Why is reaction time slower on a phone than with a mouse?
A touchscreen has more steps between your finger and the page than a mouse button does, and each step takes time. The page code is the same on both. The hardware around it is what differs.
| Step | Mouse and monitor | Phone |
|---|---|---|
| Seeing the change | A 60 Hz screen redraws every 16.7 ms, so the green arrives about 8 ms late on average | Same idea. Many recent phones refresh at 120 Hz, which halves that wait |
| Registering the press | A microswitch closes and the mouse reports it within 1 to 8 ms, depending on its report rate | The glass scans for your finger on a cycle, a touch controller filters the signal, then the operating system passes it on |
| Finger travel | The finger rests on the button, so it moves less than a millimetre | The thumb hovers above the glass and has to cross the gap. You control that distance |
| Measured total | 71 to 87 ms for the whole chain on desktop platforms (Anwyl-Irvine et al. 2021) | 35 to 140 ms for screen plus touch sensor alone, across 26 phones (Nicosia et al. 2023) |
The two totals in the table measure different things, so don't subtract one from the other. The desktop figure covers the whole path, from the code deciding to change colour to the code logging the click, software included. The phone figure covers only the hardware.
Nicosia and colleagues built a robot that tapped the glass and watched the screen. They ran it on 26 phones priced from under 100 dollars to over 1,000. What stood out was the spread. On hardware alone, the fastest and the slowest phones were more than 100 ms apart. Your own phone sits somewhere in that range, and you have no way of knowing where.
Finger travel is the one row you can shorten. On a mouse, your finger already rests on the button. On a phone, your thumb floats above the glass, and the time it takes to come down is added to your score. That is why the reaction time test page tells you to keep your thumb hovering just above the screen.
How big is the gap in practice?
On average, it is tens of milliseconds. Every large dataset that recorded the device shows the same pattern. The biggest comes from TestMyBrain, where Passell and colleagues (2021) compared 59,587 visitors. Even after controlling for age, gender and education, people on mobile devices were significantly slower on reaction time tests than people on laptops and desktops. The gap was largest on Android phones. The authors put the difference down to device delay, not to the people.
This site sees the same thing. Each run is tagged as touch or mouse, and nothing about who you are. Across about 360 reaction runs collected through October 2026, the touch median was 28 ms slower than the mouse median. We treat that as a direction only. The sample is far too small for a percentage, so the site doesn't show one.
The guide to reaction time by age covers another side of this effect. Older visitors tend to use slower phones, and that makes browser age charts look steeper than lab ones.
A gap that size is about as big as the swing inside one person. Woods and colleagues (2015) found that a person's trials vary by about 40 ms around their own average from round to round. So one phone run and one laptop run can be 60 ms apart without anything changing in you.
Which number should you trust?
Trust both, but keep them apart. The laptop number is not closer to your nervous system. It just has less hardware in the way. Neither one is a lab figure, because a lab measures its own hardware delay and subtracts it, and a browser can't. The good reaction time chart shows the screen scale and the lab scale side by side.
Compare a phone with a phone, and ideally with the same phone. Your records here are saved in the browser you used, so your phone history and your laptop history never mix. To see whether you are getting faster, watch the median on one device over several weeks.
Whether your phone is a slow phone is a different question. Only comparing it with other phones can answer that.
The leaderboard follows the same idea. Each entry has a small phone or desktop mark. So when a 190 ms score sits above your 300, check whether it came from the same kind of input before you compare yourself with it.
What does this site do about touch?
It does three things, and none of them is a correction.
First, it records the tap at the moment of contact. A browser sends one event when your finger lands and another when it lifts. Older mobile browsers held the second one back by about 300 ms, in case a double tap was coming. This test uses the contact event and turns off the double-tap wait, so the time until you lift your finger doesn't count.
Second, it doesn't subtract a phone allowance. The measured range of 35 to 140 ms is too wide for one number to be fair to everyone. A flat subtraction would flatter the slow phones and penalise the fast ones. We think showing the raw number is fairer, and this page is here to help you read it.
Third, it records only the input type. The anonymous run statistics store touch or mouse and nothing else about your hardware, and the leaderboard shows the same mark next to each name. The privacy page lists exactly what is kept.
How do you get a fair score on a phone?
- Use the same phone every time, held the same way. Its hardware delay stays fixed, so it drops out when you compare yourself with yourself.
- Keep your thumb low. Hover it a few millimetres above the screen and tap lightly. The time is taken at contact, so pressing harder does nothing. The distance your thumb falls is the only part of the phone's delay you can shorten.
- Keep the page in front. Browsers slow their timers in background tabs, so the test stops and shows a notice the moment the tab is hidden. On a phone, that also means not switching apps between rounds.
- Turn on a 120 Hz screen setting if your phone has one. It cuts the display wait from about 8 ms to about 4 ms on average. That won't close the gap to a desktop, but it helps.
Frequently asked questions
Should I subtract something from my phone score?
No. Across phones, the measured hardware delay runs from 35 to 140 ms, and you don't know where your phone falls in that range. Any number you subtract would be a guess. Keep the raw score and compare it only with your own earlier scores on the same phone.
Is a laptop trackpad a mouse or a touchscreen?
This site counts it as a mouse, because the pointer is precise and the page never sees a finger on the glass. Still, a trackpad click has its own physical travel, and most laptop screens run at 60 Hz. So a trackpad tends to score slower than a wired gaming mouse on a fast monitor. It belongs in the mouse history, but keep that in mind when you compare it with a desktop score.