"clean", "suspect", "miraculous" and "mutants"

[L'arriviste]

I got confirmation yesterday from one of the authors that neither Sastre nor J-Rod were included for reasons of lesser interest in comparison with the other cases presented.

Nevertheless they have the numbers and they do intend publish them on the website.

 

[JimmyFingers]

Thank you brodeal. I also found what Mexckx wrote...



So my apologies for not reading the entire thread. Now onto the problems inherent in the analysis. For starters, the long list of assumptions remains. No need to go over that, its all been discussed in the estimated power thread.

Something sort of novel here is the "normalization" procedure. This introduces another source of error.....

W/kg is not directly proportional to velocity (even in the absence of environmental variability). ie: 2 different riders with exactly the same w/kg do not go the same velocity. This is because bodyweight is not directly proportional to differences in energy production ie: for a 2% difference in bodyweight you do not get a 2% change in power production because some of that difference in weight has nothing to do with locomotor energy production (ie: it is bone or connective tissue or vital organs etc).

There are also differences in thermoregulation between smaller and larger individuals. Smaller individuals with exactly the same w/kg will have a comparative advantage vs larger individuals because they have a greater surface area to bodyweight ratio. This enhances cooling and generally enables smaller individuals to perform better esp in hot conditions.

So if you "normalize" to a set bodyweight + bike weight you will re-introduce these errors which are actually already present in the initial estimation of power based on VAM (ie: you double dip on those sources of error). The further away from the bodyweight of 70kg the riders actual weight is, the greater the error will be. This is why this Vayer publicity stunt is published as a MAGAZINE and not in a scientific journal, because no proper science journal would touch it with a 10 foot pole.

but hey, keep up the pseudo scientific BS and believe whatever it is you want to believe. Nobody can be better that Lemond unless they are doping.... ever.... in all the future of cycling. SOOOO much hero worship around here its nauseating.

edit: IMO such an analysis is interesting from a broad 'trending' aspect of performances. It would be useful to have data across more than 100 athletes to examine trends, but it cannot be used in the way Vayer is using it ie: to make some definitive claim about doping suspicion betwe individuals. There are simply way too many sources of error to do that with enough precision. But that is what will sell and make Vayer money and you suckers will lap it up.

Good stuff. So many times analysis that seeks to incriminate seems to purposely simplify that analysis. This looks like lots of pretty colours but little else. And again we have LeMond as the shining light against whom all riders present and future must be judged.

 

[valkus]

I got confirmation yesterday from one of the authors that neither Sastre nor J-Rod were included for reasons of lesser interest in comparison with the other cases presented.

Nevertheless they have the numbers and they do intend publish them on the website.

Good news!

 

[valkus]

With so many numbers there are bound to be mistakes here and there.

In this particular case I don't see where the mistake you point out (6.15 miraculous/6.25 suspicious) could have come from. Possibly from placing the power numbers (real and normalized power) in the wrong columns of their worksheets.

W.r.t. Vino 99 Ventoux, he finished second, 43sec behind Vaughters 56:50 (not Mayo as indicated)

While I am ready to believe that Vino was clean early on in his career, I have no such delusion w.r.t. Moreau.

I don't have the magazine, and you probably know better, but the classification miraculous/suspicious (the threshold levels) are different as a function of last climb of a big mountain stage/ ITT or at maximum second climb of the stage/ breakaway all the day

 

[vrusimov]

Good stuff. So many times analysis that seeks to incriminate seems to purposely simplify that analysis. This looks like lots of pretty colours but little else. And again we have LeMond as the shining light against whom all riders present and future must be judged.

I would not say that Lemond is the shining light but average tour performances climbed, starting with Indurain in 1993...

But question...Do you think that Bobby Julich (confessed doper) was a greater talent than Lemond? In 1998 his tour average was 420W on his way to third place.

In Not Normal, there are almost 50 guys that developed 410W or better over an entire tour. That is not including guys like Mayo, DiLuca, Berzin, Santambrogio, Ricco and many others too low profile to mention. This is not rocket science. 50 guys in 20 years doesn't fit the definition of a once or twice a generation athlete. Anyone can be exceptional with the right type of doctor. In particular, the story on Ullrich was that he was addicted to EPO as early as 1992. He would have won the tour the first time he rode it, beating Indurain no less, if it were not for Riis.

 

[wansteadimp]

A question, with these comparisons of w/kg have improvements in technology been factored in?

The bikes that Lemond (esp at the start of his career), Hinault etc rode were completely different in terms of weight & technology than the ones being ridden now.

I suppose the real question is are these differences minor and not worth worrying about or have they already been built in or are we comparing apples and oranges.

 

[L'arriviste]

A question, with these comparisons of w/kg have improvements in technology been factored in?

The bikes that Lemond (esp at the start of his career), Hinault etc rode were completely different in terms of weight & technology than the ones being ridden now.

I suppose the real question is are these differences minor and not worth worrying about or have they already been built in or are we comparing apples and oranges.

I don't have the magazine here right now to check but I think I remember reading that they do take account of that.

 

[Le breton]

I don't have the magazine, and you probably know better, but the classification miraculous/suspicious (the threshold levels) are different as a function of last climb of a big mountain stage/ ITT or at maximum second climb of the stage/ breakaway all the day

Those 3 values refer to the same Ventoux TT climb of 1999 (DL)

 

[Le breton]

I don't have the magazine here right now to check but I think I remember reading that they do take account of that.

Yes +2 kg for Hinault, LeMond, Fignon. See page 139
Also : they reduce the bicycle efficiency.
However they are ambiguous.
For post 90's they take a transmission loss of 2.5% (chain mostly).

Pre-90's they state page 139 that the efficiency is reduced by 5%.
Do they mean it was 92.5% or 95%?

I assume they mean it was 95% as elsewhere I have read that Vayer claimed that automatic pedals reduced the loss in the applied power by 2%. Which I would like to see confirmed by some scientific source.

 

[vrusimov]

A question, with these comparisons of w/kg have improvements in technology been factored in?

The bikes that Lemond (esp at the start of his career), Hinault etc rode were completely different in terms of weight & technology than the ones being ridden now.

I suppose the real question is are these differences minor and not worth worrying about or have they already been built in or are we comparing apples and oranges.

Jan Janssen 1968 rode 8.7kg
Van Impe in 1976 rode 8.3kg, which was one of the lightest at the time
Lemond/Hinault rode ~9.5kg in the eighties [9.1 for Lemond TT bike 1990]
Indurain's lightest was 8-9kg
Riis was 9kg along with Ullrich
Pantani 8kg

UCI went to 6.8kg minimum in 2000:

Armstrong 8.2kg - 2002 and 7.2kg in 2003...6.8kg all other years
6.8kg for all succeeding winners of the Tour

The 2kg difference in watts/kg:

300W at 70kg = 4.28 w/kg...68kg = 4.41 w/kg...difference .13 w/kg
350W at 70kg = 5 w/kg...68kg = 5.147 w/kg...difference .147 w/kg
400W at 70kg = 5.71 w/kg...68kg = 5.88 w/kg...difference .17 w/kg
~3% power difference in watts...

Vayer on A'lpe D'Huez:
1 kilo = :20 time saved... :40 saved at 2kg

While 2kg certainly makes a difference, there is still the weights of the riders, which must be derived from published figures. The "Armstrong fudge" as it were, could easily nullify that 2kg difference.

Some of the weights are more than Vayer's 8kg bike allowance so power required would be slightly higher for Lemond, Indurain, Ullrich, Riis, Hinault and others with heavier bikes. Vayer probably accounted for this discrepancy. And a quick look back confirms the two standards he used, 78 and 80kg.

Vayer goes into detail from pages 138-143 on the methodology of his calculations.

http://felixwong.com/2010/11/tour-de-france-bicycles-historical-bike-weights/

 

[thehog]

Good stuff. So many times analysis that seeks to incriminate seems to purposely simplify that analysis. This looks like lots of pretty colours but little else. And again we have LeMond as the shining light against whom all riders present and future must be judged.

Yes, Jimmy lots of pretty colours. That's all it is

A bit like ZTP & marginal gains. Pretty colours.

We finally have a serious thread with excellent debate and you think it's about 'pretty colours'.

I often wonder if you give any thought to what you post.

 

[Merckx index]

Normalising to a "standard" 70kg rider is redundant, since a W/kg value already normalises performance. Just plot the W/kg values.

Indeed I would argue that you don't need to calculate W/kg at all, since that simply introduces errors in assumptions (e.g. such as bike / kit mass as you say, and other things I've mentioned before), so you may as well just chart climbing times for a given climb to provide a trend analysis for each climb.

The main reason W/kg is estimated is to compare performances from different climbs. There is no need to normalise to a standard sized rider/bike to provide such comparisons. Indeed I think it just confuses people.

For these cross climb analyses, I think it would be helpful to plot W/kg with error bars.

I agree with all this, particularly the bolded.

W/kg is not directly proportional to velocity (even in the absence of environmental variability). ie: 2 different riders with exactly the same w/kg do not go the same velocity. This is because bodyweight is not directly proportional to differences in energy production ie: for a 2% difference in bodyweight you do not get a 2% change in power production because some of that difference in weight has nothing to do with locomotor energy production (ie: it is bone or connective tissue or vital organs etc).

No, this is not correct. If one rider is 2% heavier than another rider, the muscles involved in power production will also have about 2% more mass. Not exactly, because every body has a slightly different shape and proportion of parts, but certainly close enough to be well within the error margin that is inevitable in these calculations. What you are saying is that if one rider weighs 60 kg and another 66 kg (let’s use 10% difference to make the example clearer), the second rider’s power muscles will not make up all of that 6 kg difference. Of course not, but that doesn’t matter. What matters is that the 10% ratio is conserved wrt muscles. If it is, then one anticipates 10% more power.

VAM gets some criticism for being an inexact measure of power/weight, but the most important uncertainties are external to the rider--weather conditions and changes in the grade. If the grade is constant and there is no wind, then VAM is a very good measure of power/weight. Keep in mind that the original definition of power, going back several centuries to the Industrial Revolution, is the force necessary to lift a certain weight a certain height in a certain length of time. Even modern power meters are in effect calibrated to this parameter. Since riders of different size are lifting different weights, VAM is good measure of power/weight, or watts/kg. Other sources of variation such as friction and wind resistance tend to be fairly minor and in any case do not vary much from rider to rider. You would want to know them and correct them for the best estimates, but they don't completely invalidate general conclusions made without taking them into consideration.

There are also differences in thermoregulation between smaller and larger individuals. Smaller individuals with exactly the same w/kg will have a comparative advantage vs larger individuals because they have a greater surface area to bodyweight ratio. This enhances cooling and generally enables smaller individuals to perform better esp in hot conditions

Again, I have to disagree. If the smaller rider’s body cools better, this will result in a faster climb and a higher watts/kg value. The latter takes into account temperature regulation, and any other physiological variable that affects power.

What you are referring to, I think, is power as it would be measured at some intermediate stage before reaching the pedals. But while power like this could be determined, it’s not really relevant to a bike race. The most relevant measure of power is the watts actually transferred to the pedals. If a large rider and small rider put out identical watts/kg in this sense, they will make the climb in exactly the same time, again by definition.

So if you "normalize" to a set bodyweight + bike weight you will re-introduce these errors which are actually already present in the initial estimation of power based on VAM (ie: you double dip on those sources of error). The further away from the bodyweight of 70kg the riders actual weight is, the greater the error will be. This is why this Vayer publicity stunt is published as a MAGAZINE and not in a scientific journal, because no proper science journal would touch it with a 10 foot pole.

I agree with you that the bike weight introduces a potential error in the calculations. It would actually be fairer if every rider used a bike that was proportional to his weight. E.g., if a 70 kg rider uses an 8 kg bike, a 56 kg rider should use a 6.4 kg bike. That way, the total weight lifted would always stand in the same proportion to body mass. In this respect the relationship of VAM to watts/kg is imperfect. That is, it actually measures watts/total kg., not watts/kg of body weight. But of course if you know the weights of the different riders and their bikes, you can correct for this.

but hey, keep up the pseudo scientific BS and believe whatever it is you want to believe. Nobody can be better that Lemond unless they are doping.... ever.... in all the future of cycling. SOOOO much hero worship around here its nauseating.

edit: IMO such an analysis is interesting from a broad 'trending' aspect of performances. It would be useful to have data across more than 100 athletes to examine trends, but it cannot be used in the way Vayer is using it ie: to make some definitive claim about doping suspicion betwe individuals. There are simply way too many sources of error to do that with enough precision. But that is what will sell and make Vayer money and you suckers will lap it up.

If you really believe this, you need to provide error estimates to substantiate this conclusion. There certainly are sources of variation, primarily I would say in variability in the grade (when comparing times up different climbs, and even to a slight extent comparing times up the same climb, as some riders may do relatively better on some grades than others) and in weather conditions. But in the absence of detailed analysis, I don't buy that these uncertainties completely blur broad distinctions between classes like mutant, miraculous and suspicious.

 

[fmk_RoI]

Can the same questions be asked of track performances as climbing, using watts/watts per kilo? Or is a different metric used? See answer to third from last Q here http://www.podiumcafe.com/2013/6/10/4415124/interview-edward-pickering

 

[Fearless Greg Lemond]

Vetoo has done a great job making an all time list for the Alpe:

149. Alejandro Valverde __ ESP | 41:45 | 2008
150. Frank Schleck _______ LUX | 41:45 | 2008

151. Andrew Hampsten _____ USA | 41:45 | 1991
152. Vladimir Efimkin ____ RUS | 41:47 | 2008
153. Cadel Evans _________ AUS | 41:47 | 2008
154. Denis Menchov _______ RUS | 41:47 | 2008
155. Christian Vandevelde_ USA | 41:47 | 2008
156. Bernhard Kohl _______ AUT | 41:47 | 2008
157. Anthony Charteau ____ FRA | 41:48 | 2004
158. Laudelino Cubino ____ ESP | 41:48 | 1995
159. Pierrick Fedrigo ____ FRA | 41:49 | 2004
160. Luis Herrera ________ COL | 41:50 | 1987

161. Yuriy Krivtsov ______ UKR | 41:53 | 2004
162. Stefano Garzelli ____ ITA | 41:56 | 2006
163. Laurent Fignon ______ FRA | 41:56 | 1987
164. Yaroslav Popovych ___ UKR | 41:57 | 2006
165. Michael Boogerd _____ NED | 41:57 | 2006
166. Richard Virenque ____ FRA | 41:57 | 2004
167. Marius Sabaliauskas _ LTU | 41:57 | 2004
168. Marzio Bruseghin ____ ITA | 41:58 | 2004
169. Kim Kirchen _________ LUX | 41:59 | 2004
170. Carlos Contreras ____ COL | 41:59 | 1999

____ 42:00 - 42:14 | 20
171. Ruben Lobato ________ ESP | 42:00 | 2006
172. Pierre Rolland ______ FRA | 42:03 | 2011
173. Sylvain Chavanel ____ FRA | 42:04 | 2006
174. Kevin Livingston ____ USA | 42:04 | 1997
175. Oscar Sevilla _______ ESP | 42:05 | 2004
176. Francisco Mancebo ___ ESP | 42:06 | 2001
177. Igor G. de Galdeano _ ESP | 42:06 | 2001
178. Roberto Laiseka _____ ESP | 42:06 | 2001
179. Peter Velits ________ SVK | 42:07 | 2011
180. Cadel Evans _________ AUS | 42:07 | 2011

181. Thoams de Gendt _____ BEL | 42:07 | 2011
182. Damiano Cunego ______ ITA | 42:07 | 2011
183. Frank Schleck _______ LUX | 42:07 | 2011
184. Andy Schleck ________ LUX | 42:07 | 2011
185. Abelardo Rondon _____ COL | 42:08 | 1991
186. Santiago Botero _____ COL | 42:11 | 2004
187. Leonardo Piepoli ____ ITA | 42:12 | 2001
188. Ronan Pensec ________ FRA | 42:12 | 1991
189. Eddy Mazzoleni ______ ITA | 42:14 | 2006
190. Thierry Bourguignon _ FRA | 42:14 | 1997

____ 42:15 -
191. Jerome Pineau _______ FRA | 42:15 | 2004
192. Pedro Delgado _______ ESP | 42:15 | 1987
193. Pedro Delgado _______ ESP | 42:15 | 1989
194. Laurent Fignon ______ FRA | 42:15 | 1989
195. Bobby Julich ________ USA | 42:16 | 2004
196. Igor G. de Galdeano _ ESP | 42:16 | 2004
197. Pascal Richard ______ SUI | 42:16 | 1991
198. Reynel Montoya ______ COL | 42:16 | 1991
199. Iker Flores _________ ESP | 42:17 | 2004
200. Beat Breu ___________ SUI | 42:17 | 1982

Go see for yourself at:
http://www.fillarifoorumi.fi/forum/...km-h-VAM-W-W-kg-etc-%29&p=2041608#post2041608

That is frigging disgusting. Herrera not even making the top 150.
160. Luis Herrera ________ COL | 41:50 | 1987

 

[Lanark]

Wow, who would have guessed that Ruben Lobato was a better climber than Delgado and Fignon?

 

[Alex Simmons/RST]

Vetoo has done a great job making an all time list for the Alpe:

I don't read that language (Finnish?) save for a few fun phrases that I no doubt mispronounce. Does it say how they dealt with all the different timing points used over the years?

If climbing times are to be compared, then we need to know they are measured using the same start and finish points.

 

[Big Doopie]

Vetoo has done a great job making an all time list for the Alpe:

149. Alejandro Valverde __ ESP | 41:45 | 2008
150. Frank Schleck _______ LUX | 41:45 | 2008

151. Andrew Hampsten _____ USA | 41:45 | 1991

Amazing that Lemond, mottet, fignon, claveyrolat and hampsten all did the best times of their careers chasing epo-driven chiappucci, Indurain and Bugno in 1991.

That year Lemond thought he was having an off year after dominating the first week. He couldn't explain why the peloton was going so fast, instead of slowing some riders seemed to get stronger and faster as the tour grew longer. Usually the best in the third week, lemond exhausted himself chasing the new epo wonders around France trying desperately to compete with them. And all the time wondering what was wrong with himself. Meantime he actually posted the best time of his career up l'alpe...But way behind the new epo kings.

 

[Nastyy]

I don't read that language (Finnish?) save for a few fun phrases that I no doubt mispronounce. Does it say how they dealt with all the different timing points used over the years?

If climbing times are to be compared, then we need to know they are measured using the same start and finish points.

Doesn't say anything about which timing points he used.

 

[burning]

146. Greg LeMond _________ USA | 41:42 | 1991
147. Benjamin Noval ______ ESP | 41:44 | 2004

I loved that part

 

[Fearless Greg Lemond]

I don't read that language (Finnish?) save for a few fun phrases that I no doubt mispronounce. Does it say how they dealt with all the different timing points used over the years?

If climbing times are to be compared, then we need to know they are measured using the same start and finish points.

The times arent that hard to find out, a lot of work I guess, but I agree, you also have to take the course into account. When it is like 1987 with the Cucheron/Coq/Laffrey it is something completely different to lets say 1997 with the total 'flat' sprint to the Alpe.
Too bad the footage of the great stage in 1984 is too short to make a good timing. But I found this:
Herrera goes from this corner, I think it is corner 10
http://www.youtube.com/watch?feature=player_detailpage&v=xS-_8nmW_kg#t=1426s


to the finish
http://www.youtube.com/watch?feature=player_detailpage&v=xS-_8nmW_kg#t=2661s

in 20 minutes and 37 seconds.

Riis goes from that same corner in 1995
http://www.youtube.com/watch?feature=player_detailpage&v=YNB8fKFmiek#t=4371s

to the finish
http://www.youtube.com/watch?feature=player_detailpage&v=YNB8fKFmiek#t=5431s

in 17 minutes and 56 seconds

That is 2 minutes and 41 seconds faster than one of the best climbers of all time. Since Riis is clocked at 38 minutes and 16 seconds for the 1995 climb, when absolutely correct, Herrera would be at best have a time of around 41 minutes. Or would he be going faster for the first 20 minutes than an Epofuelled Riis? I bet he lost the same amount of time for the first half of the climb as well. And yes, the roads were better in 1995, the bikes weight less etc etc, but that does not make up for this timedifference. Given the best time of Herrera is listed at 41.50, from 1987, my assumption - yes assumption - isnt groundbreaking.

Then I thought, hey, why not compare with 1990 as well. Well:
http://www.youtube.com/watch?feature=player_detailpage&v=LgrnKi8xemM#t=5199s

Bugno, Claverolat, LeMond doing the same distance in around 20 minutes 40. Perico lost 40 seconds there, he did a lot of work there and payed the prize.

Please note the presence of Breukink at the finishline, he was back two minutes thirty on the beginning of the Alpe, he rode at his 'own pace' to the front.
He is quoted doing the Alpe in and around 40 minutes that day. He rode with Fabio Parra that day.

I timed the last 11K
http://www.youtube.com/watch?feature=player_detailpage&v=rIqtQL5yN8s#t=4292s

give or take 35 minutes 10 seconds.

Cant find footage for the first 2.8K of the climb, the steepest part.

Clayverolat should have won that day by the way.

Amazing that Lemond, mottet, fignon, claveyrolat and hampsten all did the best times of their careers chasing epo-driven chiappucci, Indurain and Bugno in 1991.

That year Lemond thought he was having an off year after dominating the first week. He couldn't explain why the peloton was going so fast, instead of slowing some riders seemed to get stronger and faster as the tour grew longer. Usually the best in the third week, lemond exhausted himself chasing the new epo wonders around France trying desperately to compete with them. And all the time wondering what was wrong with himself. Meantime he actually posted the best time of his career up l'alpe...But way behind the new epo kings.

How about Delgado suddenly finding two minutes from 1989? Jeff Bernard finding five minutes in comparison to 1987? It is rumoured though he overcooked himself with corticos for that Ventoux TT the day before and was total crap afterwards.

Also note:
Gianni Bugno 1989
2. Delgado à 1'09"
7. Rooks à 3'04"
8. Bugno

193. Pedro Delgado _______ ESP | 42:15 | 1989

Thats about a two minute gain, in a year, after winning the Giro from start to finish 2 months before. But hey, never use your eyes watching cycling

 

[Fatclimber]

146. Greg LeMond _________ USA | 41:42 | 1991
147. Benjamin Noval ______ ESP | 41:44 | 2004

I loved that part

Or how about this famous climber:

85. Johan Bruyneel ______ BEL | 40:52 | 1995

 

[vrusimov]

Or how about this famous climber:

85. Johan Bruyneel ______ BEL | 40:52 | 1995

I have to say that I am literally flabbergasted at the insistence of some to muddle the difference between the Hinault/Lemond era and the EPO era. A careful examination of the top of D'huez records is a virtual who's who of former dopers.

 

[Alex Simmons/RST]

Here are the top 5 ascent times per year.

 

[thehog]

Or how about this famous climber:

85. Johan Bruyneel ______ BEL | 40:52 | 1995

That's an impressive time for a non climber.

He would have beaten Contador and Rolland in 2011.

 

[manafana]

had say di marchi the cannondale rider looked pocessed riding the other day.