How the Random Pair Matcher Works
Paste a list of names and the matcher shuffles everyone and groups them two at a time. It fits paired discussions, two-person lab work, study partners, and secret gift exchanges, where the pairing has to look fair to everybody. Line breaks and commas both work as separators and empty lines are dropped, so a roster copied from a spreadsheet or a chat thread works as is.
Identical names are treated as different people rather than merged, so a class with two people called Alex still shows the right headcount. Up to 200 names are used and the summary says so when the list was trimmed.
When the headcount is odd
Nobody is quietly dropped. By default the last group becomes a trio, and you can switch to leaving one person waiting instead. Either way the summary names the group that became a trio or the person who is waiting, so it is visible on screen who was affected. With an even headcount that line does not appear at all.
How the shuffle works
The list is shuffled once with the Fisher-Yates algorithm and then read two names at a time. Nobody can land in two groups, and every possible partner is equally likely. Avoiding pairs that came up in an earlier round is not supported, so if the same pair repeats, just run it again.
Frequently Asked Questions
No. By default the last group becomes a trio, and if you choose the waiting option the person left over is named in the summary.
No. They are treated as separate people, so the headcount you pasted is preserved exactly.
Remembering previous rounds is not supported. Every run is an independent shuffle, so run it again if a repeat pair shows up.