Updated Study on Super-Shoes (Now Includes 2022 and 2023 Data)

I'll give it a read but you should have provided a summary of the findings to save other people time.

"This analysis shows that elite men have shown little improvement in race times beyond prevailing trends since the adoption of super-shoe technology in 2017, with a possible exception in the Road 10K event. For elite women, however, there has been a noticeable drop in times in the Road 10K and significant improvements in some other events. Despite these impressive results for women, the data does not uniformly indicate a pattern of decreased times correlating with the gradual adoption of super-shoes over the past five years."

Why a 2% reduction in times? They seems way, way beyond any claims of super shoe benefits.

What happened to the 4% improvement in efficiency? That doesn't equate to 2% time savings from 28 to 122 minutes.

Thank you. Before seeing the updated graphs, I was expecting to see a reversion back to existing trends in both men’s and women’s times, but the women’s times are so strong, and the men’s times so predictable, I don’t know quite what to think. And what is it about the Road 10K race that has allowed such a drop in times by both men and women?

How did you correct for the effect of drug testing in Kenya? If they could still dope like 2001, how much faster would the time be? You can clearly see how drug testing broke the improvement curve of the early 00s when you had that long period of people running slower. Then as soon as the shoes were released, those doped up times got crushed.

Plot up marathon times from 1985-2023 and this trend becomes obvious. 1985 100th time was 2:14:33. 1995 2:14:42. Notice how we didn’t drop 4 mins off the time in a decade?

Yes, way too many other variables for my taste. Too many things affecting the overly broad sample size to draw a conclusion about just one aspect. Better to use a more defined sample that hasn't been affected by other factors too much. High school boys or girls 1600m or 3200m times for instance. At least you can largely avoid the biggest problem, which is at the pro level, things are disrupted by whoever has the advantage in PED vs PED testing at that point in time.

Oh, sorry. I see this is about super shoes, so long distance road events are going to have to be your sample size. Also sorry I don't have any good suggestions to make it much better. That's a hard problem.

Why would you look at track races for road shoes? To assume young runners are not doping is foolish in the extreme or the effort of puberty on performance

unless you are incorporating every major trend in racing, and finding a way to prove the athletes are comparable year to year, you are on a quest for correlations to speculate causation about.

It is just as useful, and much easier, to examine world record progressions for those correlations. Not so much useful to prove something, as all those hordes of unmeasurable factors are still there. But useful to check whether you're wasting your time trying to find a correlation that just isn't there. And in world records, it's not. A few events have seen big improvements, but others are right where they were 10, 20 years ago. 3:26.00. 7:20. Others have improved just barely 12:35, 26:11.

Typically, when an event like women's 10k suddenly improves, it's because either a) some outlying performer appeared or b) the elite women in general started focusing on it. We have seen b) for years in HM with both men and women. Improvements in men's 5000 and 10000 and steeple have all been due to a single outlying performer, with everyone else either at about the same level as before (5000/10000) or worse (steeple).

Men's 1500 is still one of track's most hotly contested events. If super spikes worked, why are the current crop slower than during the normal-spike Kiprop era? Jakob is barely faster than Mo Farah.

Don't naively say doping has stopped, or slowed down, or whatever. If these shoes have a singificant effect, that effect should logically be visible in every event. Far from it; the correlation ain't there. Indications are that whatever is boosting performances in some events is specific to those events, not a general improvement.

Keith, nice work on collecting the data and coding it up, respect! This is a slightly more technical observation, but did you consider using a time series model instead of a trend line to forecast from the 2016 period and comparing that to the actual results and your 2% impact?

I don't disagree with your post, but this is just stupid. 3:28.81 vs 3:27:14 is miles apart.

I like your method and logic here and the final result or conclusion you have is what my personal estimate of the "true benefit" of these new products have been.
It's indisputable that especially on the roads, these shoes have had a great impact, but definitely to the degree some people have chosen to believe (which is "carbon fiber plates have turned bad runners into great runners and they don't even have to try anymore"). The reality is as you included here, there are just so many combinations of factors at play that aren't quantifiable and they work with each other to varying degrees dependent on the runners individual attributes.

Plates (and actual CF plates - not these plates which aren't even CF as most people believe CF plates to be) have been around since the lates 90's and so has the foam - it just wasn't seemed performance beneficial for running (as firmer foams that could be molded thinner were seen as the standard due to their better ground response and force transference properties). But the real genius was the marketing and the story - launching these products with a transformative athlete like Kipchoge and under conditions such as a perfect running surface, perfectly even pacing (a car), and other benefits like drafting (humans and cars) which all contributed but played second fiddle to the story of the shoes. Runners, who due to the lonely and difficult nature of the sport especially when it comes to peak performance, will just lap up and buy into anything legal they believe will make them better (we don't have any other implements and equipment to help us like many other sports do) and these products were perfect. And yeah, they looked different, felt different and were better. The power of putting on the one "implement" you have and truly believing it's going to make you better can simply not be overlooked and worked in tandem with the performance benefits of the shoes themselves.

If there is any downside it's that the industry has probably tapped out this cycle of evolution. If they could turn the current spoon shaped plates into leaf-spring geometries (like a prosthetic blade) and true springs that would be a move-on but that seems impossible given rules around the sport. Hard to see where we go with foams too - at some point the foams would be so light and compliant that you start seeing really bad ground force transference and instability which would detract from performance. Adidas were able to use their now famous "milled" midsole shoe which removed compression molding from the creation process which creates density and squashes to open cells of the foam - on the flipside it's a process that is incredibly wasteful, expensive and not durable at all. Making stack heights larger is blocked by WA rules but kind of irrelevant anyway - once we start getting over 45mm the basic stability of the shoe starts becoming questionable and many Nike elite athletes never really liked the Alphafly (and still preferred the Vaporfly) for that reason.

I can't wait to see some reports/data on spikes. We are here almost 3 years on from the natural adoption of the foams into track product and unlike road racing product there have been absolutely no claims from any brand about the quantifiable benefits of the "new" product. My guess is that simply there aren't any that a number can be attached to. The product isn't really any different weight wise and there just isn't enough foam in them (no matter what the foam is) to get the same compression and resiliency benefits of the road product (which is 80% of the benefit driver). I do believe the track products are better and help, but I think it is on a highly nuanced level that includes things like mental/emotional placebo, fatigue reduction due the mitigation of impact vibrations to lower leg muscles and the softer nature of the foams "evening out" slight imperfections in gaits that create more balance (real and perceived) for runners (what they would call "flow"). All of these are impossible to put a number on in the lab and definitely don't add up to the numbers we see on the road shoes.

I really wish that especially for elite runners (pros) that the intro of these products hadn't come along at the exact intro of wavelight, as in I wish they had been staggered. I wish we had seen 2-3 seasons of running in just the new spikes before we saw little lights moving around a track at perfect uniform speed that the best athletes know is right below their anaerobic threshold paces. If we hypothesize that the spikes work best the longer the track distance (which seems true) then same can also be said for the impact of wavelight and we only need to look at the splits of almost every WR set from 2000m onwards over the last 4 years and the absurd uniformity of the lap splits compared to previous eras to understand the power of that. Point being it's going to be so hard to look at times, especially at the pinnacle of current human performance, and deduce any conclusions that can be attributed to any other thing vs some weighted combination of them all that can never truly be known.

But nice work and looking forwards to anything else you have!

Doping hasn’t stopped. The type of doping you can get away with has changed. There is a reason why records like the woman’s 400/800 have lasted 40 years. There is also a reason why the 5k slowed down by 10s in the 2010s after dropping 20s over the previous 15 years.

And finally not that following the trend line doesn’t mean the shoes don’t work. It just means that the shoes were why we got the improvement.

Wait, isn’t a 2% reduction in time huge? Am I calculating incorrectly?

4min mile = 240 seconds -2% 3:55.2

Thats crazy!

sorry but your analysis is garbage - that's as nice as I can put it.

Open those things on the front of your face - they are called eye.

Have you noticed times getting much faster, world records getting - not just beaten - but obliterated.

You can dress it up all you want - adds chart - and science (if that is what you call it) but you are missing the simple reality. You want so much to dispel the idea that super shoes are faster that you went into this believing in the halo effect (you are not neutral). Therefore, your entire analysis is biased and useless.

the problem with comparing mass times is there are so many potential confounding factors that there is no way you can account for them.

the simplist study is get two pairs of shoes. You a set distance using each pair. Which is faster. Then have someone else try it, etc. Everyone I ever talk to who has ran in supershoes says they are faster in them - which is why they continue to buy them and race in them and no one is going back to their old shoes.

but according to your flawed analysis there are all wrong = all of them 100% are using shoes that cost a lot and offer no 0% advantage, really?

world records got obliterated, not just beaten, long before super shoes arrived.

If there are so many confounding factors, why do you assume the results are due to the shoes? One study on a small sample? How would your n of 1 study be better? It could easily suffer from a placebo effect.

You're right the analysis doesn't establish causation, but it seems unlikely to me that runners (and coaches) who were lowering times with PEDs would have suddenly decided to quit using such substances just as super-shoes were rolled out. More likely they would have continued PED use and try for an additional 2% performance with the shoes. For the men's times at least, I don't see it, but I would love to see the numbers of athletes who were tested/caught lined up against their country's times to see if there is any correlation.

People are rightly concerned about the comparability of records (not only world records, but country and regional records, too), which is why focussing on the top-100 worldwide (which contains most of those marks) is the most helpful metric in my opinion. Yes, high school records are meaningful too, but not much beyond their local audiences and don't threaten the sport broadly.

Uhh, I'm pretty sure at no point his work suggests there is no performance benefit. I'm also quite sure at no point he says there is "0% advantage" - in fact other values are given. So simple question here - "What the f--k garbage are you talking?"

The question isn't are "supershoes faster" - it's "to what degree" - and ironically you kind of summed up the essence of the work when you said, quote, "the problem with comparing mass times is there are so many potential confounding factors that there is no way you can account for them.", but I guess that went right over your head.

You even semi contradict what you write by then saying the way we just test them is "set a distance using each pair (of SS and non SS). Which is faster". Which is "faster"? That's your test? That's your objective measure - which is "faster? Haha, good one.

I don't know Keith Moulton, I have no reason other to defend him other than I find it unconscionable that guys like you come in with the flamethrower choosing violence and put utter shi-te on the table as your basis. So if it's not the reasonable numbers put on the table in the study and it's something markedly more - why don't you all enlighten us as to what they are and your level of performance footwear expertise?