Random Name Picker — Spinner Wheel, Team Splitter & Fair Draws
Paste a list of names and spin a colorful wheel to pick a winner — or switch to Pick-N mode to draw several at once, or Teams mode to split everyone into random groups. Winners can be removed and the wheel respun for runner-up draws. Because the draw runs entirely on your device there is no server-side row limit: lists of 10,000+ names work fine. Selection uses a Fisher–Yates shuffle seeded from your device’s cryptographic generator with rejection sampling, so every name has an exactly equal chance — not merely an approximately equal one.
Use Cases
How to pick a random name with the wheel
Type or paste one name per line — students, teammates, raffle entrants. The list saves to this browser automatically.
Click Spin. The wheel accelerates and eases out over 4–6 seconds, landing on a segment chosen by cryptographic randomness.
The name under the pointer pops up in a winner card and is logged to the history list with a timestamp.
Remove the winner and respin for second place; switch to Pick-N to draw several distinct names at once; or switch to Teams to deal every name into 2+ random groups of near-equal size and copy the whole roster out.
Frequently Asked Questions
- Is my name list uploaded anywhere?
- No. The list lives in your browser’s local storage only — nothing is sent to a server, and clearing it removes it completely.
- Is the wheel actually fair?
- Yes. The landing angle comes from crypto.getRandomValues — the same generator used for cryptography — with rejection sampling to avoid modulo bias. Every segment has an identical probability.
- How many names can I add?
- There is no imposed cap. The draw is a single pass over an array in your own browser rather than a server request, so 10,000+ names paste, spin and shuffle without trouble — the practical ceiling is your device’s memory, not a plan limit. Above a few thousand names the wheel segments become too thin to read, so Pick-N or Teams mode is the better view at that size.
- How do I split a list into random teams?
- Switch to Teams mode, set the number of teams, and click Split. One Fisher–Yates shuffle is dealt round-robin across the teams, which guarantees every name lands in exactly one team and that team sizes never differ by more than one. Drawing repeatedly in single-pick mode cannot do this — the same name can come up twice.
- Can the same person win twice?
- Only if you leave them on the wheel. Use "Remove & respin" after each draw to guarantee unique winners round after round.
- How many names can the wheel hold?
- There’s no hard cap. Labels stay readable to around 30–40 names; beyond that, segments get thin — Pick-N mode handles big lists better.
- What does Pick-N mode do differently?
- It skips the wheel and draws N distinct names at once using a cryptographically shuffled copy of your list — instant and guaranteed no duplicates within a draw.
- Will my list still be there tomorrow?
- Yes, on the same device and browser — it persists in local storage across visits. It won’t sync to other devices, by design.
What makes a draw actually fair (and the three ways it usually is not)
Most name pickers feel random without being random. The three failure modes below are all invisible from the outside — the wheel still spins and a name still comes out — which is why a draw used for anything that matters should say which method it uses.
| Method | How it picks | Bias | Verdict |
|---|---|---|---|
| Math.random() % n | Scales a floating-point value, then takes a remainder | Modulo bias — low indexes come up slightly more often, and the generator is not cryptographic | Fine for a coin flip, not for a prize |
| Naive shuffle (swap each item with any other) | Loops the list swapping random pairs | Provably non-uniform: some orderings occur more often than others | Looks fair, is not |
| Fisher–Yates / Durstenfeld shuffle | Walks the list backward, swapping each item with one at or before it | None — every permutation is equally likely | The correct algorithm since 1964 |
| crypto.getRandomValues with rejection sampling | Draws bytes from the OS entropy pool and discards values that would skew the range | None, and unpredictable even knowing prior draws | What this picker uses |
- Rejection sampling is the part most implementations skip: if you map 256 random byte values onto a 10-name list, the first 6 names would win slightly more often. Discarding the out-of-range tail is what removes that.
- Capacity here is bounded by your browser, not a server quota — lists of 10,000+ names paste and draw fine, since the whole draw is a single pass over an array.
- Removing a winner and respinning gives you a genuine runner-up draw. Pick-N mode instead draws several distinct names in one shuffle, which is the right method for splitting a class into teams — repeated single draws can return the same name unless it is removed each time.
Private by design. Names are drawn entirely in your browser and the list is saved only in this browser’s local storage — nothing is sent to a server.
Related tools
More free, in-browser tools from the More Tools set — every one runs locally, with no sign-up and no upload.