The Flashtap wiki
Reaction time: the test, the averages and how to improve
A reaction time test measures the delay between a signal and your response, in milliseconds: in one study of 1,469 adults tested on computers, the mean was 231 ms, of which about 18 ms came from the screen and the mouse. The idea fits in a sentence: wait without knowing when, then respond at once, without going early.
On Neurostim, Flashtap runs that protocol five times in a row and keeps your mean. This guide gives the test's exact rules, what the number really contains, the habits that bring it down and what research can actually say about a good time.
In short
- Genre: a simple reaction time test, played solo: one signal, one possible move.
- Goal: hit the target as soon as it lights up, five times. Your score is the mean of the five reactions, and the lowest wins.
- Length: each wait lasts a few seconds, drawn at random. A run adds up to about 13.5 seconds of waiting, plus whatever time you take between signals.
- Daily challenge: the same waits for everyone, renewed at midnight, Paris time. Only your first try goes on the leaderboard.
- Free run: fresh waits every game, with your best try of the day kept.
- Account: none needed. You start as a guest, with a place on the leaderboard just like a member.
- Devices: phone or tablet by touch, computer with a mouse or the keyboard.
The rules of the reaction time test
How a run goes
Five signals follow one another. For each, the target stays dark for a randomly drawn wait, then lights up: press at that moment. The time between the light and your press is your reaction for that signal.
The diagram puts one signal on a timeline, left to right: a press during the wait is a false start, a press after the light gives your reaction.
random wait reaction
|--------------------------------|----------|
light press
The score
The score adds two things: the mean of the five reactions, rounded to the millisecond, and 50 ms per false start. On a tie, the best single reaction of the run separates the players.
False starts
Pressing during the wait shows "Too soon!" and adds 50 ms to your score. The wait then restarts for the same signal, with a new length, so a false start tells you nothing about when the target will light up. A press so close to the light that it cannot be a real reaction is treated the same way.
Missed signals
A target left unpressed for 1.5 seconds counts as 1,500 ms, and the run moves on to the next signal.
One try, in one go
Reloading or leaving the page abandons the try, which then does not count at all. On the daily challenge, an abandoned first try is lost: the later ones stay practice.
What the test really measures
One number, several stages
Your score adds up the time to perceive the light, the time to trigger the movement and the delays of your equipment. In the study by Woods and colleagues, the display added 11 ms on average and the mouse 6.8 ms: taking away those 17.8 ms moved the mean from 231 to 213 ms. So part of the number is not yours.
Simple and choice reactions
As early as the 19th century, Donders observed that a simple reaction is shorter than one where you must first recognise the signal, itself shorter than one where you must choose between responses. Flashtap is the simplest case: one signal, one move. In Hueclash, by contrast, you must choose between four colours despite a misleading word, and every answer takes much longer. Dotshot adds a movement towards a target that keeps changing place, and Gridrush a visual search.
The wait is part of the test
Researchers call this wait the foreperiod. When its length varies at random from trial to trial, reactions are faster on average after long waits. The classic explanation is that the signal grows more and more likely as time passes, which lets you prepare better; Los and colleagues argue instead for an explanation based on memory traces of earlier trials. The debate is still open, but the effect shows why Flashtap varies its waits: with a fixed delay, you would end up answering the rhythm rather than the signal.
Why you should only compare on one device
A 60 Hz screen only draws a new image every 16.7 ms, and every mouse, keyboard or touch screen has its own delay. Flashtap cannot take those delays out of your score. A 10 ms gain between your phone and your computer proves nothing; the same gain on the same device, over several runs, means something.
Getting faster
- Do not guess. A false start weighs as much as a reaction 250 ms slower, and it makes you wait all over again. It is the first thing to get rid of.
- Start the signal when you are ready. You are in charge between signals: breathe, reset your finger, then go. Rushing on right after a bad reaction gains you nothing.
- Stay ready to the end of the wait. A wait that drags on makes you want to press: keep your finger still and let the signal decide.
- Work on consistency. A mean of five is hard on a lapse: one reaction at 1,000 ms among four at 250 ms gives 400 ms.
- Set your finger before you begin. Keep it just above the target or resting on the key: that takes travel out of the movement without changing your attention.
- Play short sets. In the Woods study, times rose from 228 to 237 ms between the first and last blocks of a 120-trial set: fatigue shows up within a single session.
Benchmarks: Flashtap's waits and published averages
The measured waits
Across 2,000 runs drawn by Flashtap's generator in November 2026, or 10,000 waits, a run with no false start adds up to 13.5 seconds of waiting in the middle of the series, and 90% of runs stay between 10.3 and 16.7 seconds. The draw favours no length. Within a single run, the longest wait typically exceeds the shortest by about 2 seconds: there is no rhythm to pick up.
Betting on the moment of the signal does not pay, because of how the score works. Gaining 50 ms on one signal takes only 10 ms off the mean, whereas a false start adds a full 50 ms to it: you would need five such bets to come off to pay for a single miss.
What a mistake costs
These scores come from the game's real calculation. The 250 ms is a worked example, not a level to aim for.
| Run | Score |
|---|---|
| Five reactions at 250 ms | 250 ms |
| The same, with one false start | 300 ms |
| The same, with two false starts | 350 ms |
| Four at 250 ms and one at 400 ms | 280 ms |
| Four at 250 ms and one at 1,000 ms | 400 ms |
| Four at 250 ms and one missed signal | 500 ms |
Published averages
- On a computer: 231 ms across 1,469 adults and 237.8 ms across 189 others (213 and 220 ms without the equipment delay), according to Woods and colleagues, who note that published means run from 233 ms to nearly 400 ms depending on the protocol.
- In the lab: about 190 ms for a visual signal and 160 ms for a sound among college-age people, according to Wikipedia's overview of mental chronometry.
- On the track: 166 ms for male sprinters and 189 ms for female sprinters at the 2008 Beijing Olympics, as the mean of each athlete's best time, to a gun and on blocks with force sensors, so not on a screen.
- With age: simple reaction time slows little until around 50 in a British survey of more than 7,000 adults, while choice reaction time slows throughout adult life. Woods measures about 0.55 ms more per year.
These figures come from different protocols, often more than a hundred trials with practice: none compares exactly with five signals on your phone. To see where you stand on identical waits, look at the challenge's leaderboard instead.
The history of reaction time
- 1868: Franciscus Donders publishes his subtraction method, comparing simple and choice reactions to estimate how long a mental operation takes.
- Francis Galton puts more than 10,000 men, women and children from the London public through his tests; according to Woods, his visual times fall below 190 ms in young people.
- 1952: Hick shows that reaction time rises by a constant amount each time the number of choices doubles, now known as Hick's law.
- 2009: a study commissioned by athletics' international federation finds that top sprinters sometimes react in 0.08 s, under the 0.1 s limit of the false start rule.
- 2011: working from the Beijing 2008 starts, Lipps and colleagues estimate that a man can react in 0.109 s and a woman in 0.121 s, once in every thousand starts.
- 2015: Woods and colleagues detail what makes simple reaction time vary, from equipment to age.
Flashtap borrows the protocol without claiming that heritage: a random wait, a signal, a move. It is a game, not a medical test, and its number also depends on your equipment.
Frequently asked questions
What is reaction time?
Reaction time is the delay between a signal appearing and the start of your voluntary response. On a screen, it is counted in milliseconds and also includes the delays of the display and the mouse or keyboard. It grows as soon as you have to recognise the signal or choose among several responses.
What is the average human reaction time?
For a visual signal and a single move, studies give means of roughly 190 to 240 ms in adults. Woods and colleagues measured 231 ms across 1,469 people, including 18 ms of equipment delay. Protocols vary a lot, and some published means climb to nearly 400 ms.
What is a good reaction time?
Around 230 ms on a visual signal, you match the computer-based mean measured by Woods; under 200 ms, you are nearing the 190 ms quoted for college-age people in the lab. Compare yourself with yourself first, on the same device: your false starts and your lapses count as much as your best single reaction.
How can you improve your reaction time?
Get rid of false starts and lapses first, then play short sets, rested, on the same equipment. Long training may help: 50 hours of action games cut times by 13%, against 6% for a control game, among 25 non-players, in a study on visual tasks reviewed by Dye and colleagues. A review by Simons and colleagues points out that gains show up mainly on the trained task.
Why is my reaction time inconsistent?
Some variation is normal: Flashtap's waits differ by about 2 seconds within a run, and your attention and readiness shift from one signal to the next. A single lapse moves the mean of five a lot, since one reaction at 1,000 ms among four at 250 ms gives 400 ms. Short, rested sets help.
What is the average reaction time for drivers?
France's road safety agency gives a reaction time of 1 to 2 seconds, and uses 1 second in its stopping distance chart, which is 14 metres at 50 km/h. That is far longer than on a screen: at the wheel, you must spot an unexpected hazard, then move your foot to the brake.
What is the fastest human reaction time?
As far as we know, no organisation keeps an official reaction time record. In sprinting, the rules call a false start under 0.1 seconds; a 2009 study recorded reactions of 0.08 seconds among the best, whereas a 2011 analysis puts the limit near 0.109 seconds. Times shown on online test sites depend too much on equipment to serve as records.
Is reaction time the same as a reflex?
No: a reflex is an involuntary, nearly instantaneous response, like the knee jerk or a newborn's Moro reflex. A test like Flashtap measures a voluntary reaction: you decide to press as soon as you see the signal. People often say "reflex test", but they nearly always mean a reaction time test.
Is a reaction time test better with the spacebar or the mouse?
No answer holds for everyone, because each device has its own delay: the mouse in the Woods study added 6.8 ms, and another will differ. Flashtap accepts touch, the mouse, Space and Enter: try them in a free run, then stick with the same one so that your scores stay comparable.
Is Flashtap free, with no account?
Yes. Flashtap is free to play in your browser, with nothing to install. Your first game creates a guest profile, and that profile appears on the leaderboard. If you create an account later, it takes over the same results and makes them available on all your devices.
Sources
- David L. Woods et al., Factors influencing the latency of simple reaction time, Frontiers in Human Neuroscience 9, 2015.
- David B. Lipps, Andrzej T. Galecki and James A. Ashton-Miller, On the implications of a sex difference in the reaction times of sprinters at the Beijing Olympics, PLOS ONE 6 (10), 2011.
- Geoff Der and Ian J. Deary, Age and sex differences in reaction time in adulthood, Psychology and Aging 21 (1), 2006, pp. 62-73.
- Sander A. Los et al., Outlines of a multiple trace theory of temporal preparation, Frontiers in Psychology 5, 2014.
- Matthew W. G. Dye, C. Shawn Green and Daphne Bavelier, Increasing speed of processing with action video games, Current Directions in Psychological Science 18 (6), 2009, pp. 321-326.
- Daniel J. Simons et al., Do "Brain-Training" Programs Work?, Psychological Science in the Public Interest 17 (3), 2016.
- Wikipedia, Mental chronometry, False start and Reflex, checked on 7 November 2026.
- French road safety delegation, leaflet on speed, in French, September 2008, checked on 7 November 2026.
Your turn
Today's Flashtap challenge is the same for everyone until midnight, Paris time. No account needed: you get a guest name with your first game.
Games of the same kind
This guide also exists as plain text, for AI assistants: markdown version. All the guides of the wiki.