Performances In The Pre-EPO Era?

[jyl]

Has there been a thread discussing power, VAM, w/kg estimates for climbs in the "old days", like the 1960s, 1970s, early 1980s?

I am interested in what riders were doing back when before EPO was developed - understanding that there was amphetamines, maybe some blood transfusions, etc.

I think I'm just a bozo at using the search function. I tried searching on rider names (Gaul, Herrera, Vietto, Mercxx, Coppi, Hinault, etc) but couldn't find such a discussion. I've also tried to find data on climb times for that era, but came up short.

 

[Ferminal]

There isn't much data, check out some of the posts from le breton on Gaul etc in TTs

search.php?keywords=&terms=all&author=le+breton&fid

They were slower in races but as the tactics were so different it's hard to compare.

My view is that the old doping was probably more effective than it seems, although who really knows. I'm sure I once made a post on it but can't find it.

 

[V3R1T4S]

Here is a link to Le breton's analysis of "... arguably the most outstanding climbing performance of the past 50+ years : Bahamontes' win in the Puy-de-Dôme TT for the 15th stage of the 1959 TdF on July 10th."

I have to agree.

http://forum.cyclingnews.imdserve.com/viewtopic.php?p=1129310&sid=aea7e600bfe8e9132ee4af74c5a22af1#p1129310

 

[jyl]

That is wonderful. Thank you!

Any more like that?

 

[jens_attacks]

Check my blog climbing-records.com, you will find some performances from that era. See for example lagos de covadonga to get an idea not only alpe d'huez. Or poggio from san remo

Usually for a 40 min climb at the end of a 200 km stage, the jet fuel made them faster by 3 or 4 minutes.

For the shorter efforts( mountain time trials under one hour), the cocktails used by the pre-epo riders were super efficient imo reaching sometimes post 1993 levels( zoetemelk still has the 3rd fastest time on avoriaz, he did around 6,3 w/kg in the 1977 itt, he was caught and disqualified)

 

[Gregga]

Re:

For the shorter efforts( mountain time trials under one hour), the cocktails used by the pre-epo riders were super efficient imo reaching sometimes post 1993 levels( zoetemelk still has the 3rd fastest time on avoriaz, he did around 6,3 w/kg in the 1977 itt, he was caught and disqualified)

That's consistent with Vayer's "Not Normal" estimations : the (rare) performances above (normalized) 6 W/kg for Hinault/Lemond/Fignon (pre EPO-era...) were on mountain ITT or flat stages with one MTF. I guess for this short, partly anaerobic efforts, cortisone made a big difference, but not after 5 cols like EPO / BBs

 

[Le breton]

Re:

That is wonderful. Thank you!

Any more like that?

Can't believe that was 3 years ago! I have been busy doing other things and have not gone back to that type of activity.
Actually my house is even in more of a mess than it was then.
Maybe I should hire a tidying up coach
Nice to see that some people remember that calculation, it was a long effort.

 

[laughingcavalier]

Re:

Here is a link to Le breton's analysis of "... arguably the most outstanding climbing performance of the past 50+ years : Bahamontes' win in the Puy-de-Dôme TT for the 15th stage of the 1959 TdF on July 10th."

I have to agree.

http://forum.cyclingnews.imdserve.com/viewtopic.php?p=1129310&sid=aea7e600bfe8e9132ee4af74c5a22af1#p1129310

Based on 0 knowledge...

I find general observations like jen_attacks made really interesting. With this detailed calculation, it is a really interesting attempt, but it carries so many assumptions (honestly stated as assumptions by Le Breton - thank you) that I wonder how much we can rely on it. It would be interesting to see how variations in the assumptions might affect the outcome. For example if the figures for each factor were 10% different either way than what was assumed, how would that affect the final number? The maths for that is way beyond me! I assume if a 10% variation gives us a still believable conclusion, then the estimate has some relevance, but if a 10% variation in each factor results in an unbelievable conclusion, then we'd have to question the relevance of the estimate.

 

[Le breton]

Re: Re:

For the shorter efforts( mountain time trials under one hour), the cocktails used by the pre-epo riders were super efficient imo reaching sometimes post 1993 levels( zoetemelk still has the 3rd fastest time on avoriaz, he did around 6,3 w/kg in the 1977 itt, he was caught and disqualified)

About this climb : Rettenbachferner (Tour de Suisse 2015)
On my way back from a week-long vacation in the Dolomites I stopped in Solden and climbed it in the morning.
I was wondering whether you have the detailed profile of the climb.
http://www.climbbybike.com/profile.asp?Climbprofile=Rettenbachferner&MountainID=12290

This profile does not show the correct profile for the 6th kilometer where there is a 300m long downhill, not very steep, but still it means a loss of times and a drop in average VAM.

Congratulations for all your good work.

 

[Escarabajo]

Re: Re:

Here is a link to Le breton's analysis of "... arguably the most outstanding climbing performance of the past 50+ years : Bahamontes' win in the Puy-de-Dôme TT for the 15th stage of the 1959 TdF on July 10th."

I have to agree.

http://forum.cyclingnews.imdserve.com/viewtopic.php?p=1129310&sid=aea7e600bfe8e9132ee4af74c5a22af1#p1129310

Based on 0 knowledge...

I find general observations like jen_attacks made really interesting. With this detailed calculation, it is a really interesting attempt, but it carries so many assumptions (honestly stated as assumptions by Le Breton - thank you) that I wonder how much we can rely on it. It would be interesting to see how variations in the assumptions might affect the outcome. For example if the figures for each factor were 10% different either way than what was assumed, how would that affect the final number? The maths for that is way beyond me! I assume if a 10% variation gives us a still believable conclusion, then the estimate has some relevance, but if a 10% variation in each factor results in an unbelievable conclusion, then we'd have to question the relevance of the estimate.

I doubt that it would happen to times pre-90's. I have done that for some of the specific calculations and there is still serious differences. The fact that you have to do the same for times post 90's makes the difference almost the same.

 

[fmk_RoI]

Re:

Here is a link to Le breton's analysis of "... arguably the most outstanding climbing performance of the past 50+ years : Bahamontes' win in the Puy-de-Dôme TT for the 15th stage of the 1959 TdF on July 10th."

From that link:

According to weather archives it was a very hot day.

So not a very good day for those dosed up on amphetamines, the drug of choice of that era...

 

[Le breton]

Re: Re:

Here is a link to Le breton's analysis of "... arguably the most outstanding climbing performance of the past 50+ years : Bahamontes' win in the Puy-de-Dôme TT for the 15th stage of the 1959 TdF on July 10th."

I have to agree.

http://forum.cyclingnews.imdserve.com/viewtopic.php?p=1129310&sid=aea7e600bfe8e9132ee4af74c5a22af1#p1129310

Based on 0 knowledge...

I find general observations like jen_attacks made really interesting. With this detailed calculation, it is a really interesting attempt, but it carries so many assumptions (honestly stated as assumptions by Le Breton - thank you) that I wonder how much we can rely on it. It would be interesting to see how variations in the assumptions might affect the outcome. For example if the figures for each factor were 10% different either way than what was assumed, how would that affect the final number? The maths for that is way beyond me! I assume if a 10% variation gives us a still believable conclusion, then the estimate has some relevance, but if a 10% variation in each factor results in an unbelievable conclusion, then we'd have to question the relevance of the estimate.

I am pretty confident of the assumptions I had to make :
Gravity has not changed in the recent past, g is still about 9.81 m/s^2
The Puy-de-Dôme volcano, has been extinct for 1000's of years,and if the altitudes of the bottom and top of that climb had changed since 1959 it would have been headline news. Besides, I climbed it a few times myself since then.

So I can't change g by even 0,00001%, let alone 10%
I can't change the difference of altitudes by 10%, at most 1 or 2%
I can't change Bahamontès time by 10%, not even 0.1% . they had pretty good clock then.

All the assumptions I had to make have only a tiny influence on the final result : bike weight, Rolling resistance, profile, temperature.
If you don't understand the problem, maybe you shouldn't come in the discussion until you do!.

 

[Le breton]

Re: Re:

[quote=".....

So not a very good day for those dosed up on amphetamines, the drug of choice of that era...[/quote]

For a 40' effort amphetamines wouldn't create a problem.

 

[Alex Simmons/RST]

Re: Re:

All the assumptions I had to make have only a tiny influence on the final result : bike weight, Rolling resistance, profile, temperature.
If you don't understand the problem, maybe you shouldn't come in the discussion until you do!.

wind?

BTW - I haven't looked at the original item.

 

[Rob27172]

Re: Re:

Here is a link to Le breton's analysis of "... arguably the most outstanding climbing performance of the past 50+ years : Bahamontes' win in the Puy-de-Dôme TT for the 15th stage of the 1959 TdF on July 10th."

I have to agree.

http://forum.cyclingnews.imdserve.com/viewtopic.php?p=1129310&sid=aea7e600bfe8e9132ee4af74c5a22af1#p1129310

Based on 0 knowledge...

I find general observations like jen_attacks made really interesting. With this detailed calculation, it is a really interesting attempt, but it carries so many assumptions (honestly stated as assumptions by Le Breton - thank you) that I wonder how much we can rely on it. It would be interesting to see how variations in the assumptions might affect the outcome. For example if the figures for each factor were 10% different either way than what was assumed, how would that affect the final number? The maths for that is way beyond me! I assume if a 10% variation gives us a still believable conclusion, then the estimate has some relevance, but if a 10% variation in each factor results in an unbelievable conclusion, then we'd have to question the relevance of the estimate.

I am pretty confident of the assumptions I had to make :
Gravity has not changed in the recent past, g is still about 9.81 m/s^2
The Puy-de-Dôme volcano, has been extinct for 1000's of years,and if the altitudes of the bottom and top of that climb had changed since 1959 it would have been headline news. Besides, I climbed it a few times myself since then.

So I can't change g by even 0,00001%, let alone 10%
I can't change the difference of altitudes by 10%, at most 1 or 2%
I can't change Bahamontès time by 10%, not even 0.1% . they had pretty good clock then.

All the assumptions I had to make have only a tiny influence on the final result : bike weight, Rolling resistance, profile, temperature.
If you don't understand the problem, maybe you shouldn't come in the discussion until you do!.

Le Breton

I can understand your comments - All I would say is that reading the laughing cavalier post he is only questioning the mathematical and statistical variance of the assumptions
He did give credit for the math and acknowledged his inferior understanding of the requirements to do the calculations.
He was quite rightly asking what assumptions had been made and if we changed those and ran simulations with different assumptions what would that do?
So effectively if you put your climbing model into a monte carlo simulation what are the potential 20,000 or 100,000 variable answers and then what would your median or mean be and what is the standard deviation and where would your results be on the distribution.
It is actually not a bad question to ask and I really don't think his intention was to have a go at your work.

Your reply though does come across as a little defensive and abrupt
And anyone who knows me can tell you I know that when i see it as it is my normal MO

Cheers
Rob

 

[Le breton]

Re: Re:

Alex Simmons/RST[/url]"][quote="
wind?

BTW - I haven't looked at the original item.[/quote]

I don't remember all the details, I know that I looked at all the information I could find about the weather that day, including pictures (looking for flags).
The first part of the race, to about an altitude of 900 m. could be influenced by wind.
From ~900 to 1415 m, you essentially spiral around the mountain, consequently only updrafts would have to be considered.
In the end I didn't include wind, encouraged by the fact that the calculated powers were quite close to each other for the first and the second parts of the race (for Bahamontes).

 

[Le breton]

Re: Re:

[
So effectively if you put your climbing model into a monte carlo simulation what are the potential 20,000 or 100,000 variable answers and then what would your median or mean be and what is the standard deviation and where would your results be on the distribution.
It is actually not a bad question to ask and I really don't think his intention was to have a go at your work.

Your reply though does come across as a little defensive and abrupt
And anyone who knows me can tell you I know that when i see it as it is my normal MO

Cheers
Rob[/quote]
Hi Rob
My intention was to be abrupt.

Now, I bet you will agree with me that we don't need a Monte-Carlo program to do a multiplication or a division.

There is nothing stochastic in nature in the necessary calculations, at least in the parts related to fighting gravity or Rolling resistance.

Now, concerning Wind ( Alex' point) that could be a nice project.
Be my guest.
Remember that wind impacts only about 10% of the final answer in terms of watts/kg.
Puy-de-Dôme is not in Patagonia

 

[vedrafjord]

Re: Re:

I can understand your comments - All I would say is that reading the laughing cavalier post he is only questioning the mathematical and statistical variance of the assumptions
He did give credit for the math and acknowledged his inferior understanding of the requirements to do the calculations.
He was quite rightly asking what assumptions had been made and if we changed those and ran simulations with different assumptions what would that do?
So effectively if you put your climbing model into a monte carlo simulation what are the potential 20,000 or 100,000 variable answers and then what would your median or mean be and what is the standard deviation and where would your results be on the distribution.
It is actually not a bad question to ask and I really don't think his intention was to have a go at your work.

Your reply though does come across as a little defensive and abrupt

I have some sympathy for Le breton because laughingcavalier may have been asking in good faith, but I see drive-by commenting here and on Twitter all the time trashing people who are trying to model this stuff and implying that they've forgotten something obvious, or that because a certain variable can't be known exactly, the results are nonsense or (cough) 'pseudo-science'.

I'm a risk systems developer for a reinsurance company. We insure insurance companies against big stuff like hurricanes, earthquakes, hailstones that can destroy a year's crops, as well as human stuff like medical malpractice. We make extensive use of Monte Carlo simulations to determine the likelihood of, say, a hurricane following a certain path and causing floods and wind damage in certain US counties.

I would absolutely kill for the margin of error of the guys estimating watts on mountain stages. By far the most important variable is a fixed number (gravity) that we know to a very high degree of accuracy. The other variables can be estimated too. From comparison of estimates and data posted by riders like Pinot and Ten Dam, we know that the estimates are accurate to within a couple of %. Additionally, if we average many climbs, the errors will tend to be averaged out too, and we're getting towards being able to make specific conclusions about the strength of different riders.