Nike claims that in a pre-race test, another Nike pro ran 3% faster in mile with pacers and speedsuit. So what happened today?

How many male pacers?, how much money?, how much hype?

Let’s get back to reality and get on with the summer season without this circus distraction.

Are people purposefully daft? No respected scientists would claim reliable results from a sample size of one. There are infinite things wrong with the test.

Maybe the 3D printer was running low on cartridge.

I just watched without following any commentary from this forum, and my impression is that she didn't get much of a draft. She was pretty far behind runners ahead of her and it looked like a lot of people spread out around her rather than a tighter pack. Maybe that was from aggressive pacing relative to her fitness.

I said this in another thread, Kipchoge got roughly 3.4 seconds per mile advantage in breaking 2 with all the changes. There was just no way she was getting more than 4 seconds advantage at slightly faster speeds. He had a net drop, pacers, bottle hand ups, and car creating a huge draft.

Perfect day in the same shape as 2023 she might go 4:04. The pace was just too hot for her current fitness.

Was that other runner trying to run a 8 second PB? Big difference between trying to run sub-4 and a controlled time trial. Kipyegon might have run 4:10 if not for the pacers/speedsuit.

The other pro may not have already had as ideally rabbited a race as Faith to run his best time. He may have been off on day 1. It may have been hot or cold, etc. Most of these tests that claim x product led to y percent improvement in trials are bunk anyways. Nike had it as a foregone conclusion that their new gear improves times, so the only tests they were going to run were going to prove that.

Today was an exhibition created a nike marketing vessel. This study could be viewed as just another piece of marketing.

How many times did they repeat that experiment? Here's the issue: you run a 1000m time trial today, you run 2:32. You do the same thing three days from now, 2:30. A few more days, 2:36. A few more days, 2:32 again. Unless they repeated the time trial with AND without the speedsuit + pacers multiple times each, you have no idea how much of the difference between two time trials is from your clever intervention and how much of it is from the natural day to day random variation in performance.

Incidentally you have the same issue with super shoes. Ironically, one of the authors on the "Faith can break 4:00" paper has another paper about this exact issue.

link.springer.com

The Fallacy of Single Trials: The Need for Multiple Trials in Assessing Running Economy Responses in Advanced Footwear Technology - Sports Medicine

In the quest to uncover the underlying mechanisms responsible for the performance-enhancing benefits imparted by advanced footwear technology (AFT), footwear researchers are employing an individual-level approach. In doing so...

Open link

Yeah it's flawed all the way around. To be confident in 3% gain they would need dozens of tests and among various ability levels. Otherwise it can be a case of the woman in the experiment knowing they wanted her to be faster in the suited situation and therefore being in better shape and giving greater effort.

It can also be a situation where a 4:30 miler receives 3% gain but a 4:07 miler does not. There's only so fast she can go.

However, Nike was pressed for time. They had one Faith Kipyegon and she's not 22. I always factor situational influence first and foremost. They had to have confidence in their theory and take a rushed swing.

Regardless, this thread is extremely valuable because the point is valid. I'm sure many at Nike are frustrated, and rightfully so. Everyone crowing on this site and elsewhere about predicting 4:05 or 4:06 had brutal logic but ended up being correct. In sports betting we call that winning with the wrong side. Those guys eventually leave town broke on a bus.

If this experiment were repeated time and again the benefit would be considerably greater than we saw today.

I need more details but who is the athlete? Which time trial was first? And were either efforts PRs?

I wish they had just shared any of this info beforehand. But I guess they didn't want to publicly acknowledge this was really how effective was the new drafting system.

You had me until the very last line.

What makes a PB so special is that it combines all the factors, wind, weather, fitness, competition, pace, etc and is, by definition, your personally best race ever. Particularly for a runner like Kipyegon who has run dozens and dozens of timed 1500s. She's never gone under 3:49 in the 1500, which isn't sub-4 (or even sub 4:05) pace.

People consider all the benefits that an experiment like this might bring, like pacing or fueling or shoes, and those combined I agree probably would beat your PB on aggregate. However, the issue with these single event trials is that you can't in fact control these factors. You can't control the temperature, allegedly it was a bit hotter than preferred today, you can't control the wind, the humidity, you can try to impact fitness and mental state but you can't guarantee that your athlete will be at their peak fitness either. While it's possible your trial event has the best possible conditions, it's highly statistically unlikely.

The other thing that is flawed about this experiment is that this runner who saw 3% gains, presumably ran on the same track. Possibly even on the same day with virtually the same weather. Likely close in time, so her fitness would be the same. It doesn't compare the time to the same runner's PB in the distance, which would likely cut into the 3% gain significantly.

If Kipyegon ran dozens of time trials over the span of say one year, eventually she would get a weather day and fitness that was as favorable as her 3:49, and then maybe you could expect something approximating a 3% gain in time on top of that, although I agree that a 4:30 runner likely has more room for growth than a 4:07 runner so maybe slightly less than 3%. But that kind of experiment would be A. very expensive and B. very time consuming for the athlete, who might be interested in running...any other race that year.

If you just run it back just one more time, there's no guarantee that her fitness is better, the weather is better, other conditions are better, etc.

This is it. With a mile, it's feasible to run the trial dozens of times across several athletes over a 2-4 week period. You can then average the gain/loss over a baseline. But the random variance issue is massive. If Kipyegon ran the TT again tomorrow, she could go 4:04 or 4:10. I think it's underappreciated that the only reason that Jakob isn't the mile/1500 world record holder is random variance. The difference between 3:26.00 and 3:26.73 isn't fitness

I reckon it's way simpler - it's just a case of someone parroting a Nike line without interrogating or communicating any logical challenge or issue with it. That's aside from the single trial comparison, the normal variation in performance, a placebo effect, and the fact that this athlete was highly unlikely to have been running as close to their own ceiling as Kipyegon would have been.

Insufficient sample size to make any conclusions

Right. Kipyegon was really part of the experiment. Listen to the biomechanist on this week's pod: every single foot strike is a variable. There's so many things that can vary in a running TT, the level of uncertainty is very high and even having the GOAT can only reduce that uncertainty a little.

Human factor? You can be in the best shape of your life and reach 99% probability of racing your fastest race ever (which she did btw) and having your best race but it's never going to be 100%. If you line her up 10 more times under these exact same conditions does she run it faster? Possibly. I for one hope they do it again to find out.

How about she just had an off day as one explanation? She is human (despite her remarkable athletic abilities.)

Back to the drawing board. In science there is no such thing as a failure since one can learn from any attempt,

The whole thing was dumb, why were there 25 pacers some being 30m behind her?

Why run in a 90% empty massive stadium on a Thursday in humid France?

I think she can do 4:03, 3 male pacer, small stadium, let the fans down to lane 3, DJ, and more realistic pacing.

Go out a little slower, and save a close coming home.