During a physics experiment at school we needed to determine the velocity of a car by using a stopwatch to measure the time it took to cross a line. We were told to calculate the uncertainty of the measurement using the ~250ms average reaction time, but is this correct?

The GIF below shows a test I did by myself to see how accurately I could stop the stopwatch pointer at the 5 seconds white line/mark. On average I got an error of about ~20ms.

enter image description here

In this sense, there seems to be different kinds of reaction times: one where we are able to predict or expect what will happen, and another when we have to react suddenly and without warning, which would be the commonly told 250ms time.

Are there studies on this topic that differentiate between these twos reaction times? If there is indeed a difference, does the "predictive" reaction time change depending on the context? In the stopwatch case it is possible to get an accurate result because the pointer moves at a slow and constant speed, but in the case of a more erratic movement the reaction time could be probably longer.

  • 2
    $\begingroup$ Point of order: There are 1000 ms in a second, you're prediction was 20 ms off (plus measurement error/indeterminacy). $\endgroup$ Sep 14, 2022 at 2:06
  • 2
    $\begingroup$ This idea of calibrating your measurement error in a separate experiment is very cool. You might want to consider cases where the pointer is speeding up or slowing down, and various different speeds, to see if this affects your errors. The closer the calibration test is to the real car experiment the better! The stopwatch scenario might be a bit easy so 2ms error is probably optimistic. The 250ms number is about what you'd get if you couldn't see the car coming, so pretty pessimistic. Not unreasonable to use the pessimistic one if you don't know, but if you have data use that! $\endgroup$ Sep 14, 2022 at 7:14
  • $\begingroup$ @JiminyCricket. Thanks for poiting out! I forgot to add a zero $\endgroup$
    – ordptt
    Sep 14, 2022 at 16:01

2 Answers 2


I wouldn't call this "reaction time" but rather a form of accuracy: you're timing your press to the advance of the stop watch.

Reaction time is latency to respond to some stimulus, but in the situation you describe you aren't responding to the bar reaching the mark, the stimulus starts long before that in the motion of the stopwatch.


I reckon the 250ms referenced is reaction to an unknown/random effect.

I watch F1 and at every start they show the reaction time of the first two drivers as even 100ms faster reaction can be very important in determining who'd be #1 after the first corner.

It's usually in the .2 - .3 range for them

Try using this to gage your reaction time instead of trying to time a stopwatch: https://f1-start.glitch.me/

The same F1 driver that needs 253ms to register it randomly turning green (similarly to a race start) https://www.youtube.com/watch?v=rWtWC3FsY5M

Only needs 0.0069s to time it to exactly 6 seconds https://www.tiktok.com/@mclaren/video/6953908004311846150?is_from_webapp=v1&item_id=6953908004311846150&lang=en

Meaning, as Bryan alludes that his reaction to the timer hitting 6s seems to have started at around 5.75s-5.8s

  • 2
    $\begingroup$ For top-level sprinters the false start rule is set at 100ms after the start gun fires, although lowering it to ~80ms might be good as some athletes are closer to 90ms reaction time - .2-.3 is probably a good reaction time for most people not trained, and also not anticipating (i.e. looking at something nearing completion) $\endgroup$ Sep 14, 2022 at 17:09

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.