Super spikes are worth 5 secs per mile

rekrunner wrote:

Armstronglivs wrote:

- A study of 10 runners hardly enables definitive findings when there are thousands of athletes in the sport. That is too small a sample.

- The research points to gains but again they have to be approximate when the findings are based on very small research samples. To use historical comparisons of times in the sport is further unreliable because there are a number of factors which could have produced a surge in performances across the board of which doping must be included as the practice continues unabated and mostly undetected (Howman).

- Posters, including the OP, have suggested a gain of several seconds over the 1500/mile, based on samples of athletes they have coached. These athletes aren't necessarily all highly talented - some won't be - but some will be talented and none will be hobby joggers. Whatever their level they all apparently gained by the same amount, of about 5 secs in a mile. If that is what the shoes enable a sizeable pool of trained athletes then they must enable a similar gain to the very best, as the gain as solely attributed to the shoes and not training or anything else. The difference in efficiency provided by a shoe to a 3:45 miler and a 4:10 miler will be very small. The difference in their levels will largely be based on physical and cardiovascular differences. The shoes don't change those.

10 runners is enough to build an economy curve over various speeds to show it is not linear. It wasn't used to estimate race time improvements of any athlete, but to answer your narrow question by helping show that the speed improvement from economy improvement diminishes at faster speeds, as oxygen costs increase at a faster rate.

You didn't seem convinced by the other study (linked to by "zzzz") which looked at the top-100 athletes per event (800m, 1500m, 5000m, 10000m for men and women). How big does the sample have to be?

A sample size of 10 runner is 10 more runners than your proposed model of "they must enable a similar gain to the very best". There is nothing (and by "nothing" I really mean nothing), that supports such a presumption.

Other posters did suggest other models, like a fixed "5 seconds per mile" model, but the point of pointing to these studies is to show that there are superior models based on more controlled observations of many hundreds if not thousands of athletes that the biggest economy gains are observed at sub-maximal speeds of 14-18 km/h, and the rate of time improvements declines at faster speeds. It is not a fixed rate per mile, nor a fixed percentage.

rekrunner wrote:

Armstronglivs wrote:

A University of Michigan study that also concludes a 1-1.5% increase in performance from the spikes. That would put El G, Lagat, and Ngeny in the 3:24 ball park for the 1500 and around 3:41 for the mile. Yet no runners in superspikes today are anywhere near that.

----------------

'They tested a small group of men in both types of spikes. Compared to traditional spikes, runners in super spikes enjoyed about a 2% increase in running economy, which could translate into a 1%-1.5% improvement in times, said co-author Geoff Burns, U-M adjunct assistant professor of kinesiology and sport physiologist with the U.S. Olympic and Paralympic Committee.'

'For a runner who completes 10,000 meters in 30 minutes, a 2% increase in running economy would result in a time improvement of about 25 seconds, the study shows. The 10,000-meter race is the longest track run at around 6 miles. The female world record for the outdoor 10,000-meter is 28:54 and the male record is 26:11.

"A 1.5%-2.0% increase in economy in an elite runner could be the difference between contending for a medal and not even qualifying for the (Olympic) Games," Burns said. "For non-elite runners, like high school and NCAA runners, this is still a profound difference that could open doors for competitions and opportunities they might not otherwise have."

Link:news.umich.edu/research-suggests-contr...

LOL -- Sample size of 9 male runners, running 16 km/h -- smack dab in the middle of that sweet spot of 14-18 km/h.

El G, Lagat, and Ngeny run 26 km/h.

rekrunner wrote:

liar liar wrote:

Here’s you from a few pages ago: “DIdn't we acknowledge Nick Willis at the beginning of the thread, about 2 1/2 weeks ago? He himself said superspikes would give him 2-3s in the 1500m/mile, which would make him a 3:27-3:28 guy.”

Let’s ignore your terrible arithmetic for the time being (2-3 seconds faster than 3:29 would be 3:26-3:27). You’re scoffing at the notion of El G going from 3:26 to 3:24, but are perfectly fine with Willis going from 3:29 to 3:27, even though both are running far faster than your magical 16 kph.

You and Armstronglivs are both awful, but at least he’s correct about super spikes.

I find it better math to round figures than truncate them as it produces smaller errors. Nick Willis ran 3:29.66. A 2-3 second improvement makes him a 3:26.66-3:27.66 guy, or 3:27-3:28 when rounded.

But good job ignoring it.

Furthermore, I clarified that when Nick Willis said that in Jan. 2021, he wasn't running 3:29, but 3:37 (using your math), which would make him a 3:34-3:35 guy, which he did run later that year.

In both cases, I am not scoffing at anything, but sharing Nick Willis' stated opinion after a 1200m time trial, and the results of several studies showing where the largest economy gains are, and how they decline at faster speeds.

I think El G could have gained a second. Within a margin of error, he could have been a 3:24 guy (your math).

Armstronglivs is right about what exactly? We both seem to agree that 5 seconds is wrong for El G, but disagree that it could be right for high school milers. I'm following the data. What is he following?

Armstronglivs is applying the wrong improvement model to El G, Ngeny, Lagat, etc.

Armstronglivs is wrong to think that elite athletes should gain a similar amount as NCAA and high school milers.

Armstronglivs is wrong to repeatedly ignore all these numerous studies and replace it with his own "gains should be similar" model, based on a sample size of 0, only to argue that his model leads to ridiculous consequences.