About Wheel Cue

What this is

Wheel Cue is a free online wheel spinner and random picker. You add choices, spin, and get one result. There is no account, no sign-up and no limit on how many times you can spin.

There are a lot of tools that do this. The reason this one exists is the part most of them skip: explaining how the result was chosen.

Every spin here shows its working. Open How this result was picked in any result window and you get the raw random number that was drawn, the threshold used to remove bias, how many draws were discarded, and the arithmetic that turned that number into a slice. The odds are on screen before you spin, and the exact probability of the winner is on screen after.

None of that matters for picking a pizza topping. It matters a lot for a giveaway, a classroom, a fundraiser or anything where somebody might reasonably ask how the result was reached.

The site's research runs simulated draws through the same random.js that picks every winner here, to show how random picks behave in classrooms, giveaways and games, with the numbers published for anyone to check.

How it is built

Some deliberate choices, because they affect how the site behaves:

Everything runs in your browser. Names, entries, saved wheels and results stay on your device, unless you send a wheel as a link or put it on a website. There is no server-side storage of wheels, because there is no application server and no database: the pages are static files, delivered and protected by Cloudflare, which also counts visits without cookies. The privacy policy sets out exactly what that involves.

The randomness is cryptographic. crypto.getRandomValues(), not Math.random(), with rejection sampling to remove modulo bias. The full explanation has the code.

The winner is chosen before the animation. The spin is aimed at a result that already exists, so no rendering quirk can affect it.

The pages are static HTML. Every tool is a real page at a real URL, readable without JavaScript running, which is why they load quickly on a school laptop or a phone on a bad connection.

The client-side code is unminified. /assets/js/random.js is about a hundred commented lines and contains every random draw the site makes. You are meant to be able to read it.

How the research is made

The research section answers practical questions about randomness with numbers rather than opinions. Every study is produced the same way.

It runs the real code. A study is a script that calls random.js, the same file that picks every winner on this site, and repeats the draw as many times as the question needs. No other random number generator stands in for it, and no figure is estimated by hand.

It checks itself against the math. Wherever an exact formula exists, the study works it out and reports it next to the simulated result. The two have to agree. If they do not, something is wrong with the script or the formula, and the study does not go out until the disagreement is explained.

It says what its numbers are. A study can quote a very large number of spins. Those are simulated draws, labelled as simulations on the page, and never presented as people using Wheel Cue.

It can be re-run. Each study publishes the script that produced it and the raw results it wrote. Run it again and the last decimal places will shift, because the draws are random, but the conclusions should hold. The raw results files are free to reuse under the CC BY 4.0 license, with credit to Wheel Cue; the license covers those data files, not the article text.

It is written with help, and checked by a person. Drafts are prepared with the help of AI writing tools. Every figure comes from the published script, not from the AI, and each article is checked, edited and approved by the person who runs Wheel Cue before it is scheduled for publication. That person is responsible for what it says.

It is corrected in the open. If a number looks wrong, let us know. Fixes are made on the page and dated in the changelog.

It is not for sale. There are no sponsored studies and no paid placements. Advertising, when the site carries it, is labelled as advertising, and it never decides what is studied or what a study says.

The fairness report applies the same approach to the code itself, testing millions of draws from random.js against what a fair process should produce.

What we do not do

A short list, because what a tool refuses to do says as much as what it does:

• No invented statistics. Wheel Cue is new, and it does not publish visitor or spin counts it cannot back up. The research quotes large numbers of draws, but those are simulated spins: always labelled as simulations and published with the script that produced them, never offered as usage figures.
• No fake reviews or testimonials. There is no aggregate rating in our structured data for the same reason.
• No "provably fair" badge. It would not be true. Everything runs on your device, and no client-side tool can prove anything to a third party. What we can and cannot claim sets out the limits.
• No account required. Nothing is gated behind a sign-up.
• No student data collection. Rosters typed into the classroom tools stay in the browser. We do not upload or store them.
• No ads where students are. When the site starts carrying advertising, there will be none on the classroom pages, on wheels opened from a shared link or in embedded wheels.

Who builds it

Wheel Cue is a small independent project, not a company. It is built and maintained by one person.

That is worth stating plainly rather than writing "our team" and implying otherwise. It means the site is straightforward, it means bugs get fixed when they are reported, and it means there is no roadmap driven by anything except what people ask for.

The site carries its own name rather than a personal brand, and that is deliberate: nothing here asks to be believed on the strength of who wrote it. The person responsible for it is named in the privacy policy and the terms of use. The code that picks every winner is public and commented, every study publishes its script and its data, and every claim is written so that you can check it yourself.

The changelog lists what has shipped. If something is broken, missing or worded badly, get in touch — that is the main way this gets better.