Guides › Hardware tests

How much of your reaction time score is actually your screen?

An online reaction time test reports one number, but that number is the sum of several delays: some belong to you, and some belong to your screen, your mouse and your computer. To find out how much the equipment adds, we pointed a bot at our own reaction time test. The bot had no reaction time at all: it clicked the instant the green colour reached the screen.

Short answer: the screen alone can add up to about 17 ms on a 60 Hz display, small next to a typical human score but enough to matter when comparing. A zero-delay bot still scored between 0 and 17 ms, 10.6 ms on average, purely from waiting for the next screen refresh. Real monitors, mice and keyboards add more on top. For young adults, laboratory studies typically report simple visual reaction times of roughly 200 to 250 ms. Compare scores only on the same setup.

Where the milliseconds go

From the moment the test decides to show green to the moment it receives your click, the signal passes through a chain:

  1. Waiting for the next frame. The page changes the colour, but the screen only redraws at fixed intervals: every 16.7 ms at 60 Hz, 6.9 ms at 144 Hz.
  2. The display itself. The image has to be sent to the panel and the pixels have to change. Monitors differ widely here; TVs in picture-enhancement modes can add tens of milliseconds.
  3. You. Your eyes register the change, your brain decides, and the signal travels to your finger muscles. This is the part people mean by "reaction time".
  4. The input device. A USB mouse or keyboard reports to the computer at a fixed polling rate. At the common 125 Hz that is every 8 ms, so a press can wait up to 8 ms before the computer even knows about it; at 1000 Hz, at most 1 ms. Wireless devices and touchscreens add their own processing.
  5. The browser. The operating system and browser deliver the click event to the page, which records the time.

The test: a bot with zero reaction time

We ran our test 40 times in Chromium with a 60 Hz frame rate. Instead of a human, a script inside the page watched each frame. As soon as a frame contained the green box, it fired the click immediately, with no physical mouse involved. This removes steps 2 to 4 above and leaves only the frame wait and the browser.

MeasureResult
Fastest score0 ms
Median score11 ms
Average score10.6 ms
Slowest score17 ms

The scores were spread almost evenly between 0 and one frame (16.7 ms). That is exactly what you would expect: the green colour is requested at a random moment, so sometimes the next frame comes almost immediately and sometimes the page has to wait nearly a full frame. On a 144 Hz monitor the same effect would range from 0 to 7 ms; on a 240 Hz monitor, from 0 to 4 ms.

The bot's score is the floor that no person could beat on this setup. A real monitor and mouse sit on top of it, which is why typical online results land somewhat above the laboratory figures.

What is a normal reaction time?

For a simple task like this one, where there is a single signal and a single response, researchers measuring young adults with carefully calibrated equipment typically report average visual reaction times of about 200 to 250 milliseconds. Woods and colleagues, in a large study published in 2015, also documented how reaction time lengthens gradually with age, and how much the measured value depends on the hardware used to collect it.

Online, with ordinary monitors and mice, results in the 230 to 300 ms range are common. Scores below 150 ms on a regular setup usually mean the click was anticipated rather than a reaction; our test randomises the waiting time between 1 and 4.5 seconds to make anticipation hard.

How to make your scores comparable

ChangeTypical effect on your score
60 Hz to 144 Hz monitorAverage frame wait drops from about 8 ms to about 3.5 ms
TV "game mode" on versus offCan save tens of milliseconds on many TVs
125 Hz to 1000 Hz mouse pollingAverage input wait drops from about 4 ms to about 0.5 ms
Mouse click versus tapping a phone screenTouch screens often add more delay than a wired mouse
Resting your finger on the button versus hoveringOften larger than all the hardware effects combined

The practical rule: compare yourself with yourself, on the same device, the same input method and a similar time of day. If you test a new mouse or monitor, run several full sets with each and compare the averages, not single attempts. Our mouse tester and keyboard tester can confirm that every button and key registers correctly before you start.

Why our test measures from the press, not the release

The test listens for the mouse button going down, a touch starting, or Space or Enter being pressed, rather than for the click to complete. Waiting for the release would add the time your finger takes to lift, typically several tens of milliseconds, which has nothing to do with how fast you reacted. The timer uses the browser's high-resolution clock, performance.now(), which is precise to well under a millisecond.

Common questions

Can practice improve my reaction time?

Results usually improve over the first few sets as you get used to the task, then level off. Sleep, alertness and attention also change the result.

Why is my phone result slower than my computer result?

Touch screens scan for touches at their own rate and the phone processes the gesture before the browser sees it. Many phones also have 60 Hz screens. Both add delay that is not yours.

Is this a medical test?

No. It measures visual reaction with your particular device. It cannot assess fitness to drive or diagnose any condition.

References

  1. Woods DL et al. (2015), Factors influencing the latency of simple reaction time, Frontiers in Human Neuroscience 9:131
  2. MDN: Performance.now()
  3. MDN: Window.requestAnimationFrame()
  4. USB HID specification: polling intervals (USB-IF)

Written and checked by Ria Kim for GlowHalo; figures without a source link come from the steps shown on this page. Companion tool: Reaction Time Test. Spotted an error? Write to help@glowhalo.org. Chromium is a trademark of its owner; GlowHalo is not affiliated with it.

More guides

Hardware testsWhy a website says your 1080p screen is 1536 pixels wide: resolution, viewport and pixel ratioWe ran our screen resolution tool under seven emulated device profiles. Why a phone with 1179 physical pixels reports 393, and which number matters for what.Read guide →Hardware testsSome key combinations don't register? Testing keyboard ghosting and rolloverOur browser test registered 26 keys held at once, so a limit you hit comes from the keyboard. How ghosting and rollover work, and a 5-minute test routine.Read guide →