Random grouping that actually works
Random grouping solves a real problem — students otherwise work with the same people all year — but it is not always the right call, and how you run it matters more than the randomness.
6 min read
What random grouping is actually for
Left alone, students form stable groups by friendship, and those groups harden over a year. The cost is not social — it is that students stop encountering people who think differently, and the ones without an obvious group get left to the end.
Randomising removes the social negotiation entirely. Nobody is picked last, because nobody is picked at all.
When not to randomise
Random is not automatically fair. It is fair on friendship and unfair on everything else, and there are cases where that trade is wrong.
- Long projects. Over six weeks, a badly-drawn group has time to fail properly. Randomise short tasks; construct long ones.
- Where support needs matter. A student who needs a particular partner, or needs to be apart from someone, should not be left to chance.
- When one group would end up with no strong reader, or no confident speaker, and the task depends on it.
Draw in front of the class
Groups produced in advance invite the suspicion that they were arranged. Drawing live removes the argument — students watched it happen.
It also stops the negotiation before it starts. A list handed out gets challenged; a draw everyone saw is simply the result.
Give the group a structure, not just a membership
The most common failure is not the composition — it is that four students are told to work together with no idea who does what. One takes over, one disengages, and the task gets done by half the group.
Assign roles as part of the setup: someone records, someone reports back, someone keeps time, someone checks the work against the brief. Rotate them between tasks so nobody is permanently the scribe.
Roles can be drawn the same way as the groups, which takes the sting out of nobody wanting to present.
Running it
The Random Team Generator handles the mechanics: set a team count between two and ten, and each spin sends its winner to the next team in rotation and takes that name off the wheel, so nobody is assigned twice and nobody is missed. "Assign the rest" fills the remaining slots in one action when the theatre has served its purpose.
For roles, a second wheel with the role names does the same job — or chain the two so one spin picks a student and the next picks what they will do.