Ed Coyle's paper about LA delta efficiency is a fraud.

[ScienceIsCool]

Just out of curiosity, I copied the ages of the Tour de France since 1953 into a spredsheet. There is no trend in ages for winners in the last 46 years (i.e., the age is not increasing. A running average is nice and smooth). In fact, the ages of the winners are normally distributed - actually, a convolusion of two gaussian distributions. The first is very small and has a peak at 24 years of age (sd <1 year). The second is much larger and has a peak at 28.8 years of age (sd ~1.5 years).

Lance's wins place him slightly outside this distribution (avg age = 30). THe difference is about one standard deviation and in statistical terms is "significant" even if it isn't large.

A more useful set of stats in this case might be the number of years racing as a pro prior to winning the Tour for the first time. That might be more revealing, but I don't feel looking up the numbers right now.

John Swanson
http://www.bikephysics.com

 

[ScienceIsCool]

Hmmm. But then I separated the data sets into pre-EPO (until 1990) and post-EPO.

The average age shifted by more than 2 years (post-EPO winners are older), and the distributions are completely different. When separated out, they are no longer normally distributed. That was just an artifact that coincidentally looked normally distributed.

Statistically, there's no chance that they came from the same overall data population. This suggests that some external factor such as EPO has caused a shift in how old Tour winners are.

Now to look at the Giro results to see if there's a similar trend.

John Swanson
http://www.bikephysics.com

 

[Digger]

Hmmm. But then I separated the data sets into pre-EPO (until 1990) and post-EPO.

The average age shifted by more than 2 years (post-EPO winners are older), and the distributions are completely different. When separated out, they are no longer normally distributed. That was just an artifact that coincidentally looked normally distributed.

Statistically, there's no chance that they came from the same overall data population. This suggests that some external factor such as EPO has caused a shift in how old Tour winners are.

Now to look at the Giro results to see if there's a similar trend.

John Swanson
http://www.bikephysics.com

Just as a furtherance of this point, I think the noteworthy point here is the age the cyclist was when they won their FIRST tour.

 

[Alpe d'Huez]

Interesting stuff. A bigger question I would like answered is a graph showing the longevity of riders from the EPO era to race into their 30's at a top level.

Though I have no direct proof, it seems to me that while some new faces and younger riders have been trying to clean the sport up to a degree, there have been numerous riders over the last 5 years or so from the "old days" that manage to keep riding at a high level, and many of them getting busted in their 30's (Rasmussen, Vinokourov, Landis, Hamilton, Heras, Piepoli, Schumacher, Ullrich). Of course, some younger riders have been busted as well (Ricco, Kohl), but I'd think a graph would lean towards riders that came up during the EPO era.

And I beleive that thinking is of of the reasons why many people don't like Lance, and don't want Landis or Rasmussen or Vinokourov back, and don't welcome other older riders and dopers back. It feels as though just as we're cleaning up the sport with fresh faces, the big name dopers come back to insure doping continues, the omerta remains in full force, and nothing changes.

 

[cyclingmad]

Omerta

Interesting stuff. A bigger question I would like answered is a graph showing the longevity of riders from the EPO era to race into their 30's at a top level.

Though I have no direct proof, it seems to me that while some new faces and younger riders have been trying to clean the sport up to a degree, there have been numerous riders over the last 5 years or so from the "old days" that manage to keep riding at a high level, and many of them getting busted in their 30's (Rasmussen, Vinokourov, Landis, Hamilton, Heras, Piepoli, Schumacher, Ullrich). Of course, some younger riders have been busted as well (Ricco, Kohl), but I'd think a graph would lean towards riders that came up during the EPO era.

And I beleive that thinking is of of the reasons why many people don't like Lance, and don't want Landis or Rasmussen or Vinokourov back, and don't welcome other older riders and dopers back. It feels as though just as we're cleaning up the sport with fresh faces, the big name dopers come back to insure doping continues, the omerta remains in full force, and nothing changes.

Good point about not wanting the old cheats back. Personally I dont want to see them back. One thing that confuses the argument of there being two groups of younger clean riders versus older omerta supporters is Klodens comments in last years Giro. Cant remember exact wording but along the lines of he said he was part of one of the most rigourous testing program at Saunier dont know what Ricco and Saunier Duval are doing and certainly dont think CSF Navigare are doing anything about it. That comment sounds like breach of Omerta by an older rider. Both teams got caught out later in the year

 

[Race Radio]

Just as a furtherance of this point, I think the noteworthy point here is the age the cyclist was when they won their FIRST tour.

I think the first GT would be important. The TdF has always been a big deal but in the 50's and 60's many riders focused on their national Tours and did not ride the TdF. Coppi won the Giro at 21 but did not race the Tour until 29 (WWII also caused a bit of a lag).

 

[Hugh Januss]

Interesting stuff. A bigger question I would like answered is a graph showing the longevity of riders from the EPO era to race into their 30's at a top level.

Though I have no direct proof, it seems to me that while some new faces and younger riders have been trying to clean the sport up to a degree, there have been numerous riders over the last 5 years or so from the "old days" that manage to keep riding at a high level, and many of them getting busted in their 30's (Rasmussen, Vinokourov, Landis, Hamilton, Heras, Piepoli, Schumacher, Ullrich). Of course, some younger riders have been busted as well (Ricco, Kohl), but I'd think a graph would lean towards riders that came up during the EPO era.

And I beleive that thinking is of of the reasons why many people don't like Lance, and don't want Landis or Rasmussen or Vinokourov back, and don't welcome other older riders and dopers back. It feels as though just as we're cleaning up the sport with fresh faces, the big name dopers come back to insure doping continues, the omerta remains in full force, and nothing changes.

Have the "older" riders been caught more because they have been tested more over a longer period of time or because they were actually doping more.
Also adding Rasmussen and Ullrich to the list of older riders who have been busted is a little disingenuous if by busted you mean failed a test.
I wonder if the idea that the "new generation" is actually cleaner than the "dirty old generation" is anything more than a optical illusion and that as the group as a whole spends more seasons in the pro ranks the percentages will even out.
On the other hand McQuaid's most recent statements may mean that that will not be happening whether they are infact any cleaner or not.

 

[ScienceIsCool]

The average age of all first-time winners of the Tour de France from 1953 to 2009 is 27.2 years old.

In the pre-EPO era (up until 1990), the average age of a first time winner was 26.9 years old. In the post-EPO era (1991-2009), the average age was 27.8 years old. No change in the distribution of ages, either.

It could be concluded that the advent of bicycle technology (light, aerodynamic bikes), training methods, and oxygen vector PEDs (such as EPO) had no effect on how young a rider would win their first Tour de France.

John Swanson
http://www.bikephysics.com

 

[ScienceIsCool]

And for first-time Giro winners:

The average age of all first-time winners of the Giro d'Italia from 1954 to 2009 is 27.2 years old. The same as for the Tour de France.

In the pre-EPO era (up until 1990), the average age of a first time winner was 26.5 years old. In the post-EPO era (1991-2009), the average age was 28.3 years old. No change in the distribution of ages, either.

These results are near identical to the Tour results I posted earlier. What's interesting is that although the difference is small between pre and post EPO eras, there is a difference of about 1-2 years. This could indicate a general trend involving aging athletes that has nothing to do with PEDs, but may have another physical or sociological cause such as longer careers due to better pay.

John Swanson
http://www.bikephysics.com

 

[Race Radio]

The average age of all first-time winners of the Tour de France from 1953 to 2009 is 27.2 years old.

In the pre-EPO era (up until 1990), the average age of a first time winner was 26.9 years old. In the post-EPO era (1991-2009), the average age was 27.8 years old. No change in the distribution of ages, either.

It could be concluded that the advent of bicycle technology (light, aerodynamic bikes), training methods, and oxygen vector PEDs (such as EPO) had no effect on how young a rider would win their first Tour de France.

John Swanson
http://www.bikephysics.com

Thanks for the info, but I do not think this address the issue correctly.

The challenge is prior to the introduction of EPO did riders show their GT ability early, or take "Years to develop" as they did post 1990? A good indicator would be when a Tour de France winner first stood on a podium of a GT. If you look at pre 1990 winners this was almost always 25 or Younger

Robic - 26
Bartali- 22
Coppi- 20
Kübler- 29
Koblet -25
Bobet - 25
Walkowiak- 29
Anquetil -23
Gaul - 22
Bahamontes -25
Nencini - 25
Gimondi - 22
Aimar - 25
Pingeon - 26
Janssen - 27
M jersey at 24
Thévenet- 23
Van Impe- KOM at 23
Zoetemelk- 23
Hinault - 23
Lemond- 24
Fignon- 22

 

[Race Radio]

Race Radio,

if you want to make that poin, then once again we need the full list. What is the list of those who won the tour for the first time above the age of 25 and how does that compare to now? You can't always use partiial info. But as I said, it wouldn't susprise me if younger people had a better shot at the tour in the days when it was less professional and older athletes didn't look after themselves in the way they do today. We see this pattern across most sports.

What is a "poin"? .....and what is your point?

As usual you make no sense. The fact remains that until EPO it was rare for a rider to suddenly become a GT winner in his late 20's early 30's without showing any previous ability to do so. This B.S. about riders needing "years to perfect" is just more of the Armstrong myth machine

he vast majority of pre EPO tour winners showed they had the engine age 25 or younger

 

[sashimono]

ScienceIsCool, I get an average age a year older than you, namely 28.16 for TdF winners 1953-2009 and 28.14 for 1903-2009. Some guys like Indurain have their birthdays during the tour so maybe its a roundoff thing?

Also making conclusions about trends and multiple convolved distributions in such a small data set seems to me as fishy as what Coyle did. Maybe thats why its in this thread?

 

[Hugh Januss]

I think the spoiler in this whole question is LA. He came along and invented a way of transforming himself from a good classics rider into an unbeatable TDF racer. Whether he did it with cancer, a moose heart, the ultimate coctail of PEDs or just good old gosh darn American hard work, he took out a 7 year swath of availible tour wins.
To say that since the "EPO" age the TDF winners age as increased I think is ignoring LA's impact. He blocked out an extended period of time and eclipsed whatever young tour stars may have been coming up in that time. After he retired there was a void that was filled for a couple years by older riders it's true, but that has happened in the past from time to time. Now with the emergence of Contador and Schlecklet and a couple of others this seems to be righting its self again.
So maybe it's wrong to blame EPO for this sudden change, maybe we should just blame Armstrong.

 

[Anonymous]

Race Radio,

if you want to make that poin, then once again we need the full list. What is the list of those who won the tour for the first time above the age of 25 and how does that compare to now? You can't always use partiial info. But as I said, it wouldn't susprise me if younger people had a better shot at the tour in the days when it was less professional and older athletes didn't look after themselves in the way they do today. We see this pattern across most sports.

That's GOLD JERRY, GOLD!!!!

 

[Hugh Januss]

You only provide partial info, and as someone noted, LA took up a good chunk of wins so the comparison wouldn't be fair in any event.

It stands to reason that in the modern era, where the sport is more professional and athletes look after themselves far better, that the average age would nudge up a bit. This is the same in many sports. I've nothing to counter this logic. Though as I said, it may well be that someone like LA didn't shine in the Tour in the early years because he became a young pro just as EPO started. That is an interesting scenario that I hadn't thought about before but does indeed make sense.

Oh yeah. I see what you mean. Lance hadn't yet learned how to use EPO in such a way as to enable him to win the tour.
Yes I think you are definately on to something there.

 

[ScienceIsCool]

ScienceIsCool, I get an average age a year older than you, namely 28.16 for TdF winners 1953-2009 and 28.14 for 1903-2009. Some guys like Indurain have their birthdays during the tour so maybe its a roundoff thing?

Also making conclusions about trends and multiple convolved distributions in such a small data set seems to me as fishy as what Coyle did. Maybe thats why its in this thread?

Oh, definitely. Right on all points. I took year of birth and didn't account for whether the b-day was before or after the Tour.

I pointed out that it could be a convolved set of distributions, but later showed that it really wasn't. I'm just riffing, not being totally anal about the analysis before sharing. You know. It's not like I'm submitting to a journal or anything. Just talking amongst peers.

John Swanson
http://www.bikephysics.com

 

[Race Radio]

I've nothing to counter this logic.

At least you admit it

 

[Franklin]

Anyways, I think we can safely conclude that rider peak at end 20, begin 30 is very disputable.

Its mid twenty and even if the late shift is insignificant, which I sincerely doubt, it put's some serious strain in the efficency story. *Once again I point out that the truly great tend to peak even earlier, with only a few exceptions*

On a side note, my link pointed to even a bigger data set, Giro, Vuelta, TDF and other monuments. It also showed that drop off is much bigger after 28, the data isn't bellshaped around 28.

On the point that science+progress makes riders peak later, for example Swimming, Tennis and Speedskating disagree. The case can be made that science+progress enables people to peak earlier. BPC, you have no proof either way. The only data set that we have showing a shift is the cycling one and it seems to fall right into the Epo era.

Oh and that Lance won late is of course logical due to cancer. I'm a big cynic(chances he didn't use Epo are approx zero), but that is a fair excuse. Had he won a GT two years earlier it would have been more into the general trend.

 

[Krebs cycle]

I think this "analysis" of the age of tdf winners is lacking another aspect of the tours history ie: that a number of those young first time winners went on to be multiple winners. Endurance performance in the controlled laboratory environment is highly related to 3 physiological variables, VO2max, lactate threshold and efficiency. So it stands to reason that a young gun with a high VO2max could come along and win at a young age, but then go on to win multiple times as their LT gradually improved over the yrs. The data available suggest that those athletes with the highest VO2max tend to have lower efficiency (Lucia et al., 2002) and those that improve their efficiency over time tend to have a lower VO2max to begin with (Santalla et al., 2009). So a young rider with a lower VO2max may not have the firepower to begin with, but after several yrs at pro level, their LT and efficiency could both have improved putting them into a position where they are now able to be a genuine contender. This could partially explain the difference between the average age that tdf winners first won versus the (higher) average age of overall tdf winners.

Now here comes a rather interesting possibility. In 2002, Lucia published a paper (http://www.ncbi.nlm.nih.gov/pubmed/12047807) which shows an significant relationship between [Hb] and greater than expected efficiency (at workrates above LT). Since it is very well known that doping can increase [Hb] then it is plausible that an artificial increase alone could account for the improved efficiency.

However, it is also interesting to note that both Lucia and Santalla appear on numerous research articles together and their method ofVO2 measurement have been criticised by Jeukendrup, Martin and Gore (http://www.ncbi.nlm.nih.gov/pubmed/12471319) and is known to display error as large as 30% at high breathing frequencies (http://jap.physiology.org/cgi/reprint/81/6/2495)

So the idea that cycling efficiency increases in professional cyclists is an intriguing possibility which makes sense intuitively and would certainly contribute to an explanation of why the average age of tdf winners is higher than the average age at which tdf winenrs first won the tdf, but despite the debate here with acoggan, I do not believe this question to be properly answered YET. With respect to LA, it most definately is not answered because Ed Coyle's paper is flawed.

 

[Race Radio]

This is in line with most sports.

Do you have anything to support this claim, or did you make it up like most of your claims?

Why is it the same trolls that say Armstrong was a freak physical specimen who was beating "Pro" triathletes as a 16 year old also say it took almost a decade for him to become a GT rider?

 

[Race Radio]

It wouldn't be that amazing to beat pro-triathletes in the cycling part of the course. That's not really in the same league. These are one day events as well.

In tour riding it often depends on whether your team puts up for the grand tours, whether they tell you to go for stage wins, or make you domestique for a few years to learn the trade, and all that, so there are many reasons why someone might not show their full potential until the later years. You'd probably have to look at how the team structures have changed over the years as well to make your long term analysis.

Simplistic anti Armstrong talking points maybe common currency on the internet and have developed a type of credibility through repetition, but that doesn't mean they are true.

 

[acoggan]

In 2002, Lucia published a paper (http://www.ncbi.nlm.nih.gov/pubmed/12047807) which shows an significant relationship between [Hb] and greater than expected efficiency (at workrates above LT). Since it is very well known that doping can increase [Hb] then it is plausible that an artificial increase alone could account for the improved efficiency.

Except that Coyle made his measurements below the point at which the VO2-power relationship significantly deviates from linearity.

it is also interesting to note that both Lucia and Santalla appear on numerous research articles together and their method ofVO2 measurement have been criticised by Jeukendrup, Martin and Gore (http://www.ncbi.nlm.nih.gov/pubmed/12471319)

Boy, if only Gore spent half as much time writing papers as letters-to-the-editor...

But anyway, since you bring up the Jeukendrup/Martin/Gore paper: you do realize, don't you, that much of the motivation for that publication stemmed from the famously bad blood between Coyle and Jeukendrup/Martin?

and is known to display error as large as 30% at high breathing frequencies (http://jap.physiology.org/cgi/reprint/81/6/2495)

I gather that you didn't read Dave Proctor's paper all that closely:

"It should be noted that the manufacturer
of the CPX/D system we used (Medical Graphics)
routinely performs validation testing at high respiratory
rates when in development and on each unit before
it is shipped. Our system required revalidation because
we used a mass spectrometer in place of supplied gas
analyzers. These errors were not due to errors in
measurement of V ˙ E or errors in the gain of the gas
analyzers but were due to induced errors in delay times
used to align flow and gas concentration signals."

IOW, the errors that were observed were due to the modification that they made, and hence this paper does not say anything about the validity of the data published by Lucia et al.

So the idea that cycling efficiency increases in professional cyclists is an intriguing possibility which makes sense intuitively and would certainly contribute to an explanation of why the average age of tdf winners is higher than the average age at which tdf winenrs first won the tdf, but despite the debate here with acoggan, I do not believe this question to be properly answered YET.

On this we would agree.

With respect to LA, it most definately is not answered because Ed Coyle's paper is flawed.

Ah, but here's the thing: no scientist truly worthy of the name really gives a d*** whether Armstrong's efficiency improved or not. What is intriuging is the fact that the efficiency of someone who trained a lot apparently increased over time...it is that observation that has spurred a number of additional studies, studies that probably would not have been conducted had the paper not been published. As such, publication of Coyle's paper has moved the field forward even if the observations themselves may be questioned.

 

[acoggan]

Good point.
According to Coyle, Lance's weight loss during cancer was a big factor in this improvement in efficiency.

This statement is completely incorrect.

 

[Digger]

This statement is completely incorrect.

"Lance went up 18 percent and that's because he both improved his raw power ability by 8 percent because of muscle adaptations which we ascibe to just the pure result of hard training."
"And then he lost body weight and the two combined equally to cause this 18 percent in power."

Ed Coyle

 

[Hugh Januss]

"Lance went up 18 percent and that's because he both improved his raw power ability by 8 percent because of muscle adaptations which we ascibe to just the pure result of hard training."
"And then he lost body weight and the two combined equally to cause this 18 percent in power."

Ed Coyle

Yes, but as acoggan pointed out that is an incorrect statement.