Am I missing something? (false starts)

[bman]

Was stumbling around the internets and found this- http://194.213.2.7/wps/PA_1_0_CI/IDMCom ... werServlet

The current false start criterion used by
the IAAF is based on an assumed minimum
auditory reaction time. If an athlete
moves sooner than 100 ms after the
start signal, he/she is deemed to have
false-started. The purpose of this study,
which was commissioned by the IAAF,
was to examine neuromuscular reaction
to the auditory signal used in the sprint
start and to determine whether the
100 ms limit is correct. Seven nationallevel
Finnish sprinters took part. A comprehensive
approach was used to study
force reaction on the blocks, the movements
of the arms and the activation
profiles of several muscles. The authors
found great variation in individual reaction
times and confirmed previous
reports of simple auditory reactions as
fast as 80 ms. They recommend that the
100 ms limit be lowered to 80 or 85 ms
and that the IAAF urgently examines
possibilities for detecting false starts
kinematically, so that judges’ decisions
are based on the first visible movement
regardless of the body part. This can be
done with a system of high-speed
cameras, which gives views of all the
athletes on the start line. With such a
system, it would be possible to change
the start rule so that no false starts are
permitted.

Was this discussed at some point? The reason I find this notable is that all discussions on this board led me to believe that all research concluded that anything faster than .120 in reaction time was a jumped gun.

[Half Miler]

Re: Am I missing something? (false starts)

I've mentioned this before somewhere, but I'm completely in favor of the high-speed camera solution (if one could be devised). This would eliminate the bogus "twitch" DQs that we've seen recently.

[Marlow]

Re: Am I missing something? (false starts)

I've mentioned this before somewhere, but I'm completely in favor of the high-speed camera solution (if one could be devised).

Me too! I've never heard of this, but it makes great sense. If there's no discernible movement before the gun goes off, is it a false start? With the no-false start rule (which I DO like), an athlete would be foolish to try and guess the gun, but if he does, more power to him, I guess.
[this is somewhat of a revision of my earlier stance]

[26mi235]

Re: Am I missing something? (false starts)

Count me out just because there is no movement before the gun. I do not want the winner to be, in part, the best guesser, just the best runner.

Maybe no movement before 0.09 after the gun but a look at the distribution of reaction times is quite convincing to me. It tails off very fast as you get near 0.11 and is almost non-existent below that level, with frequencies in the range of the false starts, not 'real reaction starts'. Mikli provided those data and these are from the top sprinters at the top races.

I have not even looked closely at the 0.08 piece, but it seems to me that a full-on starting reaction takes a bit more than some reaction where you are not initiating the entire body along a length of nerves running through the body to the feet pushing off.

[rainy.here]

Re: Am I missing something? (false starts)

Count me out just because there is no movement before the gun. I do not want the winner to be, in part, the best guesser, just the best runner.

Maybe no movement before 0.09 after the gun but a look at the distribution of reaction times is quite convincing to me. It tails off very fast as you get near 0.11 and is almost non-existent below that level, with frequencies in the range of the false starts, not 'real reaction starts'. Mikli provided those data and these are from the top sprinters at the top races.

I have not even looked closely at the 0.08 piece, but it seems to me that a full-on starting reaction takes a bit more than some reaction where you are not initiating the entire body along a length of nerves running through the body to the feet pushing off.

Charlie Francis said that Ben got under 0.10 (at least) once on what Ben said was a reaction. I believe it might happen once every 10 years by someone.

[lonewolf]

Re: Am I missing something? (false starts)

Uh, huh. How do they know those very fast reaction times were not also just anticipated?

[26mi235]

Re: Am I missing something? (false starts)

Looking at the distributions, I would guess that some of the marks in the range of 0.10 - 0.12 are 'flyers', especially with a higher percentage at the low end. It would be interesting to have the FS data to complete the statistical picture, but those data are not as easily available.

[Daisy]

Re: Am I missing something? (false starts)

For some reason the link to the pdf does not work for me, although, I found the article by going to their archive list (it's at the bottom, and if still can't see it search for year 2009 issue 1).

In describing the methodology they say

In the measurements, the athletes performed a total of five to eight sprint starts similar to the real race starts. Thus the conditions were as close as possible to real competitions.

With respect to Lonewolf's point about anticipation, one key thing that is not mentioned is how long was the pause from set to bang and was it random? If it is predictable then of course anticipation would be a problem. So a lot rests on what they mean by " similar to the real race starts".

The volunteers for this study were described as:

Seven Finnish national-level sprinters (four males and three women: age 24 ± 3 years; mass 71.2 ± 14.2kg; height 177 ± 7cm) participated in this study.

They measured a mean reaction time of 98 ms (standard deviation of 23) for the 'onset of leg force'. The average of the fastest for each athlete was 78 ms SD = 27

They measured a mean reaction time of 133 ms (standard deviation of 21) for 'force detection (25kg level)'. The average of the fastest for each athlete was 114 ms SD = 29

         'onset of leg force' 'force detection' (25kg level)
            avg  fastest         avg   fastest
athlete 1    97    76            150    129
athlete 2    75    58            114     88
athlete 3   123   115            151    144
athlete 4    73    58            109     88
athlete 5   128   110            158    150
athlete 6   113    92            142    121
athlete 7    79    42            112     78

I'm not sure what these numbers mean. Are the times for "force detection (25kg level)" meant to be equivalent to the IAAF reaction times published for the athletes? In the paper they say:

Until recently, the reaction movement was deemed to have started when the threshold of 25kg of force above the baseline in the set position was reached on either of the blocks. This has now been changed so the reaction movement is judged by the steepness of the rise of the force curve.

So possibly the RT's measured by the IAAF are somewhere in between the two above? It seems like a big flaw to go through this whole experiment and not have at least one of the measurements being directly equivalent to what we see in the actual races.

My nagging feeling is that how is it possible to have three of seven athletes consistently breaking the 100ms RT limit despite the fact it is rare in races all over the world? Something does not add up here.

[Pego]

Re: Am I missing something? (false starts)

I calculated those evoked responses through multiple synapses numerous times and no matter how liberal I am in allowing the best possible measurements (including possible shortcuts), I still cannot get under 120 msec. as the fastest possible reaction time.

[lonewolf]

Re: Am I missing something? (false starts)

Not even the guy behind you when the light turns green?

[26mi235]

Re: Am I missing something? (false starts)

The combination of no 'loss function' for false starting and the (seemingly real, see next) possibility of repeatability of commands makes this unlikely to be valuable. I will bet that if you timed the reaction of dragsters you would get better than 0.10 marks.

However, the notion that these 'relatively mediocre' athletes were significantly and systematically better than the highly trained athletes raises big issues. Furthermore, look at the difference in reaction times between 100, 200 and 400m starts -- you will see that the top 100m sprinters clean the clocks of those in the longer sprints, plus the sprinters in the 100 'redline' it more due to the increased importance.

Finally, there is a difference between 'reaction' as measured by the RT and the real starting reaction, which gh mentioned repeatedly in the WC 100m threads.

[Marlow]

Re: Am I missing something? (false starts)

If you just look at the empirical evidence of elite sprinters RTs in the DL and WC and OG, you see that the world's very best sprinters (presumably with exceptional RTs when compared to us 'normal' people) RARELY get down to even .120. Regardless of 'tests' (which I agree can be misleading), I think the IAAF's threshold of .100 is the right number, even if a large number of those between .100 and .120 are indeed 'anticipating'.

[mal]

Re: Am I missing something? (false starts)

I calculated those evoked responses through multiple synapses numerous times and no matter how liberal I am in allowing the best possible measurements (including possible shortcuts), I still cannot get under 120 msec. as the fastest possible reaction time.

But are you any sort of world class sprinter?

[Daisy]

Re: Am I missing something? (false starts)

I still cannot get under 120 msec. as the fastest possible reaction time.

But are you any sort of world class sprinter?

I think he's talking about theoretical reaction times, not his own.

[Pego]

Re: Am I missing something? (false starts)

I calculated those evoked responses through multiple synapses numerous times and no matter how liberal I am in allowing the best possible measurements (including possible shortcuts), I still cannot get under 120 msec. as the fastest possible reaction time.

But are you any sort of world class sprinter?

I thought the laws of neurophysiology applied to everybody. What baffles me is why do you have to be a sprinter to be able to calculate the minimum amount of time for a stimulus to travel through a few synapses.