The fastest fair way to split a class into random groups is to shuffle the full list of names and deal them out one by one into the number of groups you need, like dealing cards.
The fastest fair way to split a class into random groups is to shuffle the full list of names and deal them out one by one into the number of groups you need, like dealing cards. A random group generator does exactly that in a second, and you can do the same by hand with numbered lolly sticks, a deck of cards or counting off around the room.
Random groups work well because pupils see the process as fair. Nobody gets picked last, friendship cliques break up, and students meet classmates they'd never normally talk to. Randomness alone won't balance ability or behaviour, though, so this guide also covers simple ways to add a little structure.
Below you'll find seven tried-and-tested methods, a chart for dividing classes of 20 to 32, tips for leftovers and absences, and the pitfalls to avoid. To make groups right now, paste your class list into the free random group generator.
In this guide:
Teachers in the UK, the US and everywhere in between use random grouping for a simple reason: it removes the politics. When the process is visibly fair, pupils accept the result and get on with the task.
Research summaries on group formation, such as the University of Bath's guide to putting students into groups, note that random allocation is quick and seen as fair, while teacher-selected groups give more control over balance. Many teachers mix both approaches across a term.
An online generator is the quickest option for most lessons, especially if you project it on the board.
Projecting the shuffle helps pupils trust the result. They can see you didn't quietly arrange things.
Keep a saved class list for each set you teach so you can generate groups in seconds. Remove absent pupils before you shuffle rather than reshuffling afterwards, which avoids complaints that you "redid it" to change the result.
Sometimes you don't have a screen, or you want something more hands-on. These classic methods all work.
| Method | How it works | Best for |
|---|---|---|
| Counting off | Pupils count 1, 2, 3, 4 around the room; all the 1s form a group, and so on | Quick splits with no prep |
| Playing cards | Hand out cards; group by number for fours, or by suit for four groups | Groups of four or four large teams |
| Lolly sticks | Write each name on a stick, shuffle in a pot, and draw them out into groups | Primary classrooms |
| Coloured cards or sweets | Pupils draw a colour from a bag and find matching colours | Younger children, active starts |
| Puzzle pieces | Cut pictures into pieces; pupils find the others who complete their picture | Ice-breakers and review games |
| Matching pairs | Hand out question and answer cards; pupils find their match | Pairs with a learning twist |
Counting off has one weakness: pupils quickly learn to move seats to land with friends. Change the starting point and the direction each time, or hand out numbers yourself as they walk in.
For a one-off pick, such as who answers next or who presents first, a random name picker adds a bit of suspense without making anyone feel targeted.
Class sizes rarely divide neatly. This chart shows a balanced split for common class sizes, keeping every group within one person of the target size.
| Class size | Pairs | Groups of about 3 | Groups of about 4 | Groups of about 5 |
|---|---|---|---|---|
| 20 | 10 pairs | 4 of 3, 2 of 4 | 5 of 4 | 4 of 5 |
| 24 | 12 pairs | 8 of 3 | 6 of 4 | 4 of 5, 1 of 4 |
| 25 | 11 pairs, 1 trio | 7 of 3, 1 of 4 | 5 of 4, 1 of 5 | 5 of 5 |
| 27 | 12 pairs, 1 trio | 9 of 3 | 6 of 4, 1 of 3 | 3 of 5, 3 of 4 |
| 28 | 14 pairs | 8 of 3, 1 of 4 | 7 of 4 | 4 of 5, 2 of 4 |
| 30 | 15 pairs | 10 of 3 | 6 of 4, 2 of 3 | 6 of 5 |
| 32 | 16 pairs | 8 of 3, 2 of 4 | 8 of 4 | 4 of 5, 3 of 4 |
The quick rule: divide the class by your target size, round to the nearest whole number of groups, then spread the leftover pupils one per group. Avoid groups of one or two when you're aiming for fours, because they rarely have enough voices for a good discussion.
"Random" only feels fair if every pupil has an equal chance of landing in any group. Some homemade methods quietly break that rule.
Good digital tools shuffle the whole list using a proper algorithm. The standard one is the Fisher–Yates shuffle, which gives every possible ordering an equal chance. Once the list is shuffled, dealing names round-robin into groups keeps sizes balanced.
By hand, shake the pot properly or shuffle a deck at least seven times. For most classroom purposes, that's plenty.
Pure randomness can occasionally land all your strongest readers in one group, or put two pupils together who really shouldn't sit side by side. A little structure fixes that while keeping the process transparent.
Keep adjustments small and consistent. If you overrule the shuffle every lesson, pupils will stop trusting it. A swap or two is usually enough.
The right group size depends on what you want pupils to do. Getting it right matters as much as how you form the groups.
| Activity | Best size | Why |
|---|---|---|
| Think-pair-share | 2 | Everyone must talk, nobody can hide |
| Peer marking | 2 or 3 | Focused feedback on each piece of work |
| Discussion questions | 3 or 4 | Several viewpoints without crowding |
| Science practicals | 3 or 4 | Enough hands for equipment and recording |
| Research projects | 4 | Clear roles and shared workload |
| Debates and team games | 5 to 8 | Bigger teams create energy and competition |
Pairs suit quick checks and shy pupils, because there's no audience to perform for. Fours offer a sweet spot for most collaborative work: two pairs can split a task and then compare. Larger teams work best when the activity itself keeps everyone busy, such as a relay quiz or a debate with assigned speaking turns.
Tell pupils the size and the reason before you shuffle. "Fours today, because each of you will research one part" makes the grouping feel purposeful rather than arbitrary.
Real classrooms are messy. These quick fixes keep you moving:
For weekly activities, reshuffle often. Fresh groups every lesson or every week keep the social mixing going and stop any one group's dynamics from settling into a rut.
The same techniques work anywhere you need fair teams quickly:
For quick decisions such as who goes first, or a number between 1 and 30, a random number picker does the job and everyone can see the result.
Whatever the setting, the principle stays the same. Make the process visible, keep the rules simple, and people will accept the outcome even when they don't get the team they wanted.
A few habits undermine random grouping. Watch out for these:
That last point matters in schools. Data protection rules such as the UK GDPR mean you should think twice before pasting full names into random websites. Tools that work entirely in the browser, without saving anything, are the safer choice. First names or initials also work perfectly well.
Shuffle the full list of names, then deal them one at a time into each group in turn. A random group generator does this instantly and visibly.
Make eight groups: six groups of four and two groups of three. Alternatively, make seven groups with five fours and two fives.
Count off around the room, hand out playing cards, or draw named lolly sticks from a pot and deal them into groups.
Groups of three to four work well for most tasks. They're small enough for everyone to speak and big enough to share ideas.
Both have a place. Random groups feel fair and mix pupils socially, while teacher-chosen groups give more control over ability and behaviour.
Change them every lesson or every week for short activities, and keep them fixed for the length of a longer project.
Use tools that run in the browser and don't store data. First names or initials are enough, and they reduce data protection risk.
Last checked: 18 October 2026. Rates, rules and app settings can change, so confirm anything important with the official source linked above.
To average times in minutes and seconds, convert every time to seconds, add them up, divide by how many times you have, then convert the answer back to minutes and seconds.
If you work irregular hours or only part of the year in the UK, you build up paid holiday at 12.07% of the hours you work in each pay period. Your employer can either pay you for that holiday when you take it, or add 12.07% to every payslip as "rolled-up" holiday pay.
Stacked discounts multiply rather than add. To find the final price, multiply the original price by (1 − first discount) and then by (1 − second discount). So 20% off followed by an extra 10% off gives 0.8 × 0.9 = 0.72 of the price, which is 28% off, not 30%.