A visual reaction time test measures the gap between light hitting your retina and your finger completing a click. For most adults that gap is somewhere between 200 and 300 milliseconds, with about 250ms being typical. The interesting part is what fills those 250 milliseconds — and how little of it is under your control.
Where the 250 milliseconds go
The number is not one process. It is a chain, and each link takes time:
- Retinal processing (~20–40ms). Photoreceptors do not respond instantly; converting light into a neural signal has a measurable lag.
- Signal transmission to visual cortex (~30–50ms). Nerve conduction is fast but not free.
- Perception and decision (~80–120ms). Recognising that something changed and deciding to act. This is the largest and most variable chunk.
- Motor command and muscle activation (~50–80ms). The signal travels back down to your hand, and muscle fibres take time to contract.
- Mechanical travel (~10–30ms). Your finger physically moving the switch, plus the switch's own actuation.
Only the middle section — perception and decision — responds meaningfully to practice or attention. The rest is largely fixed biology. This is why reaction time improves with training, but not by very much: you are optimising perhaps 40% of the total.
What the distribution actually looks like
Reaction times across a large population are not symmetrical. They cluster tightly at the fast end and trail off with a long tail toward slow results, because there are many ways to be slow and very few ways to be fast.
- Under 180ms: exceptional. Common among competitive esports players and people who have practised this specific task extensively.
- 180–220ms: fast. Well above average for an adult on a decent screen.
- 220–260ms: the middle of the distribution. Most healthy adults land here.
- 260–320ms: slower than average, and very commonly explained by fatigue, a slow display, or a touchscreen.
- Over 320ms: usually a hardware or attention issue rather than a physiological one.
A single result means almost nothing. Reaction times vary by 30 to 50 milliseconds between trials for the same person in the same session, which is why any test worth taking averages several rounds.
How much of your score is your equipment
This is the part most people underestimate. A reaction test measures the whole loop, including hardware, and hardware differences are large enough to swamp real skill differences.
| Source | Typical added latency |
|---|---|
| 60Hz display refresh | 0–16ms (average 8ms) |
| 144Hz display refresh | 0–7ms (average 3.5ms) |
| Display processing (TVs, some monitors) | 10–40ms |
| Wireless mouse polling | 1–8ms |
| Touchscreen input | 20–60ms |
| Browser frame scheduling | 0–16ms |
Adding it up: the same person can score 195ms on a 144Hz monitor with a wired mouse and 260ms on a phone. Neither number is wrong; they are measuring different systems. If you want to compare yourself to a friend, compare on the same device.
What genuinely improves it
Sleep
Sleep deprivation degrades reaction time faster and more reliably than almost any other variable. A night of restricted sleep can add 30 to 50 milliseconds and, more importantly, dramatically increases the variance between trials — occasional very slow responses called lapses. If you want a good score, be rested.
Anticipation, used correctly
Knowing roughly when a stimulus is coming reduces reaction time substantially — this is why sprinters react to a starting gun faster than to a random tone. Good reaction tests defeat this deliberately by randomising the delay. In real games, though, prediction is legitimate and it is where most of the actual advantage lies.
Reducing the decision
Simple reaction time (one stimulus, one response) is much faster than choice reaction time (which of several stimuli appeared, and which response matches). Every additional option adds roughly 30 to 50 milliseconds. Training in games is largely about converting choice reactions into simple ones through familiarity.
Caffeine, modestly
Moderate caffeine reliably produces a small improvement, typically in the range of 5 to 15 milliseconds, and mostly by reducing lapses rather than by speeding up your best trials. It is a real effect and a small one.
What does not help
- Straining. Tensing up before the stimulus makes you more likely to false-start and, paradoxically, slower when you do react.
- Staring at one point. Peripheral vision detects luminance changes faster than foveal vision. Soft-focusing the whole panel beats locking onto its centre.
- Repeating the test forty times in a row. Fatigue sets in within a few minutes and your average gets worse, not better.
Reaction time versus aim
These get conflated constantly and they are different capacities. A reaction test measures how fast you can begin a response. An aim test measures how fast you can complete a movement to a target you were not already pointing at — which is governed by Fitts's law, where time scales with the distance to the target and inversely with its size.
People with excellent reaction times are frequently mediocre at aim, and the reverse is just as common. If your goal is to get better at games, aim training transfers more directly, because almost no in-game situation is a pure simple-reaction task.
A sensible way to test yourself
- Use the same device every time, ideally with a wired mouse.
- Warm up with one throwaway set. First-of-the-day results are not representative.
- Take at least five valid trials and use the average, not your best single result.
- Test at a consistent time of day — circadian effects on alertness are large.
- Track the average over weeks, not sessions. Session-to-session noise will otherwise convince you of improvements that are not there.
Done that way, the number becomes a genuinely useful signal — mostly, it turns out, about how well you slept.