I need and will pay a stats guy to compare reaction times of Tokyo 2021 vs Eugene 2022.

Didn't mean to quote that post

sprinter guy wrote:

Precious Roy wrote:

Better yet, get someone to go out and test the equipment. There is no way that equipment can be 100% accurate up to one one thousandth of a second.

I don't think anyone's going to allow that, because this already has lawsuit written all over it. Anyone involved with meet administration will go into deny, deny, deny mode either way.

Yes, it looks like they are open to a legal challenge. I can hear the lawyers now:

The meet officials had clear evidence something was wrong with the timing--all those DQs that were up near .1 were quite unusual.

Given this, the fair, just, reasonable thing to do is to let those athletes who were DQd (with times near .1) run in their event anyway. This would prevent the meet from committing an incurable injustice.

Instead, the meet officials simply shrugged their shoulders and sent athletes home. They were culpably negligent.

So, ladies and gentlemen of the jury, Devon Allen's lost income from this is in the millions, and it should come out of the hide of the folks in charge of the meet....

You had just one job

LTR5K wrote:

Are they using the same equipment as in the NCAA's??

NO. This is Seiko; they've had this gig since WC in Helsinki 1983. NCAA & USATF were done by Flash Results of the USA. Pre Classic was done by Omega (Swiss Timing). Pac-12 was done by Athletic Timing of Portland, OR.

Bra-ket wrote:

I will pool together each gender's 100, 200, and 110h reaction times (I think these should all draw from similar distributions) and see if this trend holds. Rescaling by the mean values is probably a good way to see if we are in fact drawing from different distributions.

https://imgur.com/a/jahGtUV

This sounds sensible. Depending on how much trouble it is, you might go the other way and look at each race & heat to see whether it makes sense to pool them. (If there are enough people who ran at both Tokyo & Eugene, you could compare their numbers to see if any systematic differences show up.)

It would be useful to look to see if there are any outliers that are driving the differences in means.

For the non-stats people: From the numbers we've seen so far, it sure looks like there's something different about Eugene. But, as we all know, "appearances can be decieving." Their statistician's job is to look at those numbers more carefully and systematically, to make sure they really say what they seem to be saying.

Statisticians are data-sheriffs, enforcing Feynman's law:

1) Don't fool yourself;

2) You are the easiest person to fool.

rojo, you claim that you will pay someone to do this work for you. How much are you willing to pay?

why do you need a stats guy? wrote:

Rojo, I don't think you know what "stats" is. Why do you need a stats guy to put lists of numbers into a spreadsheet? YOU could do that (if you weren't so lazy)

Here's the deal - rojo wrote that he had just woken up, and his next hourly nap was fast approaching. When is he going to fit in all this work?

Surprised that nobody has put down the figures with a confidence interval (99% looks to the point) yet...

However it s pretty clear that something is off, what exactly is less clear as it could be a lot of things.

I just want to point one thing though. An important one. To the eye in slow motion it is clear he doesn't jump the gun. That should be all that matters. But we put ourselves in the hand of a tool we made and became a slave of. It should help us and not decide for us. And still the rule is to kneel to it. There is no leeway and everybody thinks it s fine until something happens. Isn't that crazy?

Just Thinking wrote:

LTR5K wrote:

Are they using the same equipment as in the NCAA's??

NO. This is Seiko; they've had this gig since WC in Helsinki 1983. NCAA & USATF were done by Flash Results of the USA. Pre Classic was done by Omega (Swiss Timing). Pac-12 was done by Athletic Timing of Portland, OR.

Do the Seiko systems require calibration or are they completely hardwired? If they need to be calibrated, maybe somebody screwed that up. It just looks like the system was off by about -.02

You are kinda in the world beyond t tests. Multiple studies have show <.1 reaction time is possible.

I believe this problem has more to do with the very small sample sizes these reaction time standards are built on. Statistically, Allen might be an outlier with one sort of calculation approach but not another, but only in the world where we forget about 5 standard deviations.

An important issue here is the change in reaction times through rounds to finals; gender effect can't be ignored either.

I am concerned too that some of this might be due to the carbon plates, which lower race times and may also alter mean reaction time readings, especially a concern with males.

rojo wrote:

If you are a stats guy, I need your help. I'll pay you to compile a spreadsheet comparing the reaction times in the men's 100, women's 100, men's 110h from Eugene to Tokyo from last year.

Should we include the preliminary rounds of the men's 100? I say no. Takes less time and those guys are largely amateurs.

I'd do it but just woke up and am headed to do some women's interviews for the marathon. In the 5 minutes I spent googling, I can't even find the reaction times for Tokyo. Anyone know where they are to help get us started?

IF you are interested, give me a call or shoot me a text 410-929-3368.

What you need is Coach and Star man Timo Mostert of AFXC!

stat man, though Timo is known as a star in coaching circles. haha

Good way of getting all the free stats, by posting that you will pay for it.

My best guess is that they’ve adjusted the force threshold down for these championships which would lower everyone’s reaction time. Here’s the best explanation: (from https://www.basvanhooren.com/is-it-possible-to-react-faster-than-100-ms-in-a-sprint-start/)

Each competition can use a different system and each system can use a different method to determine whether a false start has taken place [3]. A Seiko system used in the past for example used force sensors in which a 20 kg limit had to be exceeded within 100 ms to trigger a false start. There is also a Seiko system that measures the speed at which the force curve rises to determine a false start. The Lynx system uses an accelerometer at the rear of the starting blocks and uses an unpublished threshold to determine the initiation of the start. These differences in determining when the start actually takes place can lead to different total reaction times. Depending on the specifications of the system, the same reaction time of an athlete (step 1-5 Figure 2) can cause differences in the total reaction time (step 1-6 in Figure 2) and thus, in one system, cause a false start (total reaction time <100 ms), while another system would give a correct start.

Anyone know if this is the same equipment Hayward field uses for all its track meets? Did they have to change equipment to host these world championships, or is it the same equipment they used a few weeks ago for USTAF Championships?

I'm betting they've been using whatever system they have at Hayward for many years, and no one has complained it was malfunctioning until Allen got DQed.

The Dreaded Pile Driver wrote:

Anyone know if this is the same equipment Hayward field uses for all its track meets? Did they have to change equipment to host these world championships, or is it the same equipment they used a few weeks ago for USTAF Championships?

I'm betting they've been using whatever system they have at Hayward for many years, and no one has complained it was malfunctioning until Allen got DQed.

You loose your bet. It's not the same equipment they used for USATF Championships. For more details read the posts above yours.

And you think they did a $270 million dollar renovation but kept all the same equipment that they've used for "many years"? Where do you think they kept the starting blocks? In Phil Knight's garage?

set bang wrote:

sprinter guy wrote:

I had two backup theories that just don't seem likely at all:

1) Athletes suddenly started training reaction time way harder, especially in the last month or so, and were collectively significantly successful in this endeavor. Not likely.

2) Possible PEDs, whether legal or illegal, improving what is possible on the average. Even less likely. Because, everyone would have had to be on board and it would have had to happen all at once, because there's likely a pattern and it is not just one athlete.

What no one seems to be talking about is how quick from set to fire the starter has been in Eugene. It is possible since the hold is consistently much shorter that the reaction times are influenced.

This is an excellent point that no one is talking about!!!! The hold times are so small, that Shelly Ane Frazier Price never even is actually set, it almost looks like a continuous motion from booty raise to getting out of the blocks

Surely this would help improve the reaction time.

Median reaction time for all block start events (100 m, 200 m, 100/110 m hurdles) for World Championship meets:

0.155 men iaaf-world-championships-london-2017-7093740
0.151 men 15th-iaaf-world-championships-7078726
0.161 men 14th-iaaf-world-championships-7003368
0.155 men 8th-iaaf-world-championships-6947294
0.156 men iaaf-world-athletics-championships-doha-2019-7125365
0.157 men 9th-iaaf-world-championships-in-athletics-6930156
0.160 men 11th-iaaf-world-championships-in-athletics-6903480
0.144 men 7th-iaaf-world-championships-in-athletics-6939522
0.152 men 12th-iaaf-world-championships-in-athletics-6998524
0.152 men 10th-iaaf-world-championships-in-athletics-6937596
0.170 men 13th-iaaf-world-championships-in-athletics-7003367
0.133 men oregon22

0.160 women iaaf-world-championships-london-2017-7093740
0.162 women 15th-iaaf-world-championships-7078726
0.171 women 14th-iaaf-world-championships-7003368
0.157 women 8th-iaaf-world-championships-6947294
0.155 women iaaf-world-athletics-championships-doha-2019-7125365
0.162 women 9th-iaaf-world-championships-in-athletics-6930156
0.158 women 11th-iaaf-world-championships-in-athletics-6903480
0.149 women 7th-iaaf-world-championships-in-athletics-6939522
0.158 women 12th-iaaf-world-championships-in-athletics-6998524
0.158 women 10th-iaaf-world-championships-in-athletics-6937596
0.171 women 13th-iaaf-world-championships-in-athletics-7003367
0.140 women oregon22

Oregon does seem to be fastest and it's not just Allen.