By the numbers, this should be impossible
Baseball is the place to begin, simply because its numbers are the tidiest. Although the mound sits 60 feet 6 inches from the plate, release happens closer to 55 feet out, which puts a 100 mph fastball in the air for roughly 375 ms and a 95 mph one for roughly 400 ms. That is everything the hitter has to spend.
Now watch it drain. For a healthy adult responding to a bare stimulus, with nothing to identify and no option to weigh, the cost is 200 to 250 ms, and that is precisely the quantity the Classic Reaction Time Test captures. Executing a competitive swing takes another roughly 150 ms, and the moment it begins it is ballistic: there is no meaningful way to steer the bat afterwards. The call on whether to swing, and where to put the bat, therefore has to be settled by about the 200 ms mark, with the ball still only halfway home.
Under a quarter of a second, then, to pick the ball up, sort fastball from slider, work out where it will cross the plate, and commit. Read the numbers honestly and the conclusion is unavoidable: reacting cannot get this done, by anyone. The work is being carried by something that is not reaction time.
| Stage | Cost | Note |
|---|---|---|
| Ball in flight (release to plate) | ~375 ms total | Everything available, at 100 mph from roughly 55 ft |
| Visual reaction to a bare stimulus | 200 to 250 ms | Simple reaction time, with nothing to identify and nothing to choose |
| Swing execution | ~150 ms | Ballistic from the start, so no mid-flight correction is possible |
| Latest possible commit point | ~200 ms after release | The ball has covered roughly half the distance |
| Time left over for reacting | None | Which is precisely why elite hitting is not reaction |
The body reveals what the ball conceals
The experiment that resolved the question is elegantly crude. Play an athlete footage of an opponent in motion, stop the clip before the ball is released, and ask where it is headed. Perform better than chance and the information cannot have come from the ball's flight, because there was none to see. It came from the body.
Abernethy and Russell applied exactly this to badminton in 1987, testing 20 experts against 35 novices while varying how much of the stroke remained visible and which regions of the display were masked. The experts called the landing position using cues that arrived earlier than anything the novices could exploit. Their attention was not on the shuttle at all, but on the stroke about to produce it.
Cricket sharpens the point. Across four occlusion experiments, Müller, Abernethy and Farrow compared world-class, intermediate and low-skilled batsmen. The obvious late cues were available to everybody. What set the world-class group apart was their ability to draw usable information from particular early sources, above all the bowling hand and arm, which the weaker players were simply deaf to. Expertise of this kind is not a quicker nervous system. It is a nervous system tuned to a signal that arrives sooner.
Tennis reveals just how far in advance that signal can sit. Using in-situ temporal occlusion on the service return, Farrow and Abernethy found that expert prediction accuracy survived even when players had to produce a genuine return stroke rather than merely say an answer aloud. A 200 km/h serve grants the receiver about a third of a second to read the flight and manufacture a reply, so direction has to be taken from the server's trunk, shoulder and toss well before racquet meets ball. Nothing on this site sits closer to that problem than the Choice Reaction Time Test: what costs you is not the movement but the decision.
What the eyes are really doing
As coaching goes, keep your eye on the ball is both physically impossible and empirically false. Bahill and LaRitz showed as much in 1984, watching graduate students, college players and one major leaguer follow balls hauled toward them at 60 to 100 mph. Close to the plate the ball's angular velocity outruns smooth pursuit, so the batters flicked their eyes out in front of it with a predictive saccade, settling them on the ball's future position instead of its current one.
Land and McLeod uncovered the identical strategy in cricket in 2000, and showed that its timing is what divides skill levels. A batsman watches the release, fires a predictive saccade at the spot where the bounce is expected, waits for the ball to arrive there, then follows it for 100 to 200 ms after it lifts. What marked out the better batsmen was a shorter latency on that opening saccade: their prediction was committed to sooner. The sample was three batsmen, so give it the scepticism it has earned, though Mann, Spratford and Abernethy reproduced the pattern in 2013 with elite batters, observing the head stay with the ball while gaze ran out ahead of it.
Widen the lens and the meta-analytic picture agrees. Drawing on 42 studies and 388 effect sizes, Mann, Williams, Ward and Janelle reported that experts deploy fewer fixations, each held longer, alongside extended quiet-eye periods, and that they come out ahead on both response accuracy and response time in sport-specific tasks. Experts are not looking more intently or more quickly. They are looking at fewer, better-chosen places, sooner.
| Event | Time available | Why reacting is not enough |
|---|---|---|
| Baseball, 100 mph fastball | ~375 ms from release to plate | A ballistic swing costs ~150 ms, forcing a commitment by ~200 ms |
| Cricket, fast bowler | Around half a second, split in two by the bounce | A predictive saccade to the bounce, then 100 to 200 ms of tracking |
| Tennis, 200 km/h serve | Roughly a third of a second | Direction must come from the server's body ahead of contact |
| Football penalty, 11 m | Ball flight 500 to 600 ms | A full-stretch dive alone needs ~600 ms, so commitment comes at or before the strike |
| F1 race start | Reaction is the event | There is nothing to anticipate: elite reactions gather near 200 ms, the best around 180 ms |
| 100 m sprint start | Reaction is the event | Under 0.100 s counts as a false start rather than a quick reaction |
The goalkeeper with no time to spare
No example is cleaner than the penalty kick, because the shortfall here leaves no room for argument. Struck from 11 metres, the ball is at the goal in roughly 500 to 600 ms. Reaching a corner at full stretch takes a goalkeeper about 600 ms. What the movement demands is not less than what the clock offers; it matches it or exceeds it.
Information from the ball's flight is therefore largely worthless. Wait until the direction is visible and the corners are already lost. Studies of goalkeeper behaviour find the dive timed to where the kicker plants the non-kicking leg, a cue that appears before contact and keeps its value no matter how little time remains.
That recasts what a save really is. Not a rapid reaction, but a wager laid early on a body cue, with the skill lying in reading that cue and in hiding your own. Every pre-emptive sport runs on the same logic: the striker's task is to delay the tell, or aim it in the wrong direction.
The single event that is purely reaction
Everything so far has been about escaping reaction time. The sprint start is where that escape is unavailable. No cue exists to read, no body to anticipate, only a gun and a rule forbidding you from being too quick.
World Athletics counts any reaction below 0.100 s as a false start, and the current one-strike policy turns that into immediate disqualification. The cut-off assumes no human can turn a sound into force against the blocks inside 100 ms, making anything faster anticipation rather than reaction. Devon Allen lost his place in the 110 m hurdles final at the 2022 World Championships over a reaction of 0.099 s, a single thousandth of a second on the wrong side of the line.
That assumption is less solid than the rule suggests. Pain and Hibbs fitted starting blocks with piezoelectric force transducers and recorded nine athletes under four conditions. Five averaged below 100 ms in at least one of them, 20% of every start in the opening two conditions landed under 100 ms, and the neuromuscular component fell as low as 85 ms, with EMG latencies beneath 60 ms. There is a floor, but it may lie below 0.100 s, which makes the rule an administrative boundary rather than a physiological one.
This site draws a comparable line for comparable reasons. A click quicker than 80 ms is read as anticipation rather than reaction: it is logged so you can see it, flagged void, and left out of the round count. Without that, a session of nothing but guessing could manufacture a flawless score. The Drag Racing Reaction Time Test and the Red Light Green Light Test come nearest to a real start line here, and each penalises the guess much as a starter would.
What a click test does and does not reveal
Honesty counts for more than marketing here. Simple reaction time behaves almost like a fixed trait. Athletes do beat non-athletes on generic reaction tasks, but only slightly, and the wide expert-novice gulfs open up elsewhere entirely: in sport-specific anticipation. A combat-sports meta-analysis put the expert advantage at d = 1.51 for response accuracy and d = -0.91 for reaction time on sport-specific tasks, effect sizes no plain stimulus comes close to generating.
In practice that makes a click test a fine instrument and a poor gym. It reads the fixed floor of your visuomotor chain, and that is genuinely worth knowing, because the floor shifts with sleep, caffeine, alcohol, fatigue and age. It will report the state you are in today. It will not improve your hitting against a curveball, for the simple reason that whatever does improve it is absent from the test.
Two consequences follow. The first: to track condition, hold every other variable steady and read the trend, since what sleep, caffeine and alcohol are doing to you surfaces plainly in a weekly median, and your stats page retains all of them. The second: to get better at your sport, train against that sport's own cues, occluded and under time pressure, rather than against a coloured square in a browser tab. The Go/No-Go Test does at least introduce the inhibition problem real sport imposes, which puts it one step nearer than a bare stimulus, and no nearer than that.
Work on the cue, not the clock
The literature converges on an awkwardly precise instruction: give up on trimming milliseconds from your reaction and start shifting your decision point forward. Everything below flows from that.
| Do this | Because | Where it comes from |
|---|---|---|
| Watch the body, not the ball | The usable information lands before release, and experts are the ones acting on it | Abernethy & Russell (1987); Müller et al. (2006) |
| Learn your sport's earliest reliable cue | Elite performers read particular early sources (bowling hand and arm, non-kicking leg) that others overlook | Müller et al. (2006); Dicks et al. (2021) |
| Commit to a prediction sooner | Better batsmen were distinguished by a shorter latency on the first predictive saccade | Land & McLeod (2000) |
| Fewer, better-placed fixations | Experts hold fewer fixations for longer, with extended quiet-eye periods | Mann et al. (2007) |
| Practise occluded | Stopping the footage before release pushes you onto the early cue rather than the ball flight | The temporal occlusion paradigm itself |
| Use click tests as a state gauge, not as training | Practice barely shifts simple reaction time; the expert gap lives in sport-specific tasks | Wang et al. (2022) |