Skip to the contentSpinRondo

Probability spinner with sectors sized to their chance

Six outcomes weighted five, four, three, three, two and one — eighteen units in total, so one unit is exactly twenty degrees of rim. Nothing is removed as it is drawn, so an outcome can come up twice and a class can spin two hundred times without touching a setting.

This wheel is already set up for the job. Open it in the editor to change the list, the machine or the odds.

Every spin draws from your browser’s cryptographic random source. How a draw is made sets out the method and its limits.

Open this wheel in the editor

Why is a spinner a better model of chance than a die?

Because here the probability is the picture. A die’s one-in-six is an assertion: you have to be told the cube is fair, and nothing you can see about it settles the claim. A spinner’s one-in-six is a sixth of a disc, and a pupil who can compare two angles checks it from the back of the room.

It is also the only model that takes unequal outcomes without new apparatus. A die cannot be made to give an outcome of 5/18; a spinner is built to. That is why textbooks draw spinners for the chapter where probabilities stop being equal, and it is why this wheel arrives weighted rather than even.

What does the gap between theory and twenty spins actually teach?

That twenty spins will disagree with the fractions, and that the disagreement is not an error. Win a point has a probability of 5/18, so twenty spins should land on it about five and a half times — but anything from three to eight happens roughly seven times in eight. A class that spins twenty times and counts four is looking at an ordinary result, not a broken wheel.

Raise the count and the gap shrinks as a share while it grows in whole spins. At twenty spins the count typically strays about two either way; at two hundred it strays about six — but six out of two hundred is a much smaller share than two out of twenty, and the proportion settles near 27.8% and stays there. That is the law of large numbers said honestly, and it is the part of the topic a worksheet cannot show.

How do you set the sectors so the picture and the fractions agree?

Give every outcome a whole-number weight and add them up. This wheel is five, four, three, three, two and one — eighteen in total — so a weight of one is 1/18 of the disc and, because eighteen divides 360 exactly, precisely twenty degrees of rim. Win a point at five is a hundred degrees, free spin at one is twenty, and every sector can be checked with a protractor.

Choosing a total that divides 360 is the part worth teaching: 8, 9, 10, 12, 18, 20, 24 and 36 all give whole-degree sectors, and a total of seven does not. The tool does the arithmetic either way — a segment is drawn from its weight and the percentage beside it is the same number again — but a lesson goes better when pupils can measure what they calculated.

What else is this wheel good for?

It is the name wheel with weights turned on, so the chance printed beside each entry is the one the draw uses. Where the outcomes are numbers rather than events — a fair die, a bag of counters — the number wheel builds the range for you.

The same wheel with multiplication facts on it is a times tables drill. The same weights with something to win behind them make a prize wheel, where visible odds stop being a lesson and become a promise.

Questions about the probability lesson

Weight divided by the total of the weights, exactly. Here the weights are five, four, three, three, two and one, so the total is eighteen and win a point is 5/18, about 27.8%. Nothing is held as a decimal anywhere in the draw: the weights are summed as whole numbers and one uniform whole number is taken from zero up to eighteen, so there is no rounding for a bias to hide in.

Yes. Every draw is written into the session log beside the wheel with the outcome, the time and an eight-character draw id, and the log exports as a file — which is the tally sheet, already made. It lives on your device and clears when the tab closes, so a two-hundred-spin experiment wants exporting before the lesson ends rather than after it.

Each spin takes a value from the browser’s cryptographic random source and discards the values that would favour the first few segments rather than rounding them in. The animation is computed from the result and not the other way round, so the spin length changes nothing. What that does not make it is audited: this is classroom apparatus, not a certified source of randomness, and the page on how a draw is made says so plainly.

Yes, and that is usually the better half of the lesson. A pupil sets six outcomes and six weights, copies the link, and it opens on somebody else’s screen as exactly the same wheel. Then set the harder task: work out each other’s probabilities from the picture alone, before anybody looks at the percentages. The wheel will not teach the sample space for you — it only makes the answer checkable.