Pair Generator

The Pair Generator draws students two at a time so partner work starts without a negotiation. The wheel begins with twelve numbered student slots; replace them with your roster and each spin produces the next name in a random order. Read the names off in twos and the pairs form themselves.

Pairs behave differently from larger groups, which is why this tool is separate from the student group maker. In a pair nobody can hide. Both students talk, both students write, and a partner who does not contribute is obvious within a minute rather than after a week.

How to Pair Students Randomly

1. Load the roster. Open Edit the list and paste one student per line.
2. Turn on no-repeat. Remove entry after it is picked takes each name off the wheel as it is drawn, which is what stops a student being paired twice.
3. Spin twice per pair. The first two names are partners, the next two are the second pair, and so on down the class.
4. Read the pairs off the Results list. Draws appear in order and numbered, so pairing them off is a matter of reading down the list in twos.

No-repeat is the setting that makes this tool work. Without it the wheel can hand you the same student twice in one round, which is fine for a cold-call picker and useless for pairing.

What to Do With an Odd Number of Students

A class of 25 gives twelve pairs and one student left over. Every teacher meets this every week, and the three usual answers are not equally good.

The group of three is usually the right default, provided the third student has a genuine job. In peer review that is a second reader; in a lab it is the person recording measurements. A trio without a third role becomes a pair with an audience.

Stopping the Same Partners Repeating

Random pairing across a semester will repeat some pairs, because randomness does that. In a class of 24 the same two students landing together twice in a term is unremarkable, and students still notice.

Two approaches keep it fresh. Export each draw with Export CSV and glance at the previous file before the new draw, removing one name if a repeat would be the third in a row. Or split the class in half yourself, put one half on the wheel, and pair each drawn name with a student from the other half — which guarantees cross-group pairs without doing any bookkeeping.

Where Random Pairs Work Best

• Think-pair-share. The pairing step is short, so the time saved by not negotiating partners is a real share of the activity.
• Peer review. Random partners give more honest feedback than chosen ones, because friends soften their comments.
• Lab partners. Rotating pairs spreads equipment skills across the class instead of concentrating them in whoever always handles the apparatus.
• Language practice. Speaking with a wide range of partners builds more flexibility than rehearsing with the same one.
• Math problem pairs. Two students working one problem out loud surfaces the reasoning that silent individual work hides.

Pairing Around Accommodations and Conflicts

Place the students who need a specific partner first, then tap both names off the wheel before you spin. That covers a documented accommodation, a language pairing you have set up deliberately, and two students you are keeping apart this week.

The class sees a random draw either way, because the placements happen before the wheel starts turning. Handling it this way avoids the alternative, which is spinning and then visibly overruling the result.

Pairing Larger Classes

The Pair Generator accepts up to 100 entries, so a lecture section pairs the same way a homeroom does. The practical limit is not the wheel but the reading: spinning fifty times in front of a class takes longer than the activity deserves.

For a large section, use Instant pick instead of the full animation. Each press returns a name immediately, and a class of 50 pairs off in well under a minute. Save the animated spin for the smaller draws where the reveal is worth the time.

Fairness, Storage and Sharing

Each draw comes from the browser’s cryptographic random number generator with biased values discarded, so every student on the wheel holds the same chance of being next. The Pair Generator selects the result before the animation begins.

Your roster and sound setting live in the browser on that device and nowhere else. No student name reaches a server, and the tool needs no account. Copy share link packs the roster into a URL, which is the tidy way to keep one bookmark per section. The classroom tool hub holds the other seven wheels.

Rotating Partners Across a Semester

A class that changes partners weekly builds a very different social map from one that does not. In a section of 24 with a fresh draw each week, most students will have worked directly with the majority of the class by the end of a semester, which is difficult to arrange deliberately and free here.

Weekly is the rhythm that suits most courses. Daily is too fast for anything longer than a warm-up, because a pair spends the first few minutes working out how to work together. Monthly is slow enough that pairs settle into a habit, which is fine for a lab bench and limiting for discussion work.

Structuring the Pair Activity Itself

A random pair improves the composition and does nothing for the structure. Pairs produce more when the task tells each student what to do rather than leaving them to divide it.

• Give both students a turn on the clock. Two minutes each, timed, beats “discuss this” — which reliably becomes one student talking for four minutes.
• Require one written output with two names on it. A shared artifact makes a passenger visible immediately.
• Name the two jobs. Explainer and questioner, writer and checker, reader and summarizer. The pairing matters less than the fact that both jobs exist.
• Close with a share. Knowing that either partner might be asked to report keeps both engaged through the pair phase.

Pairs Versus Groups of Three

Default to pairs when the goal is practice and everyone needs reps. Move to threes when the task is open enough that a third perspective changes the answer, and give the third student a defined role when you do.

Managing the Noise

Twelve pairs talking at once is louder than four groups of three, because more people are speaking simultaneously. Pair work fails in rooms where the volume climbs until nobody can hear their partner.

Set the volume expectation before the pairs form, not after. A named level works better than a request — “partner voices, which means the person next to you hears you and I do not” gives students a target they can check themselves against. Pair the instruction with a visible timer, because pair volume rises fastest when students do not know how long they have.

Pairing for Peer Review Specifically

Peer review is the activity where random pairing changes the output most. Students reviewing a friend’s work write encouragement; students reviewing a near-stranger’s work write comments. The difference shows up in the drafts.

Two structures make random peer review work. Give reviewers a short checklist rather than an open invitation to comment, because “any feedback?” produces “looks good” from almost everyone. And separate the reading from the writing: two minutes reading in silence, then two minutes writing, then the conversation. Pairs that talk first tend to agree before either has thought.

Rotating Lab Partners Safely

Science labs are the one place where a purely random pairing needs a check before it runs. Equipment training, safety certification and specific accommodations all attach to individual students, and a draw that ignores them creates a real problem rather than a social one.

Handle it the same way as any other placement: assign the pairs that must be assigned, tap those names off the wheel, and randomize the remainder. Over a semester the rotation still spreads equipment skills across the class, which is the reason to rotate lab partners at all.

Pairing Across Two Groups

Some pairings should always cross a line: a native speaker with a language learner, a student who has mastered a skill with one still practicing it, a returning student with someone who was present. A single wheel cannot enforce that, but two lists can.

1. Split the class yourself. Put one group on the wheel and keep the other on paper.
2. Turn on no-repeat. Each drawn name leaves the wheel.
3. Pair each draw with the next name on your paper list. The draw stays random inside the group, and every pair crosses the line you set.
4. Swap which group goes on the wheel next time. That keeps the randomness from sitting on the same half of the class every week.

This takes about the same time as a straight draw and gives you the structure a single wheel cannot. It is the pattern most teachers end up using once the novelty of pure random pairing wears off.

Frequently Asked Questions