Team Splitter

Why splitting by hand goes wrong

Three things happen when a person divides a group. The first few names get placed thoughtfully and the rest get dumped. Whoever is doing the splitting unconsciously balances for something — ability, friendship, volume — and the result stops being random while still being presented as random. And the last team ends up short because the arithmetic was done in someone's head.

None of that is dishonesty; it is just what happens. A tool avoids all three, and more importantly it is seen to avoid all three, which is usually the actual problem being solved.

How the split works

The list is shuffled with Fisher-Yates using your browser's cryptographic random number generator, so every possible arrangement is equally likely. People are then placed one at a time into whichever team currently has the fewest members.

That second step is what keeps sizes even. When a group does not divide cleanly, the remainder is distributed rather than concentrated: eleven people into three teams gives 4, 4 and 3. The preview beneath the controls states the split before you press the button, so you can adjust the number of teams if the shape is not what you wanted.

Splitting by team count or by team size

Both are offered because people arrive with different constraints. A classroom usually knows how many groups it wants. A five-a-side session knows how many players go in each team and needs to find out how many teams that makes. Switch between them with the dropdown.

Keeping people together, and apart

This is the feature that makes the tool worth using over a shuffle, and almost nothing else offers it.

Together pairs are bundled before the split and moved as a single unit, so siblings stay in one group, a driver stays with their passengers, and a pair working on the same project are not separated. Chains work too: if you pair A with B and B with C, all three end up together.

Apart pairs are checked as each team is filled. Two people who cannot be productive together, two strong players who would unbalance a side, a pair who need separating for reasons you do not have to explain to a web page.

Because bundles are placed largest-first into the smallest team, adding constraints does not wreck the balance the way you might expect. If a constraint genuinely cannot be satisfied — usually because there are too few teams — the split still happens and a note tells you exactly which request was dropped.

Where this gets used

Classrooms

Group work, project teams, sports lessons. The keep-apart option quietly solves a problem every teacher recognises, without anyone in the room needing to know it was used.

Sport and clubs

Pick-up games, training drills, tournament pools. Split by team size when the format fixes how many play per side.

Workshops and events

Breakout groups, tables at a dinner, icebreaker teams. Naming the teams helps more than it sounds: people organise themselves faster around a name than a number, and it gives a group of strangers something to be, immediately.

Households and chores

Dividing tasks, deciding who does what. The same reason as everywhere else: the tool takes the blame for the outcome.

Practical notes

• Names in the pair boxes must match the list exactly. Nicknames and misspellings are treated as different people.
• Use first names where you can. Shorter names make the result cards much easier to read aloud.
• Reshuffle freely. If a draw looks wrong, press it again — that is still random, as long as you are not re-rolling until you get the answer you wanted.
• Copy all puts the whole set of teams on your clipboard, formatted for pasting into a message or a document.

Related tools

To pick one winner instead of splitting a group, use the Name Wheel. For inventing team names on their own, the Username Generator in wordplay mode does the same job with more control.

Common questions

How does it keep the teams even?

People are shuffled, then placed one at a time into whichever team currently has the fewest members. When the group does not divide evenly, the remainder is spread out rather than piled onto one team: eleven people into three teams gives you 4, 4 and 3, never 5, 3 and 3. The preview under the controls tells you the split before you commit.

Can I keep two people together, or apart?

Yes, and it is the main reason to use this instead of shuffling a list by hand. Open the section below the button and enter one pair per line. People who must stay together are bundled before the split, so they move as a unit. People who must be separated are checked as each team is filled.

What if a constraint cannot be met?

You get told. If you ask for three people to be on different teams but only create two teams, the split still happens and a note appears above the results explaining which request could not be honoured. We would rather say so than quietly ignore it.

Are the teams actually random?

The shuffle uses your browser’s cryptographic random number generator with Fisher-Yates, which is the standard unbiased shuffle. Every arrangement is equally likely, subject to whatever constraints you set. Press reshuffle for a completely fresh draw.

Can it name the teams for me?

Tick the box and each team gets a generated name instead of a number, built by our username engine using animal, myth and nature vocabulary. It is more useful than it sounds: named teams are easier to talk about than Team 3 and Team 4.

Is my list stored anywhere?

No. Everything runs in your browser and nothing is uploaded. Close the tab and it is gone.