Weighted Wheel Probability Explained

The basic idea

On a standard wheel, every entry has the same chance. With ten names, each one has a one-in-ten chance, or 10%, of being picked on any spin. A weighted wheel changes this by giving some entries more "shares" than others. An entry with weight 2 has twice the chance of an entry with weight 1. The wheel adds up all the weights and gives each entry its share of the total.

The formula is simple: chance = entry weight ÷ total of all weights. If you have three prizes with weights 5, 3 and 2, the total is 10. The first prize has a 50% chance, the second 30% and the third 20%. On screen, slice sizes match these percentages, so the bigger slices really are more likely to win.

Worked examples

A reward wheel for a class might have "5 minutes extra break" at weight 1, "sticker" at weight 4 and "choose the music" at weight 3. The total is 8, so the break comes up 12.5% of the time, the sticker 50% and the music 37.5%. Over forty spins you'd expect about five breaks, twenty stickers and fifteen music choices. In practice you won't get exactly those numbers, but you'll be close.

A sales promotion might offer 10% off at weight 70, 20% off at weight 25 and a free product at weight 5. These weights already add up to 100, so they read directly as percentages. That's a useful trick: if you choose weights that total 100, everyone can understand the odds without any arithmetic.

A raffle where people bought different numbers of tickets is also a weighted wheel. If Ana bought five tickets, Ben two and Chloe one, give them weights of 5, 2 and 1. It's the same as entering Ana's name five times, but the wheel stays tidy and shows the right percentage for each person.

Weights versus duplicate entries

Typing a name twice and giving it weight 2 have the same effect on the odds. Weights are usually better because the wheel stays readable, the percentages are shown, and you can change them in one place. Duplicates are easy to create by accident when pasting lists, which is why the Dedupe tool exists. If you're using weights, dedupe first, then set the weights, so nobody ends up with more chances than intended.

Why results don't match the percentages exactly

Probability describes what happens over many spins, not over a few. A 20% slice might come up three times in a row or not at all in ten spins. Both are normal. Over ten spins you'd expect two wins, but anywhere between zero and five is quite likely. Over a thousand spins the share will settle very close to 20%.

This is the "law of large numbers". It's also why people sometimes think a fair wheel is broken: we expect randomness to look evenly spread, but real randomness has clusters and gaps. The randomness audit page lets you run thousands of simulated spins with your own weights and compare the results with the expected percentages. It's a good way to show a sceptical class or audience that the wheel behaves as promised.

How the wheel picks a weighted result

Under the hood, the wheel lines all entries up end to end on a number line, each taking up a length equal to its weight. It then picks a random point along the total length using the browser's cryptographic random number generator, and whichever entry covers that point wins. The animation is then set to stop on that slice. Because the result is chosen first and the animation follows, how hard you click or how long you hold the button makes no difference.

Choosing good weights

Start with the experience you want. If a prize should feel rare, give it a small weight, around 5% or less. If it should be common, give it the largest share. Check the percentages before you spin: the entries panel shows each entry's chance, and it's easy to misjudge how much a "small" weight adds up when there are many entries.

Keep the numbers simple. Whole numbers like 1, 2, 3 and 5 are easier for others to understand than 1.37. If the odds matter to participants, publish them. A customer who knows the grand prize has a 2% chance is far less likely to complain than one who discovers it later.

Avoid zero weights unless you mean to exclude an entry. An entry with weight 0 can never win, which may confuse people who see it on the list. If you want to remove someone for a day, hide the entry instead.

Weights in the classroom

Teachers sometimes use weights to rebalance participation, giving a little more chance to students who haven't spoken recently. That's a reasonable use, but Fair Rotation does it automatically. It raises the chance of names that haven't been picked and lowers it for names that just were, then resets as things even out. It saves you adjusting numbers by hand and is easier to explain: "everyone gets a fair share of turns over time".

For reward wheels, weights keep big prizes special without removing them. Students accept that a homework pass is rare if they can see the slice is small. Hiding the odds, on the other hand, makes the wheel feel like a trick.

Weights in giveaways and promotions

Weighted draws are fine in giveaways if the rules say so. "Each extra friend you tag gives you one more entry" is a weighting rule. Apply it consistently and show the final weights before the draw. For prize wheels in shops or at events, be careful: some places have rules on games of chance and require the odds to be displayed. When in doubt, display the odds anyway. It costs nothing and builds trust.

Common mistakes

Forgetting to reset weights after a one-off event, so next week's wheel is still unbalanced. Mixing duplicates and weights, so some entries count twice. Assuming a big slice is "due" after a run of misses (each spin is independent, so the wheel has no memory unless Fair Rotation is on). Giving joke entries large weights and then wondering why the serious options rarely come up. A quick look at the percentage column before spinning catches nearly all of these.

Summary

A weighted wheel gives each entry a share of the total, and its chance is its weight divided by that total. Slice sizes show the odds visually. Results vary over a few spins and settle over many. Use simple numbers, check the percentages, and publish the weights whenever the result matters to other people. Used openly, weights let you design exactly the experience you want while staying fair.

A quick practice exercise

Try this with a group to make the idea stick. Build a wheel with four colours weighted 1, 2, 3 and 4. Before spinning, ask everyone to work out the chances: 10%, 20%, 30% and 40%. Spin twenty times and tally the results on the board. Compare the tally with the predictions, then run the same weights through the randomness audit for ten thousand spins. The small sample will be messy and the large one will match closely. It's a memorable lesson in probability, sampling and why one surprising result proves very little.

Try it now

Put these ideas into practice with the Random Number Generator, Randomness Audit, Random Student Picker, or open the main spin the wheel tool.