Random Group Generator for Teachers

Paste your class list, pick a group size or count, get balanced random groups. Optional seed for reproducibility. Avoid repeating last week's pairs.

0 names

Saved to this browser. No account, no upload, no signup.

How the generator works

It does a uniform-random Fisher-Yates shuffle of your full class list, then distributes students round-robin into groups. In groups of K mode you choose the size (3, 4, 5…) and the number of groups follows from the class size. In N groups mode you choose how many groups and the size follows. Either way, the algorithm balances sizes — a remainder is spread across earlier groups, not bunched at the end.

Seeding for fairness and reproducibility

A seed locks in the random draw. Same seed + same names + same mode = same groups, every time. Three common uses:

  • Run the same groups across two periods. If your 1st and 3rd period both have the same project on the same day, use seed wed-april-15 in both periods to produce parallel group structures.
  • Share groups with a colleague. Two teachers in adjacent classrooms can produce identical groups by agreeing on a seed and pasting the same combined list.
  • Reproduce groups after a refresh.If you closed the browser and need to recreate this morning's shuffle, type the same seed.

Avoid repeating last week's pairs

With this toggle on, the generator stores the pair-ups from your most recent shuffle in your browser's local storage. The next shuffle runs up to ten attempts and picks the one with the fewest repeat pairs. This isn't a hard constraint — with very small group sizes in a small class, some repeats are unavoidable — but it's usually enough to keep multi-week project groups feeling fresh.

Pedagogical patterns

  • Jigsaw expert groups — Use groups of 4 for the expert phase, then re-shuffle to mixed groups for the teach-back phase. The seed feature lets you control which jigsaw a student lands in if you have specific reasons for matching.
  • Think-pair-share— Use the picker for the "share" part. The pair part can be groups of 3 if you want triads instead of pairs.
  • Random project groups for the term — Run one shuffle at the start of the unit, save the result (copy to a Google Doc), and use those teams across multiple class periods until the project completes.
  • Mixed-ability shuffling— Generic randomisation doesn't account for ability mix. For deliberately balanced mixed-ability groups, sort your list before pasting (highest → lowest performer), then groups-of-4 mode pairs row 1 with rows 8, 15, 22, etc., giving roughly balanced groups by design.

What this tool doesn't do

It doesn't hard-exclude specific pairings (no "Alice and Bob must not be together" constraint), doesn't balance by gender / reading level / IEP, and doesn't persist multi-week history beyond the most recent shuffle. For deliberately balanced groupings on multiple axes, manual sorting before pasting is the intended workflow.

Related Tools

Groups of 3 Generator

Same tool with the group size locked to 3 — useful for jigsaw activities.

Groups of 4 Generator

Same tool with the group size locked to 4 — the most-used team size.

Name Picker

After making groups, randomly call on one student to summarise their group's thinking.

Classroom Timer

Pair with a projected countdown for timed group-work sessions.

Toolkit Hub

All four free classroom tools in one place.

Random Group Generator for Teachers FAQ

How does this group generator decide who's with who?
It shuffles the full class list with a uniform-random algorithm (Fisher-Yates) and distributes students round-robin into groups. With the "Avoid repeating last week's pairs" toggle on, it makes up to ten attempts and keeps the result with the fewest pair repeats from the previous shuffle stored in your browser.
What does the seed do?
A seed locks in randomness — the same seed + same names + same mode always produces the same groups. Useful for: (a) reproducing the same groups across multiple classes on the same day, (b) sharing groups verbally with a colleague ("use seed period-3-fri"), (c) re-running the same shuffle after a refresh. Leave the seed blank for truly random.
Can the generator avoid putting two specific students together?
Not directly — there's no hard exclusion list. The simplest workaround is to remove one of the two names from the input, run the shuffle, then add them manually into the remaining group with the best fit. A built-in "keep separate" constraint is a planned addition; ask if it would block your use.
Does it handle uneven groups?
Yes. In "groups of K" mode with an uneven class (23 students, groups of 4), you'll get five groups of 4 plus one group of 3. In "N groups" mode, sizes are as balanced as possible — e.g., 23 students into 6 groups becomes four 4-person groups and two 3-person groups, distributed round-robin so the smaller groups don't cluster at the end.
Where is the "previous pairs" history stored?
In your browser's local storage on this device. It records the pairings (which students were in a group together) from your last shuffle, then uses that to bias the next shuffle. It doesn't track multiple weeks back. Clear browser storage to reset.
Can I print the groups?
Use your browser's File → Print (or Cmd/Ctrl+P) — the group cards print one-per-page-block legibly. We're tracking demand for a dedicated print-friendly button with name cards / table tents on letter paper.
Does the group generator work with 100+ students?
Yes, the algorithm handles arbitrary list sizes. UI legibility starts to suffer beyond about 80 names on a single screen — for very large pasted lists, the result grid stays readable but the input textarea needs scrolling.
Can two students who were absent last week still be paired together?
The avoid-pairs logic only looks at your last shuffle, not at attendance history. If two students who weren't pair-able last week (one was absent) are pairable this week, the generator may put them together. That's usually what you want — the avoid-pairs feature is a freshness heuristic, not a strict constraint.