bench, five-a-side football.

A size rarely divides a list exactly, so you choose what happens to the leftovers. 'Make some teams one bigger' never leaves anyone on a team smaller than you asked for: seven people in pairs become a trio and two pairs, not three pairs and someone on their own. 'Make some teams one smaller' treats the size as a hard cap — seats in a car, places in a boat — so ten people in teams of at most four become 4, 3 and 3. The line under the settings previews the sizes before you generate, so there are no surprises.

Captains, and keeping people apart

Random teams are fair, but not necessarily even: a shuffle can put the three strongest players on one side. The captain option is the simplest fix. Put the people who must be on different teams at the top of the list — the strongest players, the group leaders, two students who should not work together — and tick 'First names are captains'. Each goes to a different team, and everyone else is shuffled in around them.

That spreads the top tier and leaves the rest to chance, which is usually as much balancing as a friendly game needs. Beyond it, no shuffle can know who is good at what — and that ignorance is exactly what makes it impartial.

Making random teams feel fair

The method matters as much as the maths. Show the screen while you generate, and agree before the first shuffle whether a reshuffle is allowed — 'once, if both captains ask' is a common rule. Reshuffling until the teams look right is just picking teams with extra steps.

Random selection also removes the worst part of picking sides in school sport: being chosen last in front of everyone. Nobody is ranked; the teams simply appear. It mixes friend groups, which is often the point in classrooms and workshops, and it is fast — a class of thirty is split in a fraction of a second.

Copy teams produces plain text, one team per line — 'Red (4): Amara, Diego, Isla, Leo' — that pastes cleanly into WhatsApp, Slack, Teams or an email.

Limits and privacy

The list holds up to 1,000 names. Duplicate names are kept as separate people, since two players can share a name. At least two teams are always made, and asking for more teams than there are names gives a clear message rather than empty teams.

The names stay in your browser. They are saved in its local storage so the squad is there next week, and they are never uploaded — the shuffle, the dealing and the copy all happen on your device.