We see these windtunnel or aero sensor tests of pro triathletes claiming their testing and adjustments and new equipment has saved them 20+ watts; or pros switch from “slow†bikes to “fast†bikes. Can we point to an of these athletes objectively racing faster? The best cases to examine are those where the “30 watts saved†was determined between races that are less than two months apart so they we can roughly control for fitness. I’m skeptical, and might even say we could more easily find cases where pro’s get slower. We could pretty easily plug-in, race a and race B in best bike split to make this comparison.
Fresh out of the wind tunnel myself, but not a pro. How is this surprising that aerodynamics help? It’s the vast majority of effort on most time trials, and less drag = faster. I definitely will be faster coming out of the wind tunnel because it’s the only way to really test and see if things improve. The problem with aerodynamics is everything is quite complicated and interacting, so it can be hard to just “look” and see things being faster. I think you are right that a lot of marketing is exaggerated. Most frames aren’t really THAT much faster. But overall, if you change a couple of things, my fastest helmet was 7 watts faster, different aero socks were like 4 watts faster. Do some stuff with your position and so on, and you could get 30 watts. And that puts you, what, something like 2 minutes faster?
The question here is not whether drag is real and a thing, the question here is external validity.
Let’s put it this way, if we take several of Lionel Sanders races from last year, and we plug his results into best bike split, and then we add 30 W to his average power, do his most recent bike splits compare? I’m skeptical. yes, we can imagine that he is taking those watt saved and saving it for the run, but at some point somewhere, we should see his bike splits in the range of what we are seeing Magnus, Sam and Trevor doing. I am not seeing that.
We see these windtunnel or aero sensor tests of pro triathletes claiming their testing and adjustments and new equipment has saved them 20+ watts; or pros switch from “slow†bikes to “fast†bikes. Can we point to an of these athletes objectively racing faster? The best cases to examine are those where the “30 watts saved†was determined between races that are less than two months apart so they we can roughly control for fitness. I’m skeptical, and might even say we could more easily find cases where pro’s get slower. We could pretty easily plug-in, race a and race B in best bike split to make this comparison.
I believe the classic example was Crowie (Craig Alexander). He got tired of being outbiked at Kona. We could easily be looking at 20W here.
Yes, I am very familiar and acquainted with the testing and the differences in training that he underwent leading up to that race. Unfortunately, the change in his biking was more than two months apart, so we are unable to control for changes in fitness, and we would also have to account for severe weather changes around that period of time in his record-breaking performance in Kona that year compared to the immediate prior years.
Also fascinating: top guys nowadays are at ~0.200 CdA for long distance tri and a lot of that is their position on the bike - if wind tunnel engineers were using protocols that led to inductive reasoned outcomes, why weren’t positions better 10+ years ago? I still see ‘top’ bike fitters slamming people’s hips behind the bb (see the fit of the above poster in their separate thread with an effective 78deg sta) and lowering the front end to lower their cda without regard to power production, and going to shorter cranks without regard to glute vs quad utilization. Inductively derived results are a sign of good experimental design.
…oh, on re-read I get it a bit more. With even day-to-day physiological state changing, environmental conditions, and racing dynamics being so variable in an uncontrolled environment, I think you’d have a hard time getting data valid enough to tease out even 20W. Except maybe for pure 40K TTers in some cases. An easier metric might be performance vs. the same competition. Not just bike leg - the W savings might have been banked for the run, but overall.
Sure, I’d presume that social media claims exaggerate actual in many cases. And for a variety of reasons.
…oh, on re-read I get it a bit more. With even day-to-day physiological state changing, environmental conditions, and racing dynamics being so variable in an uncontrolled environment, I think you’d have a hard time getting data valid enough to tease out even 20W. Except maybe for pure 40K TTers in some cases. An easier metric might be performance vs. the same competition. Not just bike leg - the W savings might have been banked for the run, but overall.
Sure, I’d presume that social media claims exaggerate actual in many cases. And for a variety of reasons.
A 20w decrease in drag over 56 miles would be fairly easy to tease out (a lot gets averaged out in time series data) and you could use the athlete’s swim and run splits and compare them to the same competitors race to race as rough gauges for confounding variables such as fitness, weather and nutrition from one race, to windtunnel, to second race. We can make excuse after excuse about the impossibility of measurement (it’s not impossible to draw inferences) but then we also can’t demonstrate external validity for these massive watt savings.
A 20w decrease in drag over 56 miles would be fairly easy to tease out (a lot gets averaged out in time series data) and you could use the athlete’s swim and run splits as rough gauges for confounding variables such as fitness, weather and nutrition from one race, to windtunnel, to second race.
Based on my own Chung testing where I try to control things as much as possible, I disagree that grabbing other people’s race performances and just assuming they “peaked” in the same way for both is going to provide much in the way of useful data. You can try to play confounding variable whack-a-mole as best you can. Maybe you could get very rough indications, but I’d consider the output less trustworthy than “Watts” on an Instagram post.
A 20w decrease in drag over 56 miles would be fairly easy to tease out (a lot gets averaged out in time series data) and you could use the athlete’s swim and run splits as rough gauges for confounding variables such as fitness, weather and nutrition from one race, to windtunnel, to second race.
Based on my own Chung testing where I try to control things as much as possible, I disagree that grabbing other people’s race performances and just assuming they “peaked” in the same way for both is going to provide much in the way of useful data. You can try to play confounding variable whack-a-mole as best you can. Maybe you could get very rough indications, but I’d consider the output less trustworthy than “Watts” on an Instagram post.
agree. i think with tri cycling it’s probably a futile exercise. even just with lionel, today’s weather in quebec was so crazy that the data wouldn’t be much good, plus we know he’s coming off an injury, so that makes comparison hard. finding two totally ‘clean’ full-gas races where all else is equal bar some aero changes would be really, really hard.
The question here is not whether drag is real and a thing, the question here is external validity.
Let’s put it this way, if we take several of Lionel Sanders races from last year, and we plug his results into best bike split, and then we add 30 W to his average power, do his most recent bike splits compare? I’m skeptical. yes, we can imagine that he is taking those watt saved and saving it for the run, but at some point somewhere, we should see his bike splits in the range of what we are seeing Magnus, Sam and Trevor doing. I am not seeing that.
Get Lionel’s Ocenaside file, get his files from last year and do that exercise. And BTW, this is before his ‘30w recent improvement’ in Germany.
The people that end up optmising their aero, chip away at it very slowly, It’s not a ‘big bang’ thing.
Some AGers, coming in from scratch do see a big effect in a one time session. But Lionel last year, changing setups every race was never chipping away. Go back to September when he tested flatter arms, that stuck through another road session and 2 tunnel sessions. More reach : same. The rest has been couple here, couple there. But by the hype he found 30 in September, 30 in December, 20 in January and 30 in Germany, it’s good for clicks and sponsors.
Compare a race last year. Compare Indian Wells (compare Sam’s power/speed) to Lionel’s. The compare Oceanside. See a trend ?
I tested 12 WT guys in December. Half had never been tested. They found probably 8w each on average. We analyze every race. First race, maybe half got “all their found watts”. The others…we chip away. Derek Gee podiumed at Dauphine, 6th in the TT and he did not do peak watts. He did do 1 tunnel and 2 road sessions. Chipping away at it.
Ask Magnus. There’s a guy that chips away constantly and make no noise.
Ask the guy that just won Canadian Nats if aero testing made a difference.
Ask the guy the guy that won Boulder 35-39.
These last two guys each had a single major ‘flaw’ and some little things. The major flaw was the deal breaker that allowed a break through. They were single changes resulting in double digits, that stuck.
AGer get 80% first session and 20% chipping away at it
Pros have to chip away constantly, unless they reset to 0 every time they race, which isn’t very Pro 
And BTW, it’s no different than swimming. They get down to 1.09/100m in their latest pool swim and they do 1:22 in open water race. Making that amazing CDA stick in a race is another story. This is why I harp on the need to get to a baseline, validate in training and confirm in racing.
I’m thinking your based question is “Does wind tunnel testing actually help?”
Seems the only method for figuring this out is using a more “large-data” methodology. Overall, have cycling speeds increased /w/kg over the last twenty years? Take the top ten pros at each race and compare their power (provided or calculated) and speed numbers over twenty years.
This won’t be exact, of course, but it will indicate whether wind tunnel testing is having an overall impact on the sport.
Whether it has a specific impact on a specific athlete is kind of irrelevant (IMHO), and is as easy to calculate as calculating where a hurricane will hit land in three days. There’s two much data to gather to calculate, and most of the data needed is not available. Something as tiny as where someone rides on the road would have some effect if the road is not a perfect road.
So, if overall times have come down proportionately to any increase in w/kg, the answer is “Yes, it’s made a difference.”
Welcome to social media and marketing.
Can good aero testing improve aero? …Yes
Will some people upsell this?..