Variable-Size Group Splitter

Groups of unpredictable size

Rather than fixing the group size, you give a minimum and a maximum and the tool cuts the shuffled list into groups whose sizes vary within that range. Twenty people with a range of two to four might come out as four, two, three, four, three, four.

That is deliberately different from an even split. A fixed size makes the last group a leftover; variable sizing spreads the irregularity across the whole set so no single group is conspicuously the remainder.

When variable beats even

• Discussion groups, where a group of two and a group of four are both fine and a group of one is not.
• Rotations where identical group sizes would make the pattern obvious.
• Any activity that scales — three people can do it, so can five.
• Not for anything requiring a fixed number of roles; use the team divider for that.

When the arithmetic runs out

A minimum below one is raised to one, and a maximum below the minimum is raised to match it, so the tool corrects an impossible range rather than refusing.

The final group can still fall below your minimum when the arithmetic leaves too few people — check the last group before you use the split. There is no seed, so a split cannot be reproduced.

Why not just split evenly

An even split has one visible flaw: the leftover. Twenty-two people in groups of four is five fours and a two, and the group of two knows it is the remainder. Variable sizing distributes that irregularity so no single group is the obvious afterthought, which changes how the split feels without changing how random it is.

It also breaks the pattern that fixed sizes create over repeated use. A class split into fives every week starts to notice the shape of the groups; a class split into threes-to-fives does not, because there is no shape to notice.