The Science of Reaction Time: What Really Happens

The first time I saw my own reaction time on a screen (248 milliseconds), I didn’t believe it. A quarter of a second? I play games for a living’s worth of hours. Surely I was faster than that.

I wasn’t. Almost nobody is. And once I understood why, the number stopped feeling like an insult and started feeling like a small miracle, because a lot has to happen in that quarter of a second.

A reaction is a relay race, not a sprint

When the screen turns green and you click, it feels like one instant thing. It isn’t. It’s a chain, and each link costs time.

Here’s the route the signal takes:

Light hits your retina. Photoreceptors at the back of your eye have to turn that light into an electrical signal. This step (phototransduction) is chemical, and chemistry is slow. It’s one of the biggest single delays in the whole chain.
The signal travels to your visual cortex. Up the optic nerve, through a relay station deep in the brain, to the back of your skull where seeing actually happens. Call it 50 to 100 milliseconds before “there is light” becomes “I have seen light.”
Your brain decides. This is the part people underestimate. Recognising the change and committing to “go” is usually the largest and most variable chunk of the whole reaction. It’s also the part training can move.
The command goes out. Motor cortex fires, the signal runs down your arm, the nerve meets the muscle, the muscle contracts, your finger drops.

Add it all up and you land somewhere around 200 milliseconds on lab equipment, a bit more on a screen. The retina and the muscle are fast and fairly fixed. The thinking in the middle is where the time (and the difference between people) actually lives.

Why ~200 ms is a floor, not a target

People ask me all the time how to get under 100 milliseconds. You can’t. Not really. The plumbing won’t allow it.

That number isn’t arbitrary. Sprint officials use it: under World Athletics rules, if a runner moves within 0.100 seconds of the gun, it’s ruled a false start, because the assumption is that no human can hear the gun, process it, and react that fast: they must have anticipated. (Interestingly, lab work on elite sprint starts, Pain and Hibbs, 2007, measured some auditory reactions below 100 milliseconds, with the pure neuromuscular component under 85 milliseconds, which keeps an argument about lowering the limit rumbling on.) Either way, the message holds: somewhere around a tenth of a second is the practical floor for a genuine simple reaction. Anything faster is a guess that happened to land.

So when a reaction test shows you 80 milliseconds, it didn’t catch a superhuman reflex. It caught you jumping the gun. That’s exactly why our reaction test throws out anything under 100 milliseconds. It’s protecting the number from your own anticipation.

Sound beats sight, and touch beats both

Here’s a fact that surprises people: you react to a sound faster than to a light.

It comes back to that first step. Turning sound into a nerve signal in your ear is quicker than turning light into one in your eye: less chemistry, more mechanics. So auditory reactions tend to run roughly 25 to 50 milliseconds ahead of visual ones, and several studies put touch ahead of both. The ordering reported is consistent: tactile, then auditory, then visual.

You can feel this yourself in about two minutes. Take the audio reaction test and the visual one back to back. Most people are noticeably quicker on sound. Seeing it in your own numbers is far more convincing than me telling you.

More choices, more time: Hick’s law

Everything above is simple reaction time: one signal, one response. Real life is rarely that tidy. Usually you have to pick the right response from several, and that choosing costs extra.

There’s a neat law for it. Hick’s law says reaction time grows with the logarithm of the number of choices: RT = a + b · log₂(N), where N is how many options you have. As illustrative figures (they vary a lot across studies and tasks, not fixed constants), a lands around 200 ms and b around 150 ms per bit of information. In plain terms: going from one choice to two adds a chunk of time; going from two to four adds another, smaller chunk; and so on. Each doubling costs roughly the same, not each new option, which is why a cluttered interface punishes you more than you’d expect.

If you want to feel the difference, the choice reaction test makes you pick between multiple targets, and you’ll watch your time climb compared to the simple version.

What changes your number (and what doesn’t)

The middle of the chain, the deciding part, is sensitive to your state in ways the retina and muscle aren’t. The things that genuinely move it:

Age. Reaction time peaks in your early twenties and slows gradually after. The MindCrowd study, with over 75,000 people, found it slowing by roughly 7 milliseconds per year of age on average. It’s a curve that decelerates rather than a flat rate, so people in their seventies tend to land roughly 40 to 60 percent slower than their twenties peak. There’s more in the full average reaction time by age breakdown.
Sleep. Lose a night and your reactions get slower and, worse, less consistent. The occasional huge lapse is the real danger.
Caffeine. A moderate dose (research often uses around 5 mg per kilo of bodyweight) claws back some of that sleep-deprived slowdown. It’s a modest, variable effect, mostly from propping up attention rather than speeding the muscles.
Practice. You won’t rebuild your nerves, but you can trim the decision step. That’s why warming up before you play actually works.

What barely moves: the hardwired bits at either end. You’re not going to train your retina to transduce light faster.

The honest footnote

One last thing, because it’s the reason this whole site exists. Every number on a web reaction test (including ours) carries 10 to 50 milliseconds that has nothing to do with your brain. Your monitor only refreshes every 8 to 17 milliseconds, your mouse and operating system add a little more, and all of it lands on your score. That’s why the lab figure (200–250 ms) and the typical online figure (Human Benchmark’s public statistics, a large self-reported aggregate, put the median around 273 ms) look like they disagree. They don’t. They’re measuring the same reflex on different equipment.

We measure your display and show you a corrected time so you can compare fairly. The full breakdown is on the methodology page. And the best way to make all of this concrete is to stop reading and take the test, then come back and you’ll know exactly which part of that quarter-second is yours.

Sources

Written by Lokesh Rathore, Founder of ReactScore. Lokesh Rathore is the founder of ReactScore and the person behind everything on it. Reflexes and reaction time are a genuine fascination of his, so he built the test he wanted to use himself: one that measures honestly, corrects for the lag your screen and mouse add, and explains the numbers instead of hiding them. He keeps every figure on the site grounded in public datasets and published research. More about the author · how we measure →