What Makes a Draw Actually Fair
A picker wheel exists to remove human judgement from a choice, so the only property that matters is whether every entry genuinely has the same chance. Two things can undermine that, and neither is visible in the animation.
The first is the randomness source. Many simple tools use Math.random(), which is fast and evenly distributed but not unpredictable, it runs from an algorithm and a seed, so a sufficiently determined observer can reconstruct the sequence. For choosing who makes the tea, this is irrelevant. For a public prize draw, it is the difference between a result you can defend and one you cannot.
The second is modulo bias, which is subtler and much more common. If you take a random number from a large range and reduce it to your smaller range using a remainder operation, the low entries come up slightly more often, because the large range does not divide evenly into the small one. The effect is small with few entries and grows as the list gets longer.
This wheel sidesteps it by drawing a 32-bit value and dividing by 232 to get a fraction, then locating that fraction along the weighted segments, no remainder operation is involved, so no entry is favoured by the arithmetic.
Duplicates, Weighting and Order
The wheel takes explicit weights using a name:weight syntax, so writing Alice:2 gives Alice twice the slice of anyone entered plainly. That is the deliberate way to bias a draw, and it stays visible in the list rather than hidden.
Repeating a name achieves the same thing by accident, which is the problem: a name pasted twice silently doubles that entry’s chance and nobody notices. Use the weight syntax when you mean it, and check for accidental repeats when you do not.
Removing a winner before the next spin is the difference between drawing prizes and drawing with replacement. For "first, second and third place" you must remove each winner, otherwise the same name can take all three. For "pick a random task, repeatedly", leaving entries in is correct.
Where a Wheel Is the Wrong Tool
Randomness is the right answer when all options are genuinely equivalent and the goal is to avoid bias or deadlock, assigning chores, choosing who presents first, breaking a tie between two acceptable restaurants.
It is the wrong answer when the options are not equivalent and the wheel is being used to avoid a decision you should make. Spinning to choose between two job offers does not produce a good outcome; it produces a random one, and the useful information is usually your reaction to the result rather than the result itself. That reaction is worth noticing, if the wheel lands on one and you feel relief or disappointment, you had a preference and now you know it.
Practical Notes for Group Use
- Show the list before spinning. A draw is only credible to a group if everyone can see the entries were what they expected.
- Decide the removal rule out loud first. Arguments come from ambiguity about whether winners stay in, not from the result.
- Check for near-duplicates. Two spellings of the same name are two entries and double that person's chance without anyone intending it.
- For anything with legal or monetary consequence, the rules of prize draws and lotteries vary by jurisdiction, and a browser tool is not a compliance record.
Nothing you enter leaves your machine. The names you type are not transmitted or stored anywhere, which matters when the list is a class roster, a set of client names, or anything else you would not paste into an unknown service.