The shoes do make you faster but your attempt to prove it has a lot of flaws.
The shoes do make you faster but your attempt to prove it has a lot of flaws.
Doesn't have to be a senior thesis, you don't have to be a grad student to copy and paste some performance lists from tfrrs into an excell spreadsheet. Sure it takes longer than 5 minutes but with a few hours you could put together a statistically significant amount of data and process it. You could do it yourself or if you're too busy pay a high school kid $20 to do it. Also saying that it passes the smell test doesn't make any sense. To my understanding, you already did the smell test before when you said it seems like times are faster. You then posted an article saying there was data supporting your intuition, but then the data was incredibly lacking. Now you're saying that 50 data points that aren't significant was the smell test. If it was just a smell test, why would you title the article in a way that suggests that you found data to prove that the shoes work? Very confusing argument. As for finding out who was wearing them, not very difficult to figure out as many races have photos and all relevant nike school runners are wearing the shoes.
rojo wrote:
113 wrote:
Or at the very least, maybe note that the women's 10k is 1.46% slower and the men's 10k is only .54% faster. Those events should show the greatest improvement. Instead, it looks an awful lot like noise.
I did note it. I can't help it that you didn't actually read the article.
So far we have a bunch of criticisms from people who haven't even read the article.
Lol. Tough crowd.
Are you only using the super shoe runners or everyone in the field?
Ps- I hope Uncle Phil send you a pair.
The data ARE in. Rojo, data = plural.
bugboy98 wrote:Now you're saying that 50 data points that aren't significant was the smell test. If it was just a smell test, why would you title the article in a way that suggests that you found data to prove that the shoes work? Very confusing argument.
Actually, Rojo didn’t even use 50 data points [per distance]. He used one - the 50th fastest time. We aren’t told if the 50th fastest runner, or anyone above him or her, wore supershoes.
Don’t get me wrong - I actually think it’s likely the shoes are helping runners achieve faster times. But the numbers cited in no way indicate that.
My gosh, Rojo just keeps finding new lows.
See if you can hold two independent thoughts (that may seem conflicting) in your head at the same time:
1. Nike's new spikes are the best spikes that have ever been made for distance runners
and
2. They're a natural progression of technology, and do not give an advantage so large as to wiping out the record books.
By your constant criticisms, I can only assume that the only records you would consider are those on cinder tracks with leather shoes.
Fwiw, I think WAVELIGHT is a MUCH bigger advantage for races like the 5k/10k.
Finally, your "statistical proof" proves only one thing, that you don't understand statistics.
rojo wrote:
"But Rojo, you didn't interview the top 300 NCAA runners and figure out what shoes they are wearing this year verus last year and post it." Do you know how long that woult take?
One email, everyone bbc’d, quick google form.
So, maybe like one day? Half a day if you have a partner in this.
i have a data point:
centro
maybe don't put it behind a paywall?
bugboy98 wrote:
Whether or not the "super spikes" actually work, this argument has flaws. You fail to consider other reasons why times would be faster this year, and you only look at one other year for reference. The NCAA has more depth because of all the people who gained a redshirt year from covid, which could also contribute. Saying that because the 50th best person in the NCAA this year versus two years ago is faster in distance events. If you really want to prove that the shoes work, take a large sample of runners comparing their times to previous years and see if runners with the shoes improved more than those without them, and while you're at it, look up the phrase "statistically significant". This is lazy reporting, I'm not saying the shoes don't work, but with all the data that is available on runner's times year to year you could have looked at more than 50 per event.
and
bugboy98 wrote:
Doesn't have to be a senior thesis, you don't have to be a grad student to copy and paste some performance lists from tfrrs into an excell spreadsheet. Sure it takes longer than 5 minutes but with a few hours you could put together a statistically significant amount of data and process it. You could do it yourself or if you're too busy pay a high school kid $20 to do it. Also saying that it passes the smell test doesn't make any sense. To my understanding, you already did the smell test before when you said it seems like times are faster. You then posted an article saying there was data supporting your intuition, but then the data was incredibly lacking. Now you're saying that 50 data points that aren't significant was the smell test. If it was just a smell test, why would you title the article in a way that suggests that you found data to prove that the shoes work? Very confusing argument. As for finding out who was wearing them, not very difficult to figure out as many races have photos and all relevant nike school runners are wearing the shoes.
The only thing that is lazy - is your lazy posting.
Are you crazy? You seem to think it will take 2 hours - you suggest me paying someone $20 which is if I'm paying them minimum wage 2 hours - for someone to come up with like 50 people per event and figure out what shoes they were wearing this year and also in 2019? And compare the two?
It would take weeks if not months to do properly.
1) Look at the 2010-2019 archived lists on Tfrrs.
2) Take the 50th best time in each year in 1500 and up
3) Compare the spread for the n=10 * 8 events
4) Set a confidence interval based on 3) and see how 2021 stacks up
We all know you are right. All the non-Nike schools scrambling to blow dry the swoosh and paint Adidas, etc.
Would be interesting to see the trajectory of the 50th best since 2010. It seems like there is a lot more depth now.
More 5th and 6th years this year. Other than the Ivy League, who else isn't competing this year in Division I?
1-4 won't take you very long.
rojo wrote:
It would take weeks if not months to do properly.
This is preposterous. A high school stats student could do it with a t-test on his/her graphing calculator in an 8-hour work day. Nobody is expecting you to spend weeks/months to get a perfect sample. It just seems weird to claim that you found data to prove that the shoes work and not do any work to back it up. If you had sent a google form to 10 coaches from non nike schools to see if any of their athletes used the shoes and then looked at how those runners' times changed from 2019 to 2021 compared to runners from nike schools, you would actually have something to report. Instead you used data that shows nothing because it doesn't have any control group and not enough data points, and then you defend it like its your life's work.
maybe u should learn python
rojo wrote:
It would take weeks if not months to do properly.
Get to work
Correlation does not prove causality.
Rojo if you want to send me lists of times from those two years I can do a statistical test for you.
Also, doesn't this prove that in-person schooling leads to slower times? ;)
i'd probably guess something along these lines if you asked me how much faster the #50 mark would be after ~9 months of no racing and with 5 full classes of guys with eligibility rather than 4
David S wrote:
Rojo if you want to send me lists of times from those two years I can do a statistical test for you.
Also, doesn't this prove that in-person schooling leads to slower times? ;)
Or does it prove that in person schooling leads to slower stats students?
It is my understanding that the Vaporfly and other road super shoes are not banned from NCAA competition.