Drag the handle round to open the wedge, or use the arrow keys.
What this measures
Most tests of this kind show you a chart and ask for a number. This one runs the other way: it gives you the number and asks for the picture. That is the harder direction and the more useful one here, because it is what a wheel does — a wheel is a proportion someone has already drawn, and its honesty is the fact that the wedge and the odds are the same thing.
The edge you measure from is put somewhere different every round, and that is the part doing the work. Fixed at twelve o'clock the task stops being estimation almost immediately: you find the quarter and the half by the clock face and read off the rest. Put the edge at four minutes past eight and there is nothing to read off. What is left is whatever sense of proportion you actually have.
A minute’s exercise, not a test
Nobody is grading this and there is no score to beat. Judging a proportion by eye is a small, specific skill, and like most small specific skills it sharpens with a bit of practice and dulls without any — a few rounds now and then is genuinely a decent workout for the part of your eye that reads charts, maps and the size of one thing against another.
It also costs nothing, needs no account, and works on a phone or a tablet as well as it does on a laptop. If you come back another day the page will still remember how your last few tries went.
The line about which way you lean
After a few rounds the page starts keeping a note of the average direction of your misses — whether the wedges you draw tend to come out bigger or smaller than the number asked for. It is stored in your own browser, it is never sent anywhere, and it is cleared by the reset beside it.
It is a remark, not a result. A handful of tries is a handful of tries, and the honest thing to say about a small number of them is that they lean one way so far. If you want to see what it takes before a run of anything means something, the fairness test is the same argument made with spins.
Why we do not tell you how you compare
There is no score against other people here, and there will not be one. TrueWedge is a static site with no server and no database, so there is nowhere for anyone's results to be collected even if we wanted them — which means any ranking on this page would have to be invented. A made-up distribution is exactly the kind of dishonesty this project exists to avoid, and it would be undetectable by the person it flattered.
For the same reason the page never tells you which way people in general lean. The research on how badly we read proportions is real and worth reading — it is collected in how we misjudge proportions — but almost all of it asks people to judge a shape and name a number, which is the opposite task to this one. The two are not each other's mirror, so borrowing a direction from one and claiming it for the other would be arithmetic we cannot stand behind.