Random Team Generator

Two ways to split

Choose a number of teams and everyone is dealt between them. Choose a group size and as many groups are created as it takes to fit everyone. The second is what you want for workshops and breakout rooms, where the room dictates the group size rather than the number of groups.

Why nobody ends up in a group of one

Eleven people into three teams does not divide evenly. A lot of tools handle this by shuffling the list and slicing it into three, which quietly leaves the last team short.

This deals instead: the shuffled list is handed out one person at a time, round-robin, so eleven people become 4, 4 and 3 rather than 4, 4 and 3 by accident or 5, 5 and 1 by bad luck. No group is ever more than one person larger than any other.

What random teams are good and bad at

Random splitting is the right answer when the point is to mix people up - breaking up the usual cliques in a classroom, rotating who works with whom, or making sure nobody can claim the teams were stacked.

It is the wrong answer when ability matters. Random teams in a competitive game routinely produce a lopsided match, because the odds of the strongest four players landing together are not negligible. If you want balanced teams, seed the strongest players first and split them deliberately, then randomise the rest.

Settling the "that was rigged" argument

Whoever runs the split always gets accused of it eventually, particularly if they end up on the team that wins. Each split produces a receipt containing the full name list and the seed, so anyone can re-run it and confirm you did not shuffle until you liked the look of it. How that works.

To pick individuals rather than split a group, use the name picker. For a knockout competition, the bracket generator seeds a full draw.