Two shapes side by side. One of them is the other rotated — or it is its mirror image, rotated. Decide which, as quickly as you can be sure.
It measures spatial visualisation: turning an object in your head and comparing it with another, without turning your head or the screen.
This is a different ability from the pattern reasoning in our matrix test. People who are strong at one are not reliably strong at the other, which is why they are scored on separate axes here.
The classic result, described in 1971 by Shepard and Metzler, surprised psychologists: response time rises in a straight line with the angle of rotation — as if the mind really does turn the object at a steady speed instead of comparing features.
So besides accuracy we fit that line and report what each 45 degrees costs you in milliseconds. It is the signature of this ability, and no other test on this site measures it.
Every shape is built by a random walk on a grid and then checked in code for one property: that its mirror image is not equal to any of its own rotations. A symmetrical shape would make both answers correct and the question meaningless. Checking this programmatically removes that possibility rather than making it unlikely.
Rotation by zero degrees is also excluded: an unrotated shape can be compared by eye without any rotation at all, so it measures nothing.
Accuracy is the headline number, not speed. If speed led, random fast tapping would score well — and since there are two possible answers, guessing lands near 50 percent, which accuracy exposes immediately. Time is kept as the secondary figure, and incorrect trials are excluded from the timing, because the time of a wrong answer does not describe a rotation that happened.
Grey shapes, two buttons. Nobody is disadvantaged by colour vision deficiency, and nobody is slowed by reading a language that is not their first. The whole test can be taken without reading a single word.
The 1971 study used drawings of three-dimensional cube figures. Our shapes are flat, built from squares on a grid, which is easier and faster on a phone screen. The idea of the task is the same; the difficulty is not. Treat your number as yours, not as something comparable to a published study.
Spatial tasks are among the few where practice effects are reasonably well documented — people do get better at rotation tasks with training. Whether that improvement shows up in reading maps, packing a car or any other real task is a separate question, and one the evidence answers far less clearly.
It is free and runs in your browser with nothing to install and no sign-up. Signing in is optional and only saves your results across devices.