Mouse Accuracy Test
A structured click accuracy test based on Fitts's law. Hit 30 targets of varied size and distance to get your accuracy, average time, throughput in bits per second and a scatter of where your clicks really land.
Needs a pointer (mouse, trackpad, or touch); the keyboard is not read. Click the start circle, then each new circle as fast as you can while staying accurate. A click on the board outside the circle counts as a miss, and the circle stays until you hit it.
How to read your mouse accuracy result
Fitts's law in one paragraph
Fitts's law says the time to hit a target grows with how far away it is and shrinks with how big it is. Its difficulty is measured in bits: ID = log2(D / W + 1), where D is the distance and W the target width. A 32 px target 288 px away is log2(10) = 3.3 bits. Dividing the bits by the seconds you took gives throughput, which stays fairly steady for one person and one mouse whether the targets are easy or hard, so it is a fairer score than raw speed or raw accuracy alone.
Reading the throughput
Each of the 30 targets gets its own bits per second, and the headline is the average. Lab studies of mice usually land between 3.5 and 5 bits/s; here the time also includes spotting each new target, so 3 to 5 is typical and above 6 is very quick. A miss leaves the circle in place and does not restart the clock. The extra time lowers that target's throughput, and the miss is included in accuracy. Small targets should show higher bits/s than large ones; if they score far lower, you are slowing right down to land them.
Reading the scatter
The scatter puts every click in units of target radius, so all sizes share one chart. A cloud around the centre is normal hand shake. A cloud that sits off to one side is a constant offset: overshooting in the direction of travel usually means pointer acceleration (Enhance pointer precision in Windows) is fighting your muscle memory, and a steady drift up or sideways can point to the sensor, a worn foot or your grip. The mouse drift test checks the sensor on its own.
Comparing mice, DPI or a mousepad
Does the new mouse really make you more accurate, as the box says? Run the test three times with each setup, in the same mode and at the same time of day, and compare the averages rather than single runs, because run-to-run noise is around 0.3 bits/s. Change one thing at a time. Measure your real DPI first with the mouse DPI test so you are comparing like with like. Browsers only see pointer events, so this measures the whole chain: hand, mouse, settings and display.
Related tools
For open-ended practice use the aim trainer, check your raw speed with the reaction time test, verify your sensor resolution with the mouse DPI test, and rule out a wandering cursor with the mouse drift test.
Frequently asked questions
How is accuracy calculated on this test?
Accuracy is targets hit divided by hits plus misses, shown as a percent. The scored run is 30 circles, and each one counts as a hit when the pointer goes down inside it. Every other primary click on the board is a miss, and that same circle stays until you hit it. The opening click that starts the run is not scored. If a result is built with no finished targets and no misses, accuracy is 0 instead of a divide-by-zero.
How is throughput in bits per second calculated?
Each target gets a difficulty, ID = log2(distance / width + 1). Distance is from the centre of the previous target to the centre of the one you hit, and width is that circle's diameter. Bits per second for the target is ID divided by the seconds since the previous hit, and misses in between stay inside that time. The headline number is the average of those 30 values, not total bits divided by total time. Average time is the average of the same gaps, in milliseconds. A gap is never treated as zero seconds.
What do Standard, Precision and Touch change?
Every mode is 30 targets, 10 of each size, with short, medium and long moves mixed in. Standard uses 16, 32 and 64 px circles and shows a running clock. Precision uses 6, 10 and 14 px circles and hides the clock: the big rank uses accuracy, but each hit is still timed for throughput. Touch uses 36, 52 and 76 px circles. On the smallest circles a faint outer ring only helps you find them; the hit area is the filled circle. Your best bits/s for each mode stays in this browser.
What do the rank names mean?
In Standard and Touch the name comes from throughput: under 2.5 bits/s is Warming up, from 2.5 it is Casual, from 3.5 Solid, from 4.5 Sharp, from 5.5 Excellent and from 6.5 Elite aim. In Precision the name comes from accuracy: under 65% is Needs a steadier grip, from 65% Shaky, from 80% Decent, from 90% Steady hand and from 97% Surgeon. The names describe this run only. They are not a comparison with other people.
When does the clock stop, and what does reset clear?
The on-screen clock advances only during a Standard or Touch run. It stops as soon as you hit the 30th target, choose Stop, change mode, or hide the tab, and the 30th hit writes the score once. Stop, Try again, a mode change and a hidden tab clear the current circle, the miss count and the clock. Hiding the tab abandons the run because a background tab makes the timing wrong. Precision hides the clock for the whole run.
Do I need a mouse, and which clicks count?
The board needs a pointer: mouse, trackpad, pen or touch. Keyboard keys are not read. A hit is a primary click inside the circle, using the board's own coordinates. A primary click anywhere else on the board is a miss. Right-click is ignored. Clicks outside the board are not counted. Targets are drawn only after the board has a real width and height, and each circle is kept inside that area.
What does the click scatter show?
Every scored click is plotted from the centre of the circle you were aiming at, in units of that circle's radius. Dots are hits and X marks are misses. The inner ring is the target edge, the outer ring is twice the radius, and the plus sign is the average click. A tight cloud on that plus sign means little offset. A cloud shifted to one side is a steady bias. A cloud spread around the middle is shake without a bias.