Froome's SRM data on Ventoux

[Dear Wiggo]

Re: Re:

Given the pretty crummy analysis and ill informed commentary when data is made public, I'm not particularly surprised at people's decision to keep it private.

Confirmation bias is the order of the day.

This man has a point

I just want to enjoy the cycling.

Then what the bleedin' hell are you doing in the Clinic.

 

[neineinei]

Anders K Christiansen (one of the journalists blacklisted by Hushovd for asking him too many questions about Armstrong) interviewed Vayer today: http://www.vg.no/sport/sykkel/sykkel/jeg-har-ikke-hacket-team-sky/a/23488674/

Vayer says it's a leak from the team (and that it was Lisbeth Salander who hacked Sky's computer, a joke). He also says the video will be up on youtube again tonight.

Vayer says the video was pulled from youtube by the uploader, because Sky's lawyers gave him cold feet.

 

[vedrafjord]

Re:

2) On the flip side, the power data does seem enormously useful for competitors. As in, you need to try to withstand a x watt effort for y minutes to beat Froome, thereafter, his FTP does seem to go down a fair tick. That W balance analysis (not heard of it before), if legit, could be even more so.

Surely if we can estimate his power data to a high degree of accuracy, then other teams (who will also have tons of power data from their own riders) can also accurately estimate it, and have already done so. I think it's naive to think that other teams haven't already calculated how many watts Froome can put out at various points and how many watts their own riders have to produce to beat him.

 

[Dear Wiggo]

Re:

A couple notes (I didn't get a chance to watch the video but did see Vayer's graph):

1) As has been said before, the power data isn't a smoking gun. 5.8w/kg avg, above threshold efforts of almost 7 w/kg is what everyone expected.

lolnope it is not. Vayer predicted exactly what the file said. Try again.

2) On the flip side, the power data does seem enormously useful for competitors. As in, you need to try to withstand a x watt effort for y minutes to beat Froome, thereafter, his FTP does seem to go down a fair tick. That W balance analysis (not heard of it before), if legit, could be even more so.

lolnope. You train the best you can do have the best endurance possible. The route is released a year in advance, the power figures needed to win are known for the past 10+ years. There is no competitive value in the data whatsoever. Try again.

3) A high max HR isn't a indication of athletic performance. A low HR is. The vuelta data would seem to confirm that. As I understand it a low max and resting HR would seemingly indicate a very enlarged heart. I imagine Froome's engine must look a bit like Secretariat's.

lolnope. There are different combinations at every level of the sport. Lance's resting heart rate was 32bpm and his max was 201bpm.

 

[WillemS]

Re: Re:

Surely if we can estimate his power data to a high degree of accuracy, then other teams (who will also have tons of power data from their own riders) can also accurately estimate it, and have already done so. I think it's naive to think that other teams haven't already calculated how many watts Froome can put out at various points and how many watts their own riders have to produce to beat him.

I agree with you, but only partially.

Most estimations I've seen work with average wattages over relatively long intervals, but don't really work well for short intervals or bursts of maximal output. It would be interesting to know how Froome recovers from an attack, if there's a time period after a burst in which he will be unable to respond to counter attacks and so on. Careful analysis of behavioural data (just measured speed against time) already provided some riders with the non-responding, maintaining tempo tactic to move back up to Froome after his attack. Maybe more detailed information of the time just prior to or after attacks could hone tactics.

 

[The Hitch]

Listening to Boardman rattle of Brailsfraud talking point after Brailsfraud talking point in a 2 minute speech, I'm losing my lasts doubts that his world beating performances in the age of Indurain, were less than a coincidence.

 

[vedrafjord]

Re:

I think people are missing the really interesting thing here. It's not whether this data shows doping or not. It's that this data matches almost exactly the w/kg calculations of Vayer and Ferrari(?), even though Brailsford dismissed them as peseudo-scientific rubbish.

Exactly, there are two views on statistics. My background is in machine learning so I'll call them the Chomsky view and the Google view, as described here http://norvig.com/chomsky.html. The Chomsky/Brailsford view is that you have to perfectly model every conceivable factor with real physics etc and the model has to mirror reality. The Google/ammattipyoraily view is that the predictive accuracy of the model is what matters and not the internals.

My view is that the Chomsky way leads to whataboutery and nothing ever being done. If you have a model and it has very high accuracy despite leaving out certain variables, I think it's obvious that those variables aren't important.

 

[The Hitch]

Also hillarious to see an organization (ITV) that regularly seeks out the opinions of 3 time loser and fraud David Millar, dismiss Vayer as just some random "blogger"

 

[JimmyFingers]

Re: Re:

So we have data and the numbers are reasonable, hmm, good I'm almost glad of the back, it's reassuring.

As for Sky's reaction, easily demonised but if the data has been stolen and it affords their rivals insights into Froome's ability, why on earth would they be ok about it? Anyway glad it's out there, the legendary Ventoux alien attack exposed as being human.

I'm glad you've finally become a believer of pseudo-science jimmy! welcome onboard.

Hopefully this means that when vetooo or Vayer says Froome does something superhuman you will believe that as well?

This is his attack on the Ventoux, when has he ever been more super-human? Or as revealed by the leak, human?

 

[JimmyFingers]

Re:

Listening to Boardman rattle of Brailsfraud talking point after Brailsfraud talking point in a 2 minute speech, I'm losing my lasts doubts that his world beating performances in the age of Indurain, were less than a coincidence.

Of course you are

 

[sniper]

Re: Re:

So we have data and the numbers are reasonable, hmm, good I'm almost glad of the back, it's reassuring.

As for Sky's reaction, easily demonised but if the data has been stolen and it affords their rivals insights into Froome's ability, why on earth would they be ok about it? Anyway glad it's out there, the legendary Ventoux alien attack exposed as being human.

I'm glad you've finally become a believer of pseudo-science jimmy! welcome onboard.

Hopefully this means that when vetooo or Vayer says Froome does something superhuman you will believe that as well?

This is his attack on the Ventoux, when has he ever been more super-human? Or as revealed by the leak, human?

he has been just a tadbit stronger than pinot over the past few years though has he.

 

[WillemS]

Re: Re:

2) On the flip side, the power data does seem enormously useful for competitors. As in, you need to try to withstand a x watt effort for y minutes to beat Froome, thereafter, his FTP does seem to go down a fair tick. That W balance analysis (not heard of it before), if legit, could be even more so.

lolnope. You train the best you can do have the best endurance possible. The route is released a year in advance, the power figures needed to win are known for the past 10+ years. There is no competitive value in the data whatsoever. Try again.

So why do riders bother with short burst of maximal power to distance their main rivals if all that counts is sustained power? Your story does not match most, if not all, MTFs.

3) A high max HR isn't a indication of athletic performance. A low HR is. The vuelta data would seem to confirm that. As I understand it a low max and resting HR would seemingly indicate a very enlarged heart. I imagine Froome's engine must look a bit like Secretariat's.

lolnope. There are different combinations at every level of the sport. Lance's resting heart rate was 32bpm and his max was 201bpm.

I agree with you there, a low max HR does not tell us much, it's certainly no indication of a high level of fitness or performance, as those are seen with high and low max HRs. In my informal cycling group there are both riders with a high max HR (200+) and relatively low max HR (~150, probably due to age [60+]) who perform relatively well on climbs, especially corrected for age.

 

[the sceptic]

Re: Re:

So we have data and the numbers are reasonable, hmm, good I'm almost glad of the back, it's reassuring.

As for Sky's reaction, easily demonised but if the data has been stolen and it affords their rivals insights into Froome's ability, why on earth would they be ok about it? Anyway glad it's out there, the legendary Ventoux alien attack exposed as being human.

I'm glad you've finally become a believer of pseudo-science jimmy! welcome onboard.

Hopefully this means that when vetooo or Vayer says Froome does something superhuman you will believe that as well?

This is his attack on the Ventoux, when has he ever been more super-human? Or as revealed by the leak, human?

he has been just a tadbit stronger than pinot over the past few years though has he.

yes, it all makes sense now. That's why Pinot has always been nearly as good as Dawg.

 

[The Hitch]

Re: Re:

Listening to Boardman rattle of Brailsfraud talking point after Brailsfraud talking point in a 2 minute speech, I'm losing my lasts doubts that his world beating performances in the age of Indurain, were less than a coincidence.

Of course you are

I thought you said you preferred just focusing on the racing and all the utopianistic fantasies they bring you, rather than wasting time here. What's wrong, can't help it?

 

[the sceptic]

Interesting that Vayer says the files came from a leak not a hack. Sky lying once again, they must be really worried about this stuff.

They can scare away oufeh from twitter, but the likes of Vayer aren't gonna stop here.

 

[the sceptic]

Re: Re:

Listening to Boardman rattle of Brailsfraud talking point after Brailsfraud talking point in a 2 minute speech, I'm losing my lasts doubts that his world beating performances in the age of Indurain, were less than a coincidence.

Of course you are

I thought you said you preferred just focusing on the racing and all the utopianistic fantasies they bring you, rather than wasting time here. What's wrong, can't help it?

If only a place existed, a place where true cycling connaisseurs could muse over the beauty of cycling, the hard work of these fine gentlemen, a place where you are not allowed to talk about doping, a place where there are no conspiracy thories and jealous losers talking crap about your hero. A place that Jimmyfingers might call home.

What a shame no such place exists on the entire internet. It would be a win-win for everyone.

 

[King Boonen]

Re: Re:

I think people are missing the really interesting thing here. It's not whether this data shows doping or not. It's that this data matches almost exactly the w/kg calculations of Vayer and Ferrari(?), even though Brailsford dismissed them as peseudo-scientific rubbish.

Exactly, there are two views on statistics. My background is in machine learning so I'll call them the Chomsky view and the Google view, as described here http://norvig.com/chomsky.html. The Chomsky/Brailsford view is that you have to perfectly model every conceivable factor with real physics etc and the model has to mirror reality. The Google/ammattipyoraily view is that the predictive accuracy of the model is what matters and not the internals.

My view is that the Chomsky way leads to whataboutery and nothing ever being done. If you have a model and it has very high accuracy despite leaving out certain variables, I think it's obvious that those variables aren't important.

I'm very interested in machine learning on my current role!

I've never heard the two views described in that way (name-wise) but I fully subscribe to the Google model. My work has been in looking at non-invasive predictive and diagnostic testing which invariably involves looking at something in blood or urine and using this to develop some sort of test or model. The idea that such a model should be 100% accurate, specific, sensitive and reproducible is ridiculous. You build your model, test it, refine it, validate it, give your accuracy, specificity, sensitivity and reproducibility and then assess if it is fit for purpose. False negatives are a no-no, false positive should be kept low but are much more acceptable.

I do not know nearly enough about this kind of data, modelling or how Vayer and Ferrari have approached it (and others) so I'll always be cautious, but this has certainly made me pay a lot more attention to what they are doing and give it more credence in my assessments.

 

[vedrafjord]

Exactly. "All models are wrong, but some are useful" - George Box.

 

[Dear Wiggo]

Re: Re:

2) On the flip side, the power data does seem enormously useful for competitors. As in, you need to try to withstand a x watt effort for y minutes to beat Froome, thereafter, his FTP does seem to go down a fair tick. That W balance analysis (not heard of it before), if legit, could be even more so.

lolnope. You train the best you can do have the best endurance possible. The route is released a year in advance, the power figures needed to win are known for the past 10+ years. There is no competitive value in the data whatsoever. Try again.

So why do riders bother with short burst of maximal power to distance their main rivals if all that counts is sustained power? Your story does not match most, if not all, MTFs.

You're either a diesel or a turbo. You can train that to some extent but mentally as much as anything you will lean towards one or the other.

Curiously, the average power by the top is pretty much the same, and over 30 minutes, your normalised power won't be much different because the attacks are not all that frequent, even by the most alien of riders.

So you have a basic formula of weight, ascent and time. Regardless of how you do it, your average power being the best it can be for X km (known 12 months in advance) is the best you can produce. Knowing what someone else can produce makes not one whit of difference.

 

[BradCantona]

Re: Re:

Given the pretty crummy analysis and ill informed commentary when data is made public, I'm not particularly surprised at people's decision to keep it private.

Confirmation bias is the order of the day.

This man has a point

I just want to enjoy the cycling.

Then what the bleedin' hell are you doing in the Clinic.

I want to... but there's no point pretending. I'd just prefer if we nailed our cheats based on unequivocal evidence, rather than interpretations of data whichno one can bloody agree on

 

[BigMac]

Re:

Interesting that Vayer says the files came from a leak not a hack. Sky lying once again, they must be really worried about this stuff.

They can scare away oufeh from twitter, but the likes of Vayer aren't gonna stop here.

They didn't just scare him away, they actually got his account nuked.

 

[therhodeo]

I'm most interested in who leaked it.

 

[the sceptic]

Re: Re:

Interesting that Vayer says the files came from a leak not a hack. Sky lying once again, they must be really worried about this stuff.

They can scare away oufeh from twitter, but the likes of Vayer aren't gonna stop here.

They didn't just scare him away, they actually got his account nuked.

I thought he nuked it himself when sky started threatening with lawyers.

But anyways, the video is still out there so I doubt it will go away any time soon.

 

[hrotha]

Re: Re:

I'd just prefer if we nailed our cheats based on unequivocal evidence, rather than interpretations of data whichno one can bloody agree on

Have no fear. As has been pointed out countless times, the Clinic is not a court. We can't nail any cheats. We can only use our brains.

 

[BigMac]

Re:

I'm most interested in who leaked it.

Twitter suggests Wiggins didn't leave the Sky empty handed.

Seriously though, I take it not everyone has access to those things and that they are relatively well hidden.