Vayer

iainf72

I know I shouldn't get wound up, but why do newspapers etc give his clown attention / space? (and I know the answer)

http://www.lemonde.fr/acces-restreint/s ... _3242.html

frenchfighter

So Peraud is clean as a whistle and Nibali is down and dirty yet Peraud rode the last climb at the same time as him. Nice logic.

The most interesting thing about that article (is not as bad as his one that preceded it) is that Nibali is going to give the Tour's jersey to Pantani's mother if he wins. Maybe because he will be the first Italian to have won since Pantani.

rick_chasey

Why d'ya do it Iain, why?

RichN95.

The worst thing is that he presents Peraud as the great champion and Nibali as the doper, almost entirely on the basis that he briefly coached him a decade ago.
Nibali - 29 year old two time GT winner - bad
Peraud - 37 year old, er, 3rd place in Paris-Nice - bad.
If Peraud was British or American he'd be making all sorts of noises. (Nothing against Peraud incidentally).

He always goes on about Froome's VO2 max, but Peraud - his ultimate champion is only 85 - less than Michael Hutchinson and the same as the overweight Froome at the WCC.

He's just a fraud. Any request for raw data, mathematical methods etc is met with a request to buy his magazine (which has none of it in there - I've read it, didn't pay for it). Proper science has peer review.

What is most amazing is that he defines himself by a team which ceased to exist in 2000. Has he worked in cycling since? And that team was one of the most notorious doping teams of all time. And yet despite, according to him, being their trainer he couldn't spot they were doping but now he can with a YouTube video. How can that be? Was he ignorant or complicit back then.

But the media won't ask him this. He provides a scandal and a quote and that's what they want. He sells 'science' to non-scientists for self-interest. Snapshots of data with no context.The Andrew Wakefield of cycling.

Charles Seife, mathematician and journalism Professor, wrote a book which I have not read called Proofiness, about the BS relationship between statistics and the media. He defines Proofiness as; "the art of using bogus mathematical arguments to prove something that you know in your heart is true--even when it's not.". The opening line of the book is “If you want to get people to believe something really, really stupid, just stick a number on it.”

And Vayer caters to this.


(As an aside, I saw on another forum that Ten Dam's powermeter stats were considerably down against VAM estimates to the tune of 10%. The opinion was that he had a poor powermeter. The idea that VAM is crap was not entertained)

Coachb

+ 1 above ...why?

Will cycling ever move on? and who really cares?

frenchfighter

Good post Rich. That book sounds interesting. Currently 10 quid for a paperback is too much so will wait a bit and buy it when it is reasonable.

r0bh

RichN95 wrote:

He's just a fraud. Any request for raw data, mathematical methods etc is met with a request to buy his magazine (which has none of it in there - I've read it, didn't pay for it). Proper science has peer review.

There's a reason none of these "analyses" are peer-reviewed - they are so fundamentally flawed that any respectable journal would laugh them out of town in a couple of minutes.

RichN95.

frenchfighter wrote:

Good post Rich. That book sounds interesting. Currently 10 quid for a paperback is too much so will wait a bit and buy it when it is reasonable.

I heard him on Richard Bacon's show on Radio 5 last week. Very interesting man. He's got a new book out about the internet and why you can't trust anything on it.

It's on the Daily Bacon podcast. I'd recommend it for those who pay any consideration to such data.

frenchfighter

dsoutar

frenchfighter wrote:

For the stupid amongst the forum (blindly assuming that there's more that just me), can you explain what this is supposed to represent 'cos I can't make any sense of this

blazing_saddles

dsoutar wrote:

frenchfighter wrote:

For the stupid amongst the forum (blindly assuming that there's more that just me), can you explain what this is supposed to represent 'cos I can't make any sense of this

At a guess, Planche db Filles highlighted at the 20 minute mark, Chamrousse at about the 50 minute mark.
My reading is that Nibali nowhere near the top performer on the 20 minute climb. Half a dozen higher outputs
at up to 30 minutes.
Chamrousse, much, closer, within one performance of the top of the pops.

Oh and of course it shows that the French are clean. TJVG is clean and even more lazy than the French. :P

RichN95.

dsoutar wrote:

frenchfighter wrote:

For the stupid amongst the forum (blindly assuming that there's more that just me), can you explain what this is supposed to represent 'cos I can't make any sense of this

The beauty of graphs and charts is that the vast majority of people will either not know what they represent, not understand it or not question it as long as they can use it support their opinions. Usually you can convince a large number of people just by using colours (red=bad, green=good). Ideally you want it to represent some complex concept, preferably of your own design, which deters investigation and rigs the result.

Blazing has actually answered it for you. It's plot of climb times vs estimated power. Of course, this eliminates any context to the perfoamnces, treating them as though they are lab tests and gives no real indication as to where the data for the blue dots comes from (winners? top tens? previous performances by those riders?)

(And normalizing CdA is the biggest fraud perpetuated by these number crunchers)

frenchfighter

I should have posted along with it that Vayer should look at that as it shows little difference between Nibali, TJ and Pinot.

philbar72

It shows pretty much what we all know, that Nibali is very slightly better than the others during that time period of 20 and 60 mins.

I don’t see anything that bangs on about someone doing what the cheats did way back when….

tailwindhome

What irritates the most about Vayer, Kimmage and more recently Bassons is that anyone who points out flaws in their arguments is immediately accused of 'attacking' them.

RichN95.

TailWindHome wrote:

What irritates the most about Vayer, Kimmage and more recently Bassons is that anyone who points out flaws in their arguments is immediately accused of 'attacking' them.

You're not even allowed to point out that none of them have actually been involved in cycling since 2000 (even longer for Kimmage).

Macaloon

Worth highlighting that these power numbers are estimates. As Rich notes, all of the variables - wind, drafting, aero-efficiency, road surface (Friebe referred to tyre compounds being worth quite a few watts), etc - get lumped into the CdA fudge. It's pathetic to see people routinely ignore this.

All of the doperatti talking-points are being refuted by this Tour. Sky was a human shield: providing sad specimens protection to throw vile sh1t at people. (Shane 'Mutant' Stokes - what a dick) Perhaps their 'arguments' will now be dismantled by objective analysts. Meanwhile, I'm drowning in schadenfreude.

RichN95.

Macaloon wrote:

Worth highlighting that these power numbers are estimates. As Rich notes, all of the variables - wind, drafting, aero-efficiency, road surface (Friebe referred to tyre compounds being worth quite a few watts), etc - get lumped into the CdA fudge. It's pathetic to see people routinely ignore this.

CdA is specifically the co-efficient of drag. So it assumes that everyone has the same position with the same kit worn in the same way. Which is monumental crap.

above_the_cows

RichN95 wrote:

The worst thing is that he presents Peraud as the great champion and Nibali as the doper, almost entirely on the basis that he briefly coached him a decade ago.
Nibali - 29 year old two time GT winner - bad
Peraud - 37 year old, er, 3rd place in Paris-Nice - bad.
If Peraud was British or American he'd be making all sorts of noises. (Nothing against Peraud incidentally).

He always goes on about Froome's VO2 max, but Peraud - his ultimate champion is only 85 - less than Michael Hutchinson and the same as the overweight Froome at the WCC.

He's just a fraud. Any request for raw data, mathematical methods etc is met with a request to buy his magazine (which has none of it in there - I've read it, didn't pay for it). Proper science has peer review.

What is most amazing is that he defines himself by a team which ceased to exist in 2000. Has he worked in cycling since? And that team was one of the most notorious doping teams of all time. And yet despite, according to him, being their trainer he couldn't spot they were doping but now he can with a YouTube video. How can that be? Was he ignorant or complicit back then.

But the media won't ask him this. He provides a scandal and a quote and that's what they want. He sells 'science' to non-scientists for self-interest. Snapshots of data with no context.The Andrew Wakefield of cycling.

Charles Seife, mathematician and journalism Professor, wrote a book which I have not read called Proofiness, about the BS relationship between statistics and the media. He defines Proofiness as; "the art of using bogus mathematical arguments to prove something that you know in your heart is true--even when it's not.". The opening line of the book is “If you want to get people to believe something really, really stupid, just stick a number on it.”

And Vayer caters to this.


(As an aside, I saw on another forum that Ten Dam's powermeter stats were considerably down against VAM estimates to the tune of 10%. The opinion was that he had a poor powermeter. The idea that VAM is crap was not entertained)

Absolutely this. Interestingly I had a 30 minute conversation about the Ten Dam power meter VAM thing with the man from the bike shop yesterday when I'd only popped in for an inner-tube.

ddraver

Another "good post Rich" post...

r0bh

Fred Grappe has done some proper peer-reviewed research on the accuracy of power estimates:

http://www.researchgate.net/publication/259202034_Accuracy_of_Indirect_Estimation_of_Power_Output_From_Uphill_Performance_in_Cycling?ev=prf_pub

One of his conclusions:

The random error (+-6%) obesrved with lower winds corresponds to +-25W in professional cyclists who can sustain 30 min uphill bouts above 400W. The present study underscores that it is impossible and dishonest to make comparisons between different cyclists and to release individual values such as power output without the corresponding range of random errors.

dsoutar

RichN95 wrote:

dsoutar wrote:

frenchfighter wrote:

For the stupid amongst the forum (blindly assuming that there's more that just me), can you explain what this is supposed to represent 'cos I can't make any sense of this

The beauty of graphs and charts is that the vast majority of people will either not know what they represent, not understand it or not question it as long as they can use it support their opinions. Usually you can convince a large number of people just by using colours (red=bad, green=good). Ideally you want it to represent some complex concept, preferably of your own design, which deters investigation and rigs the result.

Blazing has actually answered it for you. It's plot of climb times vs estimated power. Of course, this eliminates any context to the perfoamnces, treating them as though they are lab tests and gives no real indication as to where the data for the blue dots comes from (winners? top tens? previous performances by those riders?)

(And normalizing CdA is the biggest fraud perpetuated by these number crunchers)

Cheers. I'd sort of worked out it was climb times vs estimated power but beyond that I couldn't work out for what. I suppose I could have been less lazy and looked at the last few climbs to see how long the leading protagonists took at come to that conclusion but I'm still lost as to what the blue dots might represent.

As usual, Vayer (if that's where the source of this is) proving that when you want to back up a b-s theory, present lots of b-s data

RichN95.

dsoutar wrote:

As usual, Vayer (if that's where the source of this is) proving that when you want to back up a b-s theory, present lots of b-s data

I don't think it is Vayer. It's got too many data points for him.

The thing with that graph is that it doesn't really tell us anything, but you could, with the correct sleight of hand, use it to illustrate just about anything.

rick_chasey

Hey Rich. Why don't you become an equity analyst? You'd earn a lot more.

RichN95.

Rick Chasey wrote:

Hey Rich. Why don't you become an equity analyst? You'd earn a lot more.

Do they let you turn up to work at 10.30?

davidof

RichN95 wrote:

(As an aside, I saw on another forum that Ten Dam's powermeter stats were considerably down against VAM estimates to the tune of 10%.

if Vayer's calcs are within 10% of power meter results I would say that is pretty good although isn't EPO only supposed to give you a 5-7% boost so even power meter data is within the margin of error.

Vayer's work could be useful for identifying Strava cheats.

iainf72

RichN95 wrote:

Rick Chasey wrote:

Hey Rich. Why don't you become an equity analyst? You'd earn a lot more.

Do they let you turn up to work at 10.30?

Sure, just deal with 'merican markets. :P

davidof

r0bh wrote:

Fred Grappe has done some proper peer-reviewed research on the accuracy of power estimates:

http://www.researchgate.net/publication/259202034_Accuracy_of_Indirect_Estimation_of_Power_Output_From_Uphill_Performance_in_Cycling?ev=prf_pub

An interesting paper. +/-6% for low wind conditions. One problem with the wind component is that a climb like Chamrousse changes direction a number of times and the wind vectors won't cancel each other out as the slope changes so will be impossible to account for in a single formula. You would have to calculate power for each leg.

The downhill section on the Chamrousse climb, albeit small, would also have affected Vayer's calculations unless he calculated the climb as two separate legs. No indication from le Monde if he did this.

Grappe spots the big elephant in the room. Pros are climbing at much higher speeds than amateur cyclists. Perhaps 20km/h compared to 13-15km/h so the wind resistance component (the hardest to calculate) will be larger for pros.

The Chamrousse climb is also below the tree line except for the final couple of kms so would tend to be sheltered from the wind - Vayer estimates it as 5km/h for the day which is probably about right but again you'd have to measure it at altitude as well as there can be a huge difference in wind speed and even direction depending on altitude.

Macaloon

Surely drafting is a big elephant too? Valverde must make it up most mountains 10% fresher than anyone else.

davidof

Macaloon wrote:

Surely drafting is a big elephant too? Valverde must make it up most mountains 10% fresher than anyone else.

at least

philbar72

davidof wrote:

Macaloon wrote:

Surely drafting is a big elephant too? Valverde must make it up most mountains 10% fresher than anyone else.

at least

His wheelsucking is amazing. Pinot the other day got really narked by him. wouldn't want him on the club run!