Random name picker
A random wheel that picks a student, draws the running order of an oral exam, assigns one topic per candidate, or builds balanced random groups.
What are you drawing?
0 of 0 still on the wheel
Paste your list to start a draw.
No draw yet.
| Participant | Draws |
|---|
| State |
|---|
Names stay in your browser: nothing is sent to a server, nothing appears in the page address, and the list is only kept in this computer's local storage.
For an oral or a written exam, Evalmee handles invitations, running order and drawn topics. Run an exam with Evalmee
How to pick a random name with the wheel
- 01
Paste your list
One name per line is enough. Paste from a spreadsheet and the text area becomes a grid: columns take their names from the first line when it looks like a header, and you choose which one appears on the wheel.
- 02
Choose what you are drawing
A name, a full running order, one topic per candidate, or groups. The button changes with the mode, and the wheel belongs to the first one only: the other three produce a list or a table.
- 03
Spin
The wheel turns for three seconds and stops on a name, shown large. The drawn name leaves the wheel for the next spin; "Put back" returns it in one click, and "instant draw" removes the animation once it has done its job.
Draw an oral exam running order
An oral exam is decided partly by the running order: going first means speaking without having heard anyone, going last means waiting three hours. Drawing that order in front of the group, rather than following the register, takes the question off the table — and it is exactly the kind of decision an exam board may be asked to justify.
"Running order" mode shuffles the whole list at once and numbers it. "Print" then produces a one-page sheet: the paper's name, the date and the names in order, in two columns above twenty-four candidates. That is the sheet you pin to the door.
You do not have to draw the whole order at once. In "one name" mode each name leaves the wheel as it comes out, and the history rebuilds the running order as candidates go through. Both routes produce the same list; one gives it in advance, the other builds it.
Assign one topic per candidate
Drawing a topic from an envelope is the oldest form of randomisationRandomisationRandom drawing of questions from a bank and of the order of the options, so that two neighbouring candidates do not see the same paper. It is the cheapest defence against copying between candidates; it assumes questions of comparable difficulty if it is to stay fair. in an oral exam: nobody speaks on the same text as the person before them. The online version does the same thing with two lists — candidates on one side, topics on the other — and assigns one topic per person, without replacement.
With fewer topics than candidates, the tool says so and serves the ones it can: useful information in a corridor, not an error that stops the board. With more, the leftovers simply stay unused, and the next draw puts every topic back in play.
The candidate-to-topic table can be copied, downloaded as CSV and printed. On paper it doubles as a record: who drew what, on what date. A question bankQuestion bankA reservoir of indexed, reusable questions, classified by topic, level or competency. It allows an exam to be assembled by random draw, papers to vary from one session to the next, and the statistical history that makes a question reliable to accumulate, item by item. plays the same role for a written paper, except that it draws per script rather than per candidate.
Build random groups
Two ways to ask: by number of groups ("eight groups"), or by size ("groups of four"). Either way the sizes are spread as evenly as possible — the remainder goes one per group, never piled onto the last. Thirty-seven students in eight groups give five, five, five, five, five, four, four and four.
Asking for groups *of* four out of thirty-seven rounds to the nearest number of groups rather than up: nine groups of four or five, not ten with a single person in the last. That is the difference between a split you can use and one you have to redo by hand.
Balancing on a column answers the other need: if your grid holds a Group, Programme or Level column, each team gets a proportional share of every value. Six A and six B in three groups give two A and two B everywhere; the draw stays random, under the constraint you set.
| Students | Request | Groups | Sizes |
|---|---|---|---|
| 37 | 8 groups | 8 | 5 · 5 · 5 · 5 · 5 · 4 · 4 · 4 |
| 37 | Groups of 4 | 9 | 5 · 4 · 4 · 4 · 4 · 4 · 4 · 4 · 4 |
| 30 | 6 groups | 6 | 5 · 5 · 5 · 5 · 5 · 5 |
| 24 | Groups of 5 | 5 | 5 · 5 · 5 · 5 · 4 |
| 12 | 3 groups, balanced on Group | 3 | 4 · 4 · 4, two A and two B each |
With or without replacement: which is fairer?
With replacement, every draw is independent: the same person can come out twice in a row, and across thirty-seven students the chance that somebody is never picked in thirty-seven draws is around 36%. That is mathematically impeccable and pedagogically awkward — what the class remembers is that three people spoke and the rest said nothing.
Without replacement, the drawn name leaves the wheel: everyone goes once, only the order is random, and the list runs out in exactly as many draws as it holds names. It is this tool's default, and it is what a running order expects. In between, the "draw the already-drawn less often" weighting keeps everyone in play while favouring the quiet ones: each weight is 1 ÷ (1 + times drawn).
| Setting | What happens | When to use it |
|---|---|---|
| Without replacement | The drawn name leaves the wheel until "Put back" | Running order, oral questioning, handing out topics |
| With replacement | Every name stays in play on every draw | Picking a one-off volunteer, drawing a prize |
| Weighted | Weight 1 ÷ (1 + draws already taken) | Getting a whole class to speak across a term |
How the draw works
Each draw takes a uniform random integer between zero and the number of names still in play, produced by the browser's cryptographic generator rather than a plain pseudo-random sequence. The wheel then turns to the name that has already been drawn: the animation illustrates the result, it does not decide it — an instant draw and a three-second spin follow exactly the same law.
P(a given name) = 1 ÷ names remaining · weighted: weight = 1 ÷ (1 + draws already taken)Frequently asked questions
Paste the list into the text area, one name per line, then press "Pick a name". The wheel turns for three seconds and stops on a name, shown very large. No sign-up is asked for and nothing is sent to a server: the draw happens in your browser, offline included once the page has loaded.
Yes. The draw uses the browser's cryptographic random generator, and the integer it returns is corrected for the bias a plain modulo introduces. The wheel's rotation is computed afterwards so that it stops on the name already drawn, so the length of the animation — or its absence — changes nothing.
That is the default: the name that comes out leaves the wheel and cannot be drawn again. A list of thirty-seven names is therefore exhausted in exactly thirty-seven draws. "Put back the last one" undoes the previous draw, "Put everyone back" restarts the whole list, and a ↺ in the grid returns one specific person.
Choose "Groups" mode, give a number of groups or a size, and draw. Sizes are spread as evenly as possible: thirty-seven students in eight groups give five groups of five and three of four. If your list has a Programme or Level column, balancing spreads its values proportionally across the groups.
No. Names never leave your browser: they appear neither in the page address nor in analytics properties, and no request carries them. The list is kept in this computer's local storage so you do not have to paste it again, and clearing that storage deletes it for good.
Yes: the "Full screen" button opens a display that fills the screen, wheel and name included, readable from ten metres. It works on iPhone and iPad too, where browser full screen otherwise applies only to video. Escape, the cross or the back gesture leave it.
Draw the order in "Running order" mode, then press "Print". The sheet holds the document only: the paper's name, the date, the head count and the numbered names — in two columns above twenty-four candidates, so that it fits on a single A4 page. The same sheet works as a door notice and an attendance list.
What if the oral organised itself?
Evalmee invites candidates, manages sessions and groups, hands out topics and collects the marks — with no spreadsheet in the middle.