How to make random pairs from a list of names
Pairing is simple until the list is odd, the activity has several rounds, or two people should not be matched. Make those decisions explicitly before generating the first roster.

Guide visual
Pair rotations
Partner work with fewer repeated pairs.
Choose an odd-count policy
Create a partner rotation
Clean the list
Remove duplicates and confirm attendance. A duplicate name can create an impossible-looking self-match or a missing participant.
Choose pairs or an explicit group count
For an odd list, choose whether the remainder becomes a triad or a bye. The preloaded example uses eight groups so seventeen people produce seven pairs and one triad.
Set every round together
For repeated peer work, set the full number of rounds and enable repeat minimization before generating.
Add only necessary exclusions
Use an apart rule for a real conflict, assessment-integrity issue, or safeguarding need—not because a previous conversation was merely awkward.
Review partner history
Check that each person receives the intended number of assignments and that the triad or bye burden rotates.
Replan after absences
One absence changes an odd list into an even one. Regenerate the remaining rounds rather than carrying forward an unnecessary triad.
Read pair history without losing the odd-person policy
- The session rosters under Your groups show whether each round actually followed the chosen pair, triad, or bye structure.
- The Pair Meeting Histogram counts pair contacts, including the three pair contacts created by a triad. Its repeated total therefore answers how many distinct pairs met again—not who repeatedly received the triad.
- The Pair Meeting Matrix identifies repeated partners and the exact sessions they shared.
- Triad and bye burden must be read from the session rosters because it is a group-format question, not a pair-count metric.
- A required apart pair has no shared session in the matrix and “No violations” in its card in a valid result. Any other result is a solver defect.
Adjust and review the pair rotation
These destinations open the current scenario rather than importing this guide example.
Worked example: seventeen people, three rounds
The preloaded example uses eight groups: seven pairs and one triad in each of three rounds. Everyone participates, and the review should verify that triad membership and repeated partners are distributed rather than concentrated.
Setup facts
- 17 participants
- 7 pairs and 1 triad per round
- 3 rounds
- two prohibited pairings
Review the result
- Confirm every participant appears exactly once per round.
- Track who enters the triad and avoid giving the same people that format repeatedly.
- Check both direct pair repeats and repeated contacts inside the triad.
Try this setup in GroupMixer
This tool is preloaded with the example from this guide. You can edit the participants, constraints, sessions, and balance settings before generating groups.