Froome's Data
Comments
-
I think he's trying to grasp the doping nettle using analysis.
What do you mean?
He's developing models to identify the probability of performances being doped. I think it's a shame he's been written off because it's another potential tool for anti-doping.
http://veloclinic.com/about/
He's also a cyclist :-)0 -
I'm with ddraver:
pseudoscience
ˈsjuːdəʊˌsʌɪəns/Submit
noun
a collection of beliefs or practices mistakenly regarded as being based on scientific method.
If you both believe Puchowicz is pushing beliefs under the guise of science, then that's your interpretation. I don't. I think he's trying to grasp the doping nettle using analysis. We can disagree, that's fine.
Using analysis, and masquerading it as scientific methods?
so that's a yes, then?0 -
I think he's trying to grasp the doping nettle using analysis.
What do you mean?
He's developing models to identify the probability of performances being doped. I think it's a shame he's been written off because it's another potential tool for anti-doping.
http://veloclinic.com/about/
He's also a cyclist :-)
What are you talking about? We have written off his analysis of Froome's Data for reasons to do with inappropriateuse of data. In no way have we written off anything else he has said.We're in danger of confusing passion with incompetence
- @ddraver0 -
I think he's trying to grasp the doping nettle using analysis.
What do you mean?
He's developing models to identify the probability of performances being doped. I think it's a shame he's been written off because it's another potential tool for anti-doping.
http://veloclinic.com/about/
He's also a cyclist :-)
What are you talking about? We have written off his analysis of Froome's Data for reasons to do with inappropriateuse of data. In no way have we written off anything else he has said.
Er, I was answering dish_dash's question.0 -
I think he's trying to grasp the doping nettle using analysis.
What do you mean?
He's developing models to identify the probability of performances being doped. I think it's a shame he's been written off because it's another potential tool for anti-doping.
http://veloclinic.com/about/
He's also a cyclist :-)
What are you talking about? We have written off his analysis of Froome's Data for reasons to do with inappropriateuse of data. In no way have we written off anything else he has said.
And I disagree. And I think the use of 'pseudoscience' was inappropriate. Don't think we're going to progress this one much further tbh.0 -
I think he's trying to grasp the doping nettle using analysis.
What do you mean?
He's developing models to identify the probability of performances being doped. I think it's a shame he's been written off because it's another potential tool for anti-doping.
http://veloclinic.com/about/
He's also a cyclist :-)
So long as his model is recognised as another tool, rather than the be all and end all identifier... presume his models are being peer reviewed etc.0 -
.Ok - how has Dish written off everything else that Veloclinic has done by his use of the words "What do you mean?"
He hasn't - that sentence was a reference to those who have declared his recent efforts as pseudoscience. I umdestand now that you do respect Puchowicz but you take particular issue with his Froome analysis. Is that fair?0 -
I'm with ddraver:
pseudoscience
ˈsjuːdəʊˌsʌɪəns/Submit
noun
a collection of beliefs or practices mistakenly regarded as being based on scientific method.
If you both believe Puchowicz is pushing beliefs under the guise of science, then that's your interpretation. I don't. I think he's trying to grasp the doping nettle using analysis. We can disagree, that's fine.
Using analysis, and masquerading it as scientific methods?
so that's a yes, then?
Yes, can you explain the real science of his methods?0 -
I take particular exception to that article, and he appears to have Ross Tucker's problem with Twitter - in that he really doesnt come across as impartial or unbiased. I don't really see how the rest of his work on ProCycling really helps answer the overarching question of this whole subject though, i.e Is Chris Froome*, and by extension Team Sky*, doping?
(*other cyclists and teams are available but in reality the focus is purely on Froome and Sky)We're in danger of confusing passion with incompetence
- @ddraver0 -
I'm with ddraver:
pseudoscience
ˈsjuːdəʊˌsʌɪəns/Submit
noun
a collection of beliefs or practices mistakenly regarded as being based on scientific method.
If you both believe Puchowicz is pushing beliefs under the guise of science, then that's your interpretation. I don't. I think he's trying to grasp the doping nettle using analysis. We can disagree, that's fine.
Using analysis, and masquerading it as scientific methods?
so that's a yes, then?
Yes, can you explain the real science of his methods?
Sorry, but as a genuine, qualified, experienced, published sports science and medical professional I rather think the onus is on the accuser to make the case that he has written an article that is pseudoscience. I don't think that case has been well made and I can't get my head round why it is clear to you. He's used some of the tools (which are, of course, limited, with limited input data) at his disposal to question whether weight loss can explain an improvement. It may be a straw man to some but it still needs to be assessed and there is value in doing so. I can only assume he's made reference to doping because it is the elephant in the room that Swart won't/can't discuss (and Swart is right to do so, based on his involvement in the testing).0 -
I remember the day when the scales fell from my eyes about scientific papers and their accuracy...sounds like Prof hasnt had that yet (which unless you work in science there is absolutely no reason why you should have had...)We're in danger of confusing passion with incompetence
- @ddraver0 -
I take particular exception to that article, and he appears to have Ross Tucker's problem with Twitter - in that he really doesnt come across as impartial or unbiased. I don't really see how the rest of his work on ProCycling really helps answer the overarching question of this whole subject though, i.e Is Chris Froome*, and by extension Team Sky*, doping?
(*other cyclists and teams are available but in reality the focus is purely on Froome and Sky)
Having spent several years in academia and sen it myself, it's a sad truth that you often chase the money for funding. High profile stuff is good for business. As you may well know, he has taken a close look at a variety of cyclists over the years, not just Froome's/Sky.0 -
I remember the day when the scales fell from my eyes about scientific papers and their accuracy...sounds like Prof hasnt had that yet (which unless you work in science there is absolutely no reason why you should have had...)
<showing off>I have a Chemistry D.Phil</showing off>0 -
I remember the day when the scales fell from my eyes about scientific papers and their accuracy...sounds like Prof hasnt had that yet (which unless you work in science there is absolutely no reason why you should have had...)
<showing off>I have a Chemistry D.Phil</showing off>
Which college?0 -
well you ll know then, getting something published is no indication of accuracy!
Plus, it was published on tumblr, he could have just written Humpty Dumpty out a hundred times...
Indeed - I accept that. I admire academics that are pushing stuff out via social media/blogs for debate - it's the future (maybe)! Not exclusively, of course.0 -
I remember the day when the scales fell from my eyes about scientific papers and their accuracy...sounds like Prof hasnt had that yet (which unless you work in science there is absolutely no reason why you should have had...)
<showing off>I have a Chemistry D.Phil</showing off>
Which college?
Trinity. I now have nothing to do with Chemistry but it kept me busy for a few years...0 -
I remember the day when the scales fell from my eyes about scientific papers and their accuracy...sounds like Prof hasnt had that yet (which unless you work in science there is absolutely no reason why you should have had...)
<showing off>I have a Chemistry D.Phil</showing off>
Which college?
Trinity. I now have nothing to do with Chemistry but it kept me busy for a few years...
Like so many of us... 0 -
well you ll know then, getting something published is no indication of accuracy!
Plus, it was published on tumblr, he could have just written Humpty Dumpty out a hundred times...
Indeed - I accept that. I admire academics that are pushing stuff out via social media/blogs for debate - it's the future (maybe)! Not exclusively, of course.
There is still value in peer reviewed material in recognised journals. Happy to be proven wrong but I struggle to find much published by Puchowicz in such journals, particularly on his analytical model. He's a medical doctor - qualified in Pediatrics. But tbh he's not a sports scientist on par with Tucker or Swart...0 -
well you ll know then, getting something published is no indication of accuracy!
Plus, it was published on tumblr, he could have just written Humpty Dumpty out a hundred times...
Indeed - I accept that. I admire academics that are pushing stuff out via social media/blogs for debate - it's the future (maybe)! Not exclusively, of course.
There is still value in peer reviewed material in recognised journals. Happy to be proven wrong but I struggle to find much published by Puchowicz in such journals, particularly on his analytical model. He's a medical doctor - qualified in Pediatrics. But tbh he's not a sports scientist on par with Tucker or Swart...
Fair enough. As I said earlier, he has a rather idiosyncratic approach which is not to everyone's taste.0 -
He didn't develop the 'models', he took them of a post on a website by someone else - cyclismas I believe. The model attempts to get past the stumbling block of having to consider the many factors that determine climb times (particularly race circumstances) by homogenising them in to and average predicted performance. This is justified by using the false belief that everything averages out over time. As a result the model gets to compare selected individual results with favourable context to a theoretical norm devoid of context, justified by a belief that the only variable over time and races is doping. It also has further biases such as comparing a first place in a Tour stage against a model derived from top tens in all Gramd Tours.I think he's trying to grasp the doping nettle using analysis.
What do you mean?
He's developing models to identify the probability of performances being doped. I think it's a shame he's been written off because it's another potential tool for anti-doping.
http://veloclinic.com/about/
He's also a cyclist :-)
It's a statistical con trick rigged to almost guarantee 'suspicious' result at some point during a TDF.Twitter: @RichN950 -
.
He didn't develop the 'models', he took them of a post on a website by someone else - cyclismas I believe. The model attempts to get past the stumbling block of having to consider the many factors that determine climb times (particularly race circumstances) by homogenising them in to and average predicted performance. This is justified by using the false belief that everything averages out over time. As a result the model gets to compare selected individual results with favourable context to a theoretical norm devoid of context, justified by a belief that the only variable over time and races is doping. It also has further biases such as comparing a first place in a Tour stage against a model derived from top tens in all Gramd Tours.I think he's trying to grasp the doping nettle using analysis.
What do you mean?
He's developing models to identify the probability of performances being doped. I think it's a shame he's been written off because it's another potential tool for anti-doping.
http://veloclinic.com/about/
He's also a cyclist :-)
It's a statistical con trick rigged to almost guarantee 'suspicious' result at some point during a TDF.
Again, the dismissive attitude. As far as I'm aware he is trying to improve and refine them. Unfortunately it is often impossible to take account of all factors so you have to start somewhere. These things take time and may never be right. But he is trying and I think that's to be lauded.0 -
He didn't develop the 'models', he took them of a post on a website by someone else - cyclismas I believe. The model attempts to get past the stumbling block of having to consider the many factors that determine climb times (particularly race circumstances) by homogenising them in to and average predicted performance. This is justified by using the false belief that everything averages out over time. As a result the model gets to compare selected individual results with favourable context to a theoretical norm devoid of context, justified by a belief that the only variable over time and races is doping. It also has further biases such as comparing a first place in a Tour stage against a model derived from top tens in all Gramd Tours.I think he's trying to grasp the doping nettle using analysis.
What do you mean?
He's developing models to identify the probability of performances being doped. I think it's a shame he's been written off because it's another potential tool for anti-doping.
http://veloclinic.com/about/
He's also a cyclist :-)
It's a statistical con trick rigged to almost guarantee 'suspicious' result at some point during a TDF.
And throwing in a para about transfusions a la Tucker-style...
Nice.0 -
Is veloclinic the guy that used to post as a pirate? I've lost track.0
-
Is veloclinic the guy that used to post as a pirate? I've lost track.
Ooh arrr
Yes. Thats the one, Jim lad
Also has a little consulting sideline with Quintana's mentor and ex-team's management in Colombia on anti-doping0 -
I don't know whether Veloclinic is perpetrating pseudoscience, but in my view he's being intellectually dishonest. Rich has pointed out some inherent biases in the assumptions his models are built on (have these assumptions been peer reviewed, is there any attempt to mitigate them?), which qualifies at best as poor science. But further than that is the contextual framing of the work he's doing. This reveals a fundamental dishonesty regarding what conclusions his models can actually support.Warning No formatter is installed for the format0
-
.
He didn't develop the 'models', he took them of a post on a website by someone else - cyclismas I believe. The model attempts to get past the stumbling block of having to consider the many factors that determine climb times (particularly race circumstances) by homogenising them in to and average predicted performance. This is justified by using the false belief that everything averages out over time. As a result the model gets to compare selected individual results with favourable context to a theoretical norm devoid of context, justified by a belief that the only variable over time and races is doping. It also has further biases such as comparing a first place in a Tour stage against a model derived from top tens in all Gramd Tours.I think he's trying to grasp the doping nettle using analysis.
What do you mean?
He's developing models to identify the probability of performances being doped. I think it's a shame he's been written off because it's another potential tool for anti-doping.
http://veloclinic.com/about/
He's also a cyclist :-)
It's a statistical con trick rigged to almost guarantee 'suspicious' result at some point during a TDF.
Again, the dismissive attitude. As far as I'm aware he is trying to improve and refine them. Unfortunately it is often impossible to take account of all factors so you have to start somewhere. These things take time and may never be right. But he is trying and I think that's to be lauded.
His work is dismissed because it is fundamentally flawed. As RichN95 said, how can you compare the first place rider against an average of top 10s? Of course the first place rider will have a fast time/high power compared with an average. You're comparing apples with oranges.0 -
It is pseudoscience - I use the word very deliberately - because it purports to offer a scientific approach to a problem when in reality, the approach is entirely non-scientific, contains massive holes in reasoning and, in fact, does nothing to offer an explanation or solution to the hypothesis or problem.
Not offering and explanation or solution does not make something pseudoscience. He is publishing thoughts and calculations that interrogate the hypothesis that Froome's improvement can be explained by weight loss alone, which seems reasonable. You may have used the word very deliberately but you picked the wrong one.
That's a straw-man hypothesis though. The idea that there must a single defined variable that accounts for all change is entirely unscientific. Puchowitz goes on to provide a single alternative hypothesis, that the improvement is down to doping. He thus invites the conclusion that if it's not all down to weight loss then it must be down to drugs. This is shoddy in the extreme.
Tucker does similar when he reduces marginal gains to not eating Nutella and having your own pillow. No, those two things don't win you a TdF. Nobody has ever claimed they have.
With respect straw man, I'm assuming the article was in response to the quote in Esquire:
“The engine was there all along,” says Swart. “He just lost the fat.”
I'm not defending the entire article - I just find the bandying about of 'pseudoscience' as lazy and tiresome as some of the 'he must be doping because xxx' stuff, especially when it's directed at bona fide sports scientist who is trying to apply their models to real problems. It's how progress and improvements in understanding are made. We all know these models cannot take into account all racing parameters and human psychology.
To be fair to Swart, Esquire is a magazine not a scientific journal - they want a catchy hook, or soundbite, or whatever you want to call it, at the end of the article.
If you listen to what Swart has to say on the cycling podcast it's a lot more measured on that point, and he points out that there are many other factors. Certainly I found it hard the impression he's a rabid Froome defender as some on the Clinic seem to be making out. It'll also be interesting to read the proper paper.0
