Are online random wheels actually fair?

A wheel can look fair and still be rigged — here's how to tell the difference.

If a spin wheel is going to decide who wins a prize, who goes first, or who gets the last slot, it's reasonable to ask whether the thing is actually fair. The honest answer is: a good one is, and you can check. This guide explains, in plain English, how online wheels pick a winner, what "random" really means on a computer, and the practical signs that separate a trustworthy picker from a dodgy one.

What "fair" actually means

Fairness comes down to two things. First, every eligible entry has the chance it's supposed to have — no more, no less. Second, the result isn't decided in advance or quietly nudged toward a particular answer. A fair wheel doesn't necessarily mean every entry is equally likely; it means the odds are whatever they're declared to be, and nothing is happening behind the scenes to override them.

That distinction matters because "fair" and "equal" aren't the same thing. A draw can be perfectly fair while deliberately favouring some entries, as long as that favouring is out in the open. The unfairness comes from hidden influence, not from uneven odds.

What "random" means on a computer

Computers can't be spontaneous, so they fake randomness with maths. The most common approach is a pseudo-random number generator (PRNG): a clever formula that starts from a number called a "seed" and churns out a stream of values that look random. They pass as random for most purposes, but there's a catch — because it's just a formula, anyone who knows the seed and the method could, in principle, work out exactly what comes next. It only looks unpredictable from the outside.

Browsers also offer a second, stronger option: a cryptographically secure generator, exposed through the Web Crypto API (crypto.getRandomValues). This one is specifically designed to be unpredictable — you can't reverse-engineer the next value from the ones you've already seen. It's the same class of randomness used to protect passwords and secure connections, which is exactly why it's the better choice for draws that people actually care about. SpinKit uses the browser's cryptographic generator for its picks, rather than a basic PRNG, precisely so the outcome can't be predicted or steered.

The "modulo bias" trap

There's one subtle pitfall worth knowing about. To turn a big random number into "pick one of these eight names", code has to squash it down into the right range. The naive way of doing that can very slightly favour some entries over others — a tiny, lopsided wobble that you'd never notice on a single spin but that's real over many. Careful pickers correct for it so every slot stays evenly weighted. SpinKit corrects for this, so the maths underneath doesn't quietly tilt the result.

Weighting is still fair — when it's visible

A wheel where one entry has a bigger slice than the others isn't broken. A bigger slice simply means a higher chance, and that's completely legitimate when the weights are visible and everyone has agreed to them. The classic example is a raffle: someone who buys five tickets should have five times the chance of someone who buys one. Nobody calls that rigged, because the reason for the uneven odds is plain to see.

What turns weighting into cheating is hiding it. If a slice looks the same size as the others but is secretly more likely to win, or if the odds have been tilted without anyone being told, that's the dishonest version. The test is simple: can you see why one entry is more likely than another? If yes, it's fair. If the favouring is invisible, it isn't.

Independent spins vs. removing names

People often suspect foul play when the same name comes up twice in a row. But if each spin is independent — the wheel doesn't remember the last result — then a repeat is just luck, not unfairness. A fair coin can land heads five times running; nothing is wrong with the coin.

Removing a name after it's drawn is a different mode entirely, and it's also fair — it just changes the odds on purpose. Take the winner out and the remaining entries each become more likely, which is exactly what you want for a no-repeats draw where everyone should get a turn before anyone goes twice. Both behaviours are legitimate; what matters is knowing which one you're using and why.

How to sanity-check any picker

You don't need to read the code to judge a wheel. A few visible checks go a long way:

• Can you see the full list? Every entry should be shown before the spin, so nobody is secretly missing or duplicated.
• Does every entry visibly have a slice? If an option is on the list but has no real presence on the wheel, the odds don't match what you're being shown.
• Can you watch the spin? A result that simply appears, with no visible spin, is asking you to take it on trust.
• Are the odds explained? If some entries are weighted, you should be able to see why.

And the red flags are just as telling: a wheel where results always seem to favour the same option, a picker with no way to see the entries at all, or a "spin again" that keeps going until a particular answer — say, a sponsor — happens to win. Any of those means the outcome is being managed rather than drawn.

The bottom line

Online wheels can absolutely be fair, but "looks like a wheel" isn't proof. A trustworthy picker uses genuinely unpredictable randomness, corrects for the small biases that creep into the maths, and shows you everything — the entries, the slices, the spin and the odds — so you can verify the result rather than just believe it. A spin wheel built that way gives you a result the whole group can watch and agree with. If you ever can't see how a winner was chosen, that's your answer. For more on running draws people trust, browse the rest of our guides.