Splitting a list into groups that don't all have to match

Most team-splitting tools ask one question: how many groups do you want? That works fine when every group should come out the same size, but plenty of real activities don't work that way. A mentoring pod might want three to five people. A study group might do best at two or three. A breakout discussion might need groups anywhere from four to six so each one has enough voices without turning into a lecture. Forcing those into perfectly even teams either strands an awkward leftover pair or pads a group past the size that actually works for the activity. A variable-size group splitter solves a different problem than a fixed-team divider: instead of picking a group count, you pick a size range, and the tool figures out how many groups that range actually produces.

You set a range, not a count

The variable-size group splitter takes your full list and two numbers: a minimum group size and a maximum. Set them to two and four, and the groups the tool produces land in that window — no lone singletons, no crowd of six — with one exception at the end of the list, covered below. The tool doesn't ask how many groups you want, because that's the number it's solving for. Fourteen people with a two-to-four range might come back as four groups, three groups, or something in between — whatever the random chunk sizes add up to.

Under the hood, the list is shuffled first, then sliced into consecutive chunks whose sizes are drawn randomly from your min-max window as it goes. That randomness in the chunk sizes is the whole point: two runs over the same fourteen people can come back with a completely different mix of three-person and four-person groups, so nobody can predict or angle for a particular group shape.

No orphaned leftover group

The awkward case with any chunking approach is what happens to whatever's left over at the end of the list — a lone person who didn't fit a full-size group. The splitter handles that automatically: if the last group would fall below your minimum size, it gets folded back into the group before it rather than standing alone. You'll never end up with a "group of one" dangling off the end just because the math didn't divide evenly. The trade-off is that the very last group in the result can occasionally run a little larger than your stated maximum, which is the honest price of never leaving somebody by themselves.

Reroll instead of re-typing

Because the grouping is random every time, a result you don't love — say, the same two people keep landing together, or one group feels lopsided — is one click away from a fresh attempt. Hit split again and the tool reshuffles the whole list and re-chunks it inside the same size range. There's no need to manually drag people between boxes to fix a group you're unhappy with; it's usually faster to just reroll the whole arrangement than to hand-edit one result.

This also makes the tool forgiving for recurring sessions. A weekly discussion group that meets with the same two-to-four range every time doesn't need a new setup each week — reroll at the start of each session for a fresh random mix. The splitter keeps no memory of last week's groups, though, so each split is independent and some repeats will happen by chance; if avoiding repeat pairings actually matters, that's a job for the no-repeat group mixer mentioned below.

It's also worth deciding your range with the activity in mind rather than defaulting to whatever the tool starts with. A brainstorming exercise usually wants smaller groups — two or three — so quieter participants still get airtime. A longer workshop discussion can support four to six without losing anyone in the shuffle. The range is the one lever you're pulling, so it's worth a moment of thought before you hit split.

When you actually want equal groups instead

If what you actually need is a fixed number of same-size teams — say, exactly four teams of five for a bracket — a variable range is the wrong tool for that job; a straight divider that takes a team count and balances sizes evenly will get you there more directly. Variable sizing earns its keep specifically when the activity itself tolerates a range and benefits from one: breakout discussions, mentoring pods, review circles, or any grouping where "somewhere between three and five" is the actual requirement, not a compromise.

When a neighbor tool fits better

If you want a fixed number of evenly-sized teams instead of a size range, the random team / group divider is the more direct fit. If you're re-running groups across multiple sessions and want to avoid the same two people getting paired again, the no-repeat group mixer tracks past pairings and steers around them. And for splitting a video call roster specifically — with copy-friendly output for each room — the breakout room assigner is built for that exact workflow.

Let the range do the work

Not every group needs to match. The variable-size group splitter lets you set the boundaries that actually matter — a minimum and a maximum — and handles the shuffling, the chunking, and the leftover-person problem automatically. Paste your list, set your range, and split.