Générateur d'équipes

What the random team generator does

This tool takes a list of names and splits it into groups, either a fixed number of teams (say, 4 teams from any list size) or teams of a fixed size (say, groups of 5). Internally the full list is shuffled once using a Fisher-Yates shuffle powered by the browser's cryptographic random source, then the shuffled list is sliced into consecutive chunks. Because the shuffle happens before the slicing, every person has an equal chance of landing in any team, and every possible grouping of the list is equally likely to occur.

When to use it (and when not to)

Reach for this tool whenever you need to divide a group of people into subgroups for a project, a sports match, a classroom activity, or breakout discussions. If you only need to pick one or a few winners from a list rather than split everyone up, use the random winner picker instead. If you want to pair people up two at a time rather than into larger teams, the random pair generator is the better fit. For classroom-specific advice on splitting a roster fairly, see the guide on splitting people into fair teams.

A worked example: 23 people into 4 teams

Say you have 23 students and need exactly 4 teams. 23 does not divide evenly by 4, so the split will be uneven: three teams of 6 and one team of 5 (since 6+6+6+5 = 23), or some tools default to 5+6+6+6 in whatever order the shuffle lands. The generator does not try to hide this unevenness or pad it out artificially; it simply distributes the shuffled list as evenly as the numbers allow, typically giving the "extra" members to the earliest teams in the shuffled order. If you instead choose "teams of size 6" rather than "4 teams," the tool will produce three full teams of 6 and one team of 5 left over, which is mathematically the same partition but framed differently depending on whether you care more about the number of teams or the size of each one.

For example, after shuffling the 23 names, Team 1 might get Priya, Marcus, Fatima, Diego, Wei, and Hannah; Team 2 gets Oliver, Grace, Tomás, Aisha, Noor, and Liam; Team 3 gets Sofia, Ben, Ana, Kwame, Yuki, and Ravi; and Team 4, the smaller one, gets Ella, Sam, Zara, Malik, and Ines. Nobody chose these groupings; they fell out of a single shuffle of all 23 names.

Team naming and re-roll etiquette

• Naming teams: assign names or colors (Team A, Team Blue, "The Falcons") after the split, not before — deciding names in advance and then trying to match them to specific people reintroduces bias.
• Re-rolling: if a grouping looks obviously unworkable (e.g., a team ends up with everyone who is absent that day), it is fairer to reshuffle the whole list again from scratch than to manually swap one or two names, since manual swaps quietly break the equal-probability guarantee.
• Announce the rule first: decide before shuffling whether re-rolls are allowed and under what conditions, so nobody can be accused of re-rolling until they get a favorable team.

Common mistakes

A frequent error is running the shuffle multiple times and picking whichever result "looks best" — this defeats the purpose of random assignment and can create the appearance (or reality) of favoritism. Another is manually moving one person between teams after the fact to "balance" skill or friendships; if balance matters more than randomness for your situation, that is a legitimate choice, but it should be made transparently rather than disguised as a random result.

Fairness note

Every possible way of partitioning the list into teams of the given sizes is equally likely, because the underlying shuffle treats each name identically regardless of when it was added to the list. We avoid saying results are "100% random" — instead, the guarantee is that the shuffle is driven by crypto.getRandomValues() with Fisher-Yates, which does not favor any position or person over another.