Enter names or options, spin the wheel, and let it pick a fair, random winner. Perfect for classrooms, raffles, giveaways and quick decisions.
🔒 Your list is saved only in this browser — nothing is uploaded.
Call on a random student fairly, or pick teams and turns without any bias.
Draw winners live, and remove each winner so there are no repeats.
Can't decide where to eat or who goes first? Let the wheel settle it.
The spin here is driven by the browser's cryptographic random source, crypto.getRandomValues, rather than Math.random. The distinction matters less for a raffle than for a password, but it is worth knowing why the two differ. Math.random is a fast pseudo-random generator seeded once per page and is not required by the specification to be unpredictable — given enough consecutive outputs, its internal state can be reconstructed and future values predicted. It is fine for animation jitter and wrong for anything anyone might want to game.
There is also a subtle bias that most naive implementations carry. Converting a random 32-bit integer into a range by taking the remainder — value % n — distributes the leftover values unevenly whenever n does not divide 2³² exactly, so some entries become marginally more likely than others. The bias is tiny for small lists but it is real, and it is avoided by rejecting and redrawing the small band of values that would skew the result.
The visual spin is presentation, not mechanism: the winner is selected first and the wheel is animated to land on it. Easing curves, spin duration and the pointer position have no influence on the outcome, which is the honest way to build this — an animation that genuinely stopped by physics would be far easier to bias accidentally.
The choice that actually changes the mathematics is whether a name goes back in after being drawn. Drawing with replacement means every spin is independent and the same person can win twice in a row; over ten spins of a ten-name list, the chance that someone wins twice is close to certain, which surprises people who expect ten spins to produce ten different winners.
Drawing without replacement — removing each winner — guarantees no repeats and is what you want for assigning tasks, picking a running order or dealing out prizes. It also changes the odds as you go: with ten names, the first draw gives each a one-in-ten chance, but a name still in the pool at the last draw is certain to be picked.
Neither is more fair; they answer different questions. Announcing which one you are using before spinning avoids the argument afterwards. For splitting a group into balanced sides rather than picking one winner, the team generator handles the constraint properly.
Duplicating a name to double its chance is the simplest form of weighting and works exactly as expected: with twelve total entries, a name appearing twice wins one time in six. It is transparent, which is its main advantage — everyone watching can count the segments.
The perception problem is that people are poor judges of what randomness looks like. A genuinely random sequence produces clusters and streaks, and three wins from the same department in a row feels rigged even when the process is sound. Runs of that kind are expected: in twenty draws from a five-name list, a repeat somewhere is more likely than not.
If a visibly even spread matters more than statistical independence — rotating a chore, assigning a weekly duty — then a shuffled queue is the better structure. Shuffle once, work through the list, reshuffle when it empties. That produces no repeats within a cycle while still being unpredictable, and it is what most people actually mean when they ask for something to be random.
Yes — the winner is chosen with your browser's cryptographic randomness, so every entry has an equal chance.
Yes — enable "Remove winner after spin" to draw multiple winners without repeats.
Yes — your entries are saved in your browser for next time. Nothing is uploaded.
Classrooms, raffles, giveaways, chores, deciding where to eat, or any fair random choice.
Yes, if you spin with replacement — each spin is independent, so repeats are normal and even likely over several spins. Remove winners after each draw if you need every name to come up once.
No. The winner is chosen by the random source before the animation starts, and the wheel is then animated to land on it. Spin duration and easing are presentation only.