Slightly Better Response Time Test
Most reaction tests time a single click after a light changes. This one also times how fast you correct a movement you're already making, which is what aiming, driving and catching actually ask of you.
Why a click isn't the whole story
One way to sort inputs is by how many continuous dimensions they carry. A button carries none, so call it 0D. A trigger or slider carries one, and a mouse or thumbstick carries two. Nobody seems to have formally coined those labels, but the idea is old: Bill Buxton's 1983 taxonomy sorted continuous input devices by how many dimensions they sense, one to three, and left buttons out entirely.
A classic reaction-time test is 0D: a light changes and you press a button. It times how long you take to notice something and start a movement from rest. In a large, carefully calibrated study that came to about 230 ms, and the researchers estimate that only about 130 ms of it was spent detecting the light (Woods et al., 2015). Online tests often read higher, since the computer and screen can add up to 100 ms of their own.
Most real tasks aren't like that. Aiming, driving and catching are closed loops: you're already moving, you see what your movement did, and you keep correcting it. Corrections to a movement that's already underway come much sooner than a fresh reaction. People sliding a finger across a screen to intercept a target began adjusting 114 ms after it jumped (Brenner, Bom & Smeets, 2026), and the same fast corrections show up with a computer mouse (Brenner & Smeets, 2003). Models of pilots in tracking tasks put the human's delay at 0.1 to 0.2 s (McRuer & Jex, 1967). So a 0D score mostly tells you how fast you start, not how fast you steer.
1D and 2D tests measure the loop itself. In the strafe test you track a target along a line, and it times how long your hand keeps going the old way after the target turns. A 2D test adds direction: the correction could point anywhere in the plane, so you have to work out where as well as when. You might expect that to cost time, but adjustments to direction start about as fast as adjustments to distance (Oostwoud Wijdenes et al., 2013). What tells you the target moved matters more: corrections cued by a change in color or shape start about 50 ms later than ones cued by brightness or size (Veerman et al., 2008).
This site runs a 0D test and a 1D test on the same computer, so the gap between your two numbers is the interesting part. Both include your computer's own input and display delay, which is why neither compares directly with lab figures. A 2D test is the natural next step.
- Buxton, W. (1983). Lexical and pragmatic considerations of input structures. Computer Graphics 17(1).
- Woods, D. L. et al. (2015). Factors influencing the latency of simple reaction time. Frontiers in Human Neuroscience 9, 131.
- Brenner, E. & Smeets, J. B. J. (2003). Fast corrections of movements with a computer mouse. Spatial Vision 16, 365–376.
- Brenner, E., Bom, S. & Smeets, J. B. J. (2026). Intercepting moving targets: does the visuomotor latency depend on whether one taps on the target or slides through it? Experimental Brain Research.
- McRuer, D. T. & Jex, H. R. (1967). A review of quasi-linear pilot models. IEEE Transactions on Human Factors in Electronics 8(3), 231–249.
- Oostwoud Wijdenes, L., Brenner, E. & Smeets, J. B. J. (2013). Comparing online adjustments to distance and direction in fast pointing movements. Journal of Motor Behavior 45(5), 395–404.
- Veerman, M. M., Brenner, E. & Smeets, J. B. J. (2008). The latency for correcting a movement depends on the visual attribute that defines the target. Experimental Brain Research 187, 219–228.