Greetings from Amsterdam, and apologies for the long delay in articles. I was in Leeds last week for the World Congress on Cycling Science, timed to co-incide with the buildup to the start of the 2014 Tour de France in Yorkshire. It was a stimulating conference, one which gave me some good impetus for the next three weeks (and also some great content and power output data to add to this year’s discussion)
The 2014 Tour promises to be a much more competitive and exciting race than last year’s edition, if the recent Dauphine is anything to go by. For an excellent preview of all the main contenders, check out this excellent article by Inrng.
Race analysis resumes
For my part, the race will once again be the platform to discuss the performances and their physiological implications, wherever that may lead. Of course, we already know where that is likely to lead, as it has done for the last five years that I’ve done it. You can read the history of my articles on power output, doping, performance limits at this link, which is a timeline of all the articles I’ve written during the Tours since 2009, when the issue of performance analysis as a means to detect changes in the doping culture of the sport first came up on The Science of Sport.
Every year I seem to say the same thing, and there is always misunderstanding. Last year, the analysis, which had expanded to include some great minds and methods drew the attention of The New York Times. There’s no reason to change the process now, despite the lack of understanding and resultant criticism. In particular, in the last two years, since Sky have begun to dominate the race, the criticisms flow much more freely, and seemingly no amount of explanation or caution changes the nature of discussion.
Ironically, it’s not too different from what happened in 1999 when Lance Armstrong won the race, except this time the criticisms aren’t flowing across the Atlantic from West to East. If you doubt this, read Juliet Macur’s excellent “Cycle of Lies” and document what happens on social media over the next few weeks.
Nevertheless, the process will resume this year, at least from my perspective, because I believe that it adds value, and while I certainly don’t think it’s perfect to analyze performance, I think (and I’m emboldened by those I meet at scientific conferences who work in professional cycling. Andrew Coggan is the exception. Hello Andrew) it helps to advance a necessary discussion. So here we go…
Performance analysis: Purpose, pixellation and processes
Before that, I would like to particularly draw your attention to this article, which I wrote last year just after the Tour in response to all the criticism of trying to analyze performance: Performance analysis, guided: Don’t lose your senses just because one is not perfect The main point is this: Don’t choose to be blind just because you don’t have 20-20 vision.
In effect, people who dismiss performance metrics and their implications as “worthless” because we are estimating power output are analogous to a blind man, offered 40% vision, but who refuses it because he only wants 100% sight. It’s 20/20 or nothing, and I believe that’s a flawed and narrow understanding of the world. It would be the same as a head coach saying to his analysts “If you can’t guarantee with 100% certainty what my opposition are going to do, I don’t want to know it at all”.
Before I’m seen to be proclaiming that say, 40% is “good enough”, I will say once again that we all recognize that it’s not. We want 90%, 100%. That’s why the process began with a call for the data, biological and performance. That’s because the “blurred” image offered by 40% vision, similar to the “performance pixellation” I wrote about after many stages in the 2013 Tour de France, may well lead a person into many blind alleys and unseen obstacles.
So what we need to avoid is what I last year termed “Performance pixellation”, where you look so closely at a single performance, that one ‘pixel’, and then decide what the picture is. Taking a single climb, or even a single rider, and making sweeping judgments on the plausibility of performances goes BEYOND what this method and concept will allow. I realize that some do that – I ask that you direct your questions to them. I will do my very best to remain objective, conservative and ‘big picture driven’ over the next three weeks.
To that end, I was recently interviewed by a Dutch newspaper about the analysis of the Tour, and to start this particular journey off, here are some thoughts I sent to her via email questions, which explain a little bit about the plan moving forward, so I share them below:
Q: What is the concept of analysing performances, and can we identify doping using power outputs?
Performances have physiological implications. At some point, those implications start to ask real questions about their realism or plausibility.
By analysing performance, we can get an idea of the historical evolution of the sport, and ask pointed questions about whether there are changes that are good or bad.
However we CANNOT diagnose or judge doping on the basis of performance. At least, we can’t for most performances. At some point, we can – if a rider suddenly produces a climb that is much, much higher than any physiology can predict, then I would be the first person to say “This is not possible”.
However, the power outputs we are seeing at the Tour are in the “grey area”, so you can’t say for sure yes or no.
There are some other problems. We are currently estimating power output, not measuring it, and that means error. However, I am satisfied that the error is small enough for the purposes we are using it. If someone starts using the estimations to CONVICT a person, or to JUDGE them, then it is a problem. But I for one am using it to judge trends, in large groups, and then I believe it is OK. More than that, I believe it is valuable, because if you know for instance that everyone is slowing down by 5% compared to the height of doping, then that’s worth something, it’s a currency of trust (which the sport needs).
On the other hand, if riders start to produce performances equal to, or better than anything we saw with rampant doping, then you must ask questions. Not hand down convictions, but it helps nobody to remain ignorant. The whole purpose of this exercise is to look for evidence of positive change. Not to go on witch-hunts. If there’s no evidence of positive changes, then we take it from there.
Q. What have learned from the debate during the Tour last year?
That people get very emotional! People react very strongly to even the smallest hint of doping accusation. During last year’s Tour, it was often parochialism that drove the emotion, from people in the UK who were defending Team Sky and Chris Froome. Ironically, it is the same response as we saw from the USA when Lance Armstrong was accused of doping in 1999. That is a cause for concern because it makes me wonder why we have not learned anything from the past.
That’s not to say that because Armstrong was doping then, so is Froome now. But it is crazy how people react to the mere question!
The other thing last year showed is that with the exception of Froome and Quintana, everyone in the race is slower than they have been for many years. The overall picture is one where the times are slower, the power outputs are lower. This is obviously a good thing.
However, we are still in the early days of doing this. Time will reveal the full picture, including the implications for Froome and Quintana. This year, it will be Froome and Contador, most likely, but the whole race is worth analyzing to get a bigger picture of what is different.
Q. Will you approach it differently this year?
No, same approach. Until the very top riders actually provide power output, we continue to estimate their values. What will be different this year is that we’ll have SRM files and thus power numbers for many more riders, and so thanks to the excellent work by Ammatti Pyoraily, we will be able to compare estimations to real files and increase confidence in the method, and also start to understand the context and specific situations far better. That will be the big progress in 2014.
Q. Is there an advice you would like to give to the yellow jersey?
Don’t let your family and close relationships handle your Public relations and media discussions!
Ok but seriously: Don’t get emotional. Expect that the yellow jersey comes with questions, and that most of those questions will be easy to think of as unfair.
They aren’t fair, I appreciate this, but cycling is full of people who have been burned too many times before, and so you can’t use the same answers as before to deflect attention. So stop telling us how hard you train, and that you go to altitude, and that you do innovative things like mixing pineapple juice with your water, and that you’ve eaten less Nutella and lost 1kg. That’s the kind of thing the world heard in 1999, 2000, 2001. And people have memories.
That said, I appreciate how difficult the situation is for the guy who wears yellow, whether that is Froome or Contador or Talansky or Nibali or anyone else. The person who wins the race is going to be questioned unfairly because that’s the baggage the sport carries. It’s not the media’s fault, it’s the people who doped before you, so blame them. Lead from the front. If you’re genuine and clean, then take the fight to the sport’s past, don’t try to deflect it.
And be transparent!
Who else to follow for insight
There is a group of very good people who will be providing performance data and analytics on the Tour. These include:
- The previously mentioned Ammatti Pyoraily, who within minutes of a stage finish will tell you the climbing times, and the estimated power outputs using a few different calculation methods. He then gets access to SRM files, makes those comparisons and is invaluable. He is ridiculed by the fan-boys on Twitter, but his numbers are actually good, and they form the basis for what many other people explain and interpret. Worth a follow on Twitter.
- Dr Mike Puchowicz – a medical doctor in the USA, who goes by @veloclinic on Twitter. He introduced and provided an outstanding comparative analysis of current performances relative to previous ones using a method called pVAM. Basically, the method predicts what the VAM (vertical ascent meters) should be for a climb based on factors like gradient and distance, so that climbs can be compared over the course of the race. What advanced the insight enormously in 2013 was that he added a dpVAM, which is a estimate of what a doped rider would produce based on historical performances during the known heavily doped era of cycling. Comparing the actual VAM from the race to the pVAM and the dpVAM gives an indication of whether riders are going as well as they did during the doped era, or whether performances have slowed (as you’d hope for if the sport is changing). The beauty of the method is that any climb can be compared because it has its place in the history of the performances. Look out for this analysis throughout the Tour, as well as other great technical insights from @veloclinic
- Antoine Vayer, who I realize is not everyone’s favourite (he is, after all, Dave Brailsford’s first and most famous “pseudoscientist”. The original!). Formerly with Festina, however, this is someone who has seen what we are speculating about, and knows the inside of the worst aspects of the sport. I can’t see how he can be so easily dismissed. He is irreverent, abrasive and pulls no punches, and sometimes (in my opinion) goes too far in making outright judgements about performances and doping. However, you ignore insider knowledge at your peril, and so if you fancy the extreme view (which I expect to be pulling back against this month, but that’s fine!), he’s worth following on Twitter, and also over at a new website which documents historical performances from the mountains of the Tour, called ChronosWatts (which is also on twitter).
So, let the fun begin! Ross