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

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.

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.

2022 WC Results I see (hopefully formatting holds)

avg max min
w 100 m heats 0.141387755 0.229 0.110
w 100 m semi 0.146136364 0.194 0.119
w 100 m final 0.136125 0.150 0.116
110 h heats 0.138631579 0.198 0.114
110 h semi 0.130333333 0.154 0.101
110 h final 0.121571429 0.145 0.099
m 100 m heats 0.131625 0.191 0.104
m 100 m semi 0.131043478 0.146 0.108
m 100 m final 0.1295 0.155 0.104

I hate to break it to you, but no "stats guy" is going to do this analysis in Excel. There are many statistical software packages that could be used to explore this with legitimate tests to compare Eugene vs other championship, and I hope this gets taken on by one of Letsrun's smart guys. Or women - come on, it's 2022!

And if the starter has a predictable (and short) cadence, more may be willing to try "guessing" it here.

Yeah! You tell 'em! It's the current year!

Gets really interesting when you add Tokyo 2021 and London 2017

London - Avg. 154.9

Tokyo - Avg. 143.6

The half ms difference between Doha and London is what SHOULD happen.

looking at things another way for 2022 WC - interesting no Women

# < 0.11
w 100 m heats 0
w 100 m semi 0
w 100 m final 0
110 h heats 0
110 h semi 1
110 h final 3
m 100 m heats 4
m 100 m semi 1
m 100 m final 1

Do you really think excel can't run a T test...? What super advanced statistical test do you think is so much more legitimate?

Interesting thought? Do you think this we'd see RTs improve through the rounds if this were the case? It seems the RTs were virtually identical in the heats vs semis/finals. I'm not a sprinter so correct me if I'm wrong but if the starter was abnormally quick I think we'd see a difference between the first experience with it and repeated exposures. Thoughts?

Comparing men to womens times for anything is always dumb. Compare like for like.

In 2019 no women under .12 in the 100 m heats. 5 under it this year. And if you look at the number under .13 or .14 from 2019 and 2022 the difference is even clearer.

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)

is the 0.1 standard the same for both sexes?

So now we have an explanation for why there are competing studies giving support for either 0.08 or 0.10 seconds being appropriate limits for reaction time: It depends on the actual set of blocks used in the studies because of differences in sensors used, construction, etc. They might need to do a full statistical study on each set of blocks/reaction timing system to calibrate a different reaction time limit for each specific design. Or, more likely, adjust the sensitivity of the sensors or whatever to equalize the results to the set of blocks used in the study that came up with 0.10 - some kind of certification that different sets of blocks will be the same. In any case, World Athletics screwed up.

Didn't two women get DQ from the Caribbean? On in 100M and one in hurdles?

Would be interesting to see if any of them were in the same lane or if the same equipment was used in these lanes when they set up the races.

Would be good to have a proper statistical analysis done. Someone trained in statistics.

Is the data from the US Champs in Eugene last month available? Presumably, they would be using the same equipment as they are using this week.

It was mentioned in the podcast that one of the DQs was also in lane 3, Allen's lane. But I don't know if the same blocks are used for the same lanes. Allen also seemed to say to the officials that there was a lot of crowd noise at the start, so if that's a contributing factor you would expect there to be more false starts in lanes 7,8,9 than 1,2,3 over a large enough dataset (closer to crowd)