Share and Share Alike?

Tom Dumoulin has shared his power data from the Vuelta a España. Dutch newspaper Algemeen Dagblad has a summary of the rider’s performances on various climbs during the race.

Power is measured by strain gauges, often on the cranks, and logged on a bike computer. It’s a handy way to measure effort and in recent years has become a lively topic, in part because more and more riders are using power data making the topic more accessible and also because high numbers are linked to doping and recent editions of the Tour de France have seen controversy and accusations flying although curiously this hornet’s nest only seems to get poked in July.

Power data is seen as absolute but be careful as even the best power meters on the market are sold with an accuracy of +/- 1% and in reality this can be greater with big variations in temperature because the strain gauges are sensitive to temperature. So when it comes to a mountain stage the numbers can vary significantly. Whether it’s a racer putting data into a newspaper or an amateur checking on their training the best use is comparisons over time rather than one set of numbers.

Anecdotes and data

Why release the data? It doesn’t prove a rider is clean but it does suggest transparency. There seem to be three audiences for the data:

  • it provides real world data for sports scientists to evaluate what happens in a bike race
  • there’s the geek tragedy of fans pouring over the numbers to see the high score and perhaps make comparisons, either with their own numbers or with others. They’re sensitive to the marketing of the data and this validates a manufacturer like Pioneer or SRM
  • the largest audience has to be the wider public who are more likely to think “whats per kilo?” than W/kg but they’ll take the transparency as symbolism and see a rider putting their data online as a useful step and take expert opinion

This is probably why Dumoulin’s data went to AD.nl and in 2013 Team Sky didn’t put Chris Froome’s data online to be crowdsourced but instead L’Equipe got FDJ coach Fred Grappe to review and pronounce.

Nothing to hide? As for cleanliness, transparency is good but not proof of cleanliness. You can’t prove a negative but publishing is always of interest and coming out and publishing the numbers looks good, it is inviting judgement rather than trying to run from it. However just remember that if a rider has made the lifestyle choice of spending years covertly taking microdoses of EPO and draining off miniature blood bags all to manipulate the bio-passport then doctoring a data file is child’s play compared to trying to swim through the holes of the bio-passport. In other words we can get stuck on the circular arguments of faith in the numbers as much faith in a rider and their entourage.

There’s now an asymmetry because once someone shares data it puts the spotlight on those who don’t. Are there reasons not to share? Yes, there’s the embarrassing matter of inaccurate or incomplete power data which would show up some sponsors in a bad light. There is also fear of giving data away to rivals who could find weak points. If a rider cracks above nW/kg then all a rival has to do is crank the pace to n+0.1 W/kg and they’ve got them on the ropes: simple? Well this isn’t Pro Cycling Manager, it’s not so easy in real life.

How to beat Froome: When L’Equipe published “Le Dossier Froome” there was so much noise about Froome, doping and credibility that the underlying data were almost ignored. For Fred Grappe saw a weakness in Froome where he displayed excellent power for 20 minutes but the longer the duration the more the power fell. Normal? Yes but the slope of Froome’s power curve apparently fell away faster than others. So one way to beat Froome is to stick it to him on a 40-60 minute climb… which is what Nairo Quintana did on Alpe d’Huez this year. Easier said than done.

However nobody needs power meter publications to find these weaknesses, a stopwatch and and a keen eye will often do. Some scorn “armchair analysis” but it can be rather accurate. Certainly a clever team coach can combine their stopwatch plus data and anecdotes from their own riders on the wind direction and road surface to explore more. So it all means that if you wanted to use power data to hunt for weakpoints then you don’t need to wait for a rider to put their data online.

Training sessions are more private. Dumoulin has made his Vuelta numbers public but not his training data nor his time trial stage numbers. To put everything online could allow people to see training sessions which is more sensitive. If you knew a rider was doing long climbs at full speed and was a big diesel then you might prefer to attack them with waspish accelerations. But many riders do share training in full on Strava.

The real benefit comes from analysis over time in that if a rider suddenly makes a big jump in power to weight then this can be questioned and explored. It could even be an alarm bell for the anti-doping agencies to investigate and perhaps retest but it’s only an alarm bell, it’s not evidence in itself.

Conclusion
Tom Dumoulin’s shared his Vuelta data. One lesson seems to be that this year’s edition was relatively slow one, at least when measured by effort. There’s a wider view with more and more riders sharing race data and making wattages a topic of conversation and post-race analysis, often with the subtext of transparency. Given amateurs can use formulas to make estimations that are often as accurate as a power meter releasing power data shouldn’t be a big deal, it won’t give away too many insider secrets, especially if it’s kept to race data. When relayed by trusted newspaper the data can reassure the wider public and it’s also good marketing for power meter manufacturers.

75 thoughts on “Share and Share Alike?”

  1. Excellent summary and nice word play. For the amateur rider the advent of power meters has added another great level of interest to the post-ride analysis aspect of cycling. Apart from the immediate numbers hovering on the screen at you during a ride to push you to kill yourself on the bike, for amateurs and pros alike, as INRNG says, the key is comparison of saved data over time. For many years I used to calculate an average power output per ride using heart rate and calories counted; it was rubbish but that did not matter so long as the method was consistent so that the comparison remained valid. I agree that if estimated power output disproportionately increased it must have been because I had taken to riding in hurricane conditions every ride or someone had put something in my baked beans. Since the general introduction of power meters, all one does now is log more accurately failing power through age. Published pro data does indicate just how good these guys are. In that way transparency adds to the mystique. But anyone who draws too much from occasionally published power data from a race is nuts.

    • I heartily agree on your age comment.
      Power meters should not be sold to anyone under 35.
      Thereafter it should be a case of listening to your body and riding accordingly.
      But most importantly, enjoying it !

      • i’m comfortably over 35… 🙁 i recently started riding “outside” again (and was reminded how much better it was than riding “inside” and never going anywhere)…

        when riding outside, i’m too occupied by cars, potholes, dogs, random crap flying across the road, etc., to even think abut using a power meter… living in a highly populated area, my primary concern is “survival” when i’m on the road… i’m generally riding to “get somewhere” (like work), and since i’m not treating it as “exercise”, i usually am well below my limits…

        when riding my bike that doesn’t go anywhere though, i always use a HRM and the power numbers* on the bike… that is “exercise”, and i use it primarily for tracking progress…

        * which, i am forced to admit, are pitiful… 🙁 it would take 10 of me to make one tom…

    • “Comparing his numbers with what Warren Barguil [Giant-Alpecin] did last year in the Vuelta a España, the level looked higher last year,” team coach, Adriaan Helmantel told Cycling Weekly.

      “You can see that Joaquím Rodríguez, Alejandro Valverde, and Nairo Quintana are not performing at the same level as they did in the Tour de France. The combination of things was good for Tom.”
      cyclingweekly.co.uk

      • Loved your ‘geek tragedy’ comment Inrng !
        I don’t own a power meter, though I can see the benefits of their use, and I do take some interest out of the data – perhaps put me in the ‘minor geek tragedy’ category 🙂

        But what stood out for me was his final Morcuera data on stage 20. This was the climb where he ‘cracked’ and yet he still put out 5.9 w / kg, comparable to many other climbs throughout the race.
        Now what would be interesting would be to compare what Aru and Landa’s comparable figures were…

        • 5.9 W/kg was TD’s average power for the last part of Morcuera. That last part took him some 16 minutes. I did not see him crack there though. He lost maybe 20 seconds to the Aru group at the top of that climb, so he just was not able to go with their attack. During Astanas attack Aru and Landa were surely pushing close to 7 W/kg and TD was surely also above 6,5 W/kg but he lacked maybe 20 W to close the gap.
          In the following descent you could see that Aru and Landa were clearly paying for that big effort as Dumoulin managed to claw back time although he was also very tired and exhausted. And it was only when the (kind of) rested LLS and Andre Seitz joined Aru coming from the front of the race that TD’s game was over.

          • I don’t think he “cracked”, but I suspect he burned himself out a little bit more than what was opportune.
            On the descent he was getting in touch, but his moment should have been before, on the long and nearly flat section on the top of the Morcuera, where he let Nieve lead the pace (!) since he appeared totally worn out.
            I wonder if it exists any given optimal watt segment which would have allowed him to lose some more ground on the climb, say 30-some secs instead of about 16″, but coming out stronger and able to get back time on the flat section and not just downhill.

      • Thanks for the insight, well done.

        It would be nice to know the ambient temperature superimposed over the PM data.
        Temp would certainly influence physiological output. Difficult to compare one year to another?

        “Easier said then done”

  2. Excellent analysis of the situation. I am in favour of data being made public, the more riders that are prepared to do so will further isolate those who refuse.

    I very much agree that the experienced eye can pick up many of the nuances which require careful extrapolation from bucket loads of data. This makes the reluctance of some to publish data all the more difficult to understand.

  3. As the article points out, this is largely a PR exercise. If the data were going to prove anything ‘bad’ about a rider, they wouldn’t be released or, at least, not as they are. It’s all about ‘being seen’ to be transparent.
    The media desperately wants to keep the doping issue to the forefront (ironically, having ignored it for decades), because it sells their drivel.
    Nothing any rider does is going to prove anything – nor will it stop the media casting aspersions.

    • Yes, I would agree to a large extent.
      And Inrng’s point about potentially ‘doctoring’ data is a salient one too.

      I know that you don’t like to talk of nationalistic bias JE, but the comparison with Froome in this article is relevant ; it was largely the French media that was the driver on the doping accusations.
      Do you suppose that if Dumoulin had held on during that 20th stage and won La Vuelta, there wouldn’t have been similar aspersions cast by the Italian and Spanish media ?

      It’s become a sad fact that athletic excellence and improvement draws so much suspicion.
      If Dumoulin were riding for Sky there would have probably been an uproar already, sad to say.

      • I agree completely – and posted on this article saying that UK riders have been targeted by the French media: nationalistic bias absolutely isn’t only limited to the British (it’s just that I tend to hear the British side more often).
        But the comment was deleted.
        The accusations against Wiggins were particularly absurd – his victory could scarcely have been less explosive or more inkeeping with his previous riding.

        • Your earlier comment got sucked up by the spam filter, probably because you’re flooding the comments with too much, too soon. But you do post a lot of Sky-centric comments, I want this piece to be about the broader idea of data sharing rather tracing everything back to your regular British/Sky view.

          • It’s odd because I’m so often accused on these pages of being anti-Sky and anti-British.

            I disagree that my comments are often Sky-centric – those are the ones that are most often responded to, because (as others have noted) people tend to take umbrage the moment someone on this site criticises any British or Sky rider.

            I think those making accusations often have to take a look at their own motivations/biases.

            I certainly don’t have a ‘regular British/Sky view’ – don’t know where you’re getting that from.

            (Also, that was the first comment I’d posted today.)

    • To see how not to do it, look at how British long distance runner Paula Radcliffe handled the data leak. First refusing to put her data out under some very flimsy logic and then releasing it after the negative impression has already been imprinted on the casual fan and her reputation tarnished.

      It’s fat better to be on the front foot and manage the messaging with data (like Dunmoulin) than to be seen to be bullied into putting out data and still have accusations of massaged data and appeasement (Sky).

      • Actually, Paula only released the “top line” number, then compared it to other athletes. That’s not how off-scores work.

        If her biography is any indicator at all, Seppelt’s data is very detailed. Ashenden or Parisotto could have put this to rest months ago if she permitted them to discuss her values in the press. Funny how that didn’t happen.

        • Actually, Radcliffe didn’t release the numbers at all. They were leaked. She then gave her view of what they meant, but didn’t provide any more information.

          And there is a difference between the personal data being discussed in Radcliffe’s case (details of the characteristics of her blood) and the power data being discussed here. My own view is that privacy in medical affairs is a more important issue than confidence in the integrity of sports, so I can sympathise with Radcliffe (or other athletes) not wanting to disclose this information. Power readings are different, particularly as they can effectively be calculated by people watching the race with a stopwatch.

  4. I find it completely pointless. A professional cyclist knocked out mind bending levels of wattage for a ridiculous length of time, over and over again, for 3 weeks. That’s nothing new and doesn’t prove or disprove anything. If Dumoulin did exactly the same rides last year and put out way less power that might be suspicious, but we don’t know. Watts prove nothing other than the performance level of the rider (which is incredible, which we already knew). It’s how he got to that level that is important, whether he did it legally or not. Did he take weight loss drugs, does he drink a cocktale of opiate based painkillers during training, did he sit on top of a mountain with a blood bag on each arm munching EPO? We’ll never know from looking at that.

    • From a sports science perspective, it has great value.
      Varying a rider’s weight and measuring power is a core training tool, particularly for someone like Dumoulin who has to balance such gains against a potential off-set on his TT performance.

      For us punters, as you say, it’s probably only becomes meaningful over time as comparable data records can be built up and looked at.
      But, regardless, if it’s a way of interaction and engagement between the pro’s and the wider audience, that’s a good thing also surely ?

    • The reality is, in some sense, cycling is a very simple sport. Having a great big engine gets one very far in the sport. This is similar to distance running. Yes, there are 100’s of other things that have to go right, but, physically, it is very simple.

      Because the sport of cycling is relatively simple, power profiles can be indicative of doping. Combine them with blood scores over time and one gets a very complete picture of the cleanliness of an athlete.

      Unfortunately, the UCI still feels that zero transparency is best. It’s run more like the WWE with “good guys” and “bad guys.”

      I agree that a single set of power data isn’t very useful. More performance data shared frequently would raise the credibility of the sport.

    • Same here. ZZZZzzzz. F1’s the sport for fans of science – they try to take the human/driver element out of the equation as much as possible…and look how boring F1’s become. I hope pro cycling doesn’t end up there, but the likes of Velon and others seem hell-bent on this direction. I remember ol’ Heinie’s excuse for doping in pro cycling, some baloney about how it had to be tolerated because the public would not enjoy a TdF run off at an average speed of 25 kph. Who cares how fast they go? I want to see RACING, a struggle of human vs human rather than the fruits of sport scientists and pharmaceuticals.

  5. If anything this kind of thing works in favour of Dumoulin by intimidating rivals. They might look at that and think ‘Christ, how am I going to beat him if he can do that?!’, a bit like a boxer releasing footage of him in training hammering the hell out of the heavy bag.

  6. My view, and I might be wrong. Having watched Dumoulin carefully during the Vuelta, his performance was almost to be expected, given the results he obtained in past performances this season. It was exciting to watch, and ended in the way many who are clean and do not manage their efforts find to their cost. That he performed at the level he did, as a one man team, spoke eloquently on the doping issue in interviews and then released some of his performance data should be welcomed by all who argue for transparency.

    Data release is only a part of what is required from riders and teams.

    I

    • I don’t think it was, how could I say?, *unnatural* – yet, “almost to be expected” is maybe a bit too far, IMHO. What past performances are you thinking about?

      • Gabriele:

        Leaving out prologue and tt wins, which are clearly important in most three week stage races.

        Mountains jersey 2013 Tour a Andalucía. Points classification 2014 Eneco Tour. 2nd Tour of Edmonton 2014. 4th Santos Tour down under 2015. 3rd Tour de Suisse 2015. Just a selection of recent performances which indicated that Dumoulin was much more than just a time trialist.

        I think Dumoulin only took up bike riding in 2011, so fairly constant and rapid progress.

        • And from those results you would ‘almost expect’ a rider to be challenging right at the end of a grand tour?
          Even the TDU and TdS results don’t suggest that – they were won by Rohan Dennis and Simon Spilak.

          • J Evans: No, not just from those few results, they were a selection illustrating Dumoulin’s range of qualities above just testing. – I do know the race winners, thank you. I don’t intend to bore posters by giving a race result history for Dumoulin. If his sports directors were not surprised by his performance – and they were not, who am I to disagree. Of course performing in a three week race is very different to one week events.

            People like Dumoulin, Landa and might I add Thomas have that indefinable asset that most of us mere mortals lack called class. No one can really know for certain where these riders limits until are, until they are tested in the appropriate environment. If you know the race history of these riders, you should not be surprised they have performed well in three week races in 2015. It’s of course not a given, but watching the potential of such riders makes for interesting observations.

          • His palmares is good, but nothing suggested his Vuelta performance – and no-one was saying any of this without the benefit of hindsight.
            As for his directors, they were so unsurprised by his performance that they brought precisely no-one to support him.

  7. Be careful with the temperature information => obviously every serious power meter has an auto offset in order to adapt to temperatures changes during the ride!
    You could not imagine without such a feature having something correct in such a (common) situation: starting with a cold 8°C temperature in the morning and finish your ride 5h later in the heat at 1600m altitude

    • J. Evans My final word.
      No one can predict a GT performance with any certainty before the event, there are simply too many variables.
      Of course no one was claiming Dumoulin would definitely perform as well as he did in the Vuelta. That doesn’t mean his performance was a total surprise given his recent record. You are simply wrong to say his DSs were not aware of his potential, they certainly were and stated this publicly. Dumoulin is in a team which is structured around sprinters, not GT winners and their only other real climber had already ridden the Tour. No one predicted the performances of Landa or Thomas either, but many were well aware of their potential to perform in three week events.

      If you have valid points to make fine, but comments just for the sake of an argument are not helpful to anyone, especially when they bear little relationship to the actuality.

      • You’re changing history.
        Originally, you said ‘almost to be expected’.
        Now, you’re saying ‘doesn’t mean his performance was a total surprise’.
        Those are very different statements.

        All I have said is that it was not ‘almost to be expected’. A valid point.
        I never argued that his ‘performance was a total surprise’ – because you’ve only introduced this now. (What was that you said about ‘especially when they bear little relationship to the actuality’? Seems like perhaps you shouldn’t be bandying about accusations.)

        Also, if G-A were so sure of his GC prospects, why had they a) Not given him a contract beyond 2016 and b) Not hired anyone to back him up – not even at the end of this season?

        A lot of people claiming that they knew all along, but without any evidence to back that up.

  8. If the idea would be to create ‘power passports’ in the future with this sort of data in the future, would someone like Juan José Cobo be banned for an against the odds/suspicious/brilliant/out of the ordinary (delete as appropriate) performance? Also would someone be able to ‘sandbag’ in races to confuse rivals if they’re regularly performing under the standard that they show in target races? Wouldn’t that get them banned?

    • And, speaking of that Vuelta, what about Froome? You’d have to ban him or whatever you imagine, say erasing him from the GC?
      Only to give him back that result once he proves, some months later, that he was worth such a performance.
      But, wait wait… neither that 2012 Tour had been announced by comparable performances, since nor the Vuelta nor any race he rode in-between had him show that level of performance.
      The ‘true Froome’ appears in 2013… and even then, some of his wattages were probably unprecedented.
      I think it’s quite obvious that the whole watt-passport doesn’t make much sense. Sport is about optimising your results and also getting more watts along the way. How do you know when should a rider stop getting better? Is some court of inquiry what evolution is natural? And if a rider limits itself to block the well-known and above cited effects of age?
      As inrng says, at most an alarm bell – but I’m a bit skeptical also about that.

  9. Observe how much Dumoulin’s data appear to vary in different race condition.
    We must really think about what happened in the race – and use a bit of imagination – to explain why that (in an absolutely reasonable way, I’m not hinting at anything like doping).
    Why was Fuente del Chivo worse than Sotres? How could Ermita del Alba be so good? And how does La Quesera work, exactly?
    The numbers look like they go crazily up and down, if we don’t include some sort of more or less obvious technical explanation. We Ancient Geeks might try to provide some – haruspicy, or the likes.
    Consider that the difference between riders in terms of W/kg is usually way inferior to the difference between Dumoulin’s own performances through the race!
    Yeah, we can solve the easy questions: the shorter the time, the higher the W/kg ratio you can afford. Fine.
    But given that such relation is not linear, how can we judge if those variations make sense? One could argue that Cumbre del Sol was too fast when compared with Caminito del Rey… imagine when you need to do that through difference races!
    The difference needed to put time on rivals in a mountain stage is often significantly lower than 0.2 W/kg… Dumoulin’s performances cover, for different reasons, a range that is ten time that difference. Too many factors, too hard to keep them under control.
    Cross checking with other riders? Give a look to the ITT. We can say more or less who was performing better than expected and who wasn’t, but it’s all so highly hypothetical – and very relative.

    Besides, the shape of the power curve through time is rider-specific (in fact, it defines *what kind* of rider are you), but same goes for the effect of previous altitude gain in the same stage, number of previous stages, number of previous races during the season… these are just the simplest factors which change your numbers *a lot* – and, even more important, the proportion by which anyone is affected varies hugely from rider to rider. Go and compare that all!
    In a *carbon copy* Vuelta only three among Tom’s watt results were pretty similar between them: Alpujarra, Fuente del Chivo y Quesera – three very different durations, stage developments and results.

  10. I’m frustrated in how the cycling media continue to tell the world that adding a bit more “geek” to cycling coverage will ruin it for fans that don’t understand.

    More than other sports the fanbase of cycling is people who actually ride. Maybe that is because fans don’t appreciate the work and tactics of cycling unless they closely understand what it is like to be in the draft of a group or on a 10% pitch. Not the same for other major sports. There are a thousand bars here in NYC filled with baseball fans who haven’t touched a bat in 20 years.

    Maybe adding more “geek” is just the thing the networks need to bring in non-riding fans. Currently fans can’t empathize with the efforts they are seeing. They don’t need to know exactly what 300 watts feels like, but (just as you said) they only need the comparison. “Ruben Plaza has held 300 watts for 5 hours, while Aru and company in chase have averaged only 215 so far while in the peloton”.

    On board cameras and live wattage? If cycling had the budgets and the power of networks of major league sports in US, I would think there would not only be live wattage, speed, GPS… but even breakdowns of % of work done by teams or % of work done by individuals in the breakaway. This allows the commentators to paint a picture of the drama on the road. The numbers can help make heroes out of hard working domestiques in real time and show the villains of a breakaway who are avoiding their turns at the front.

    Not that Inrng did exactly, but those who claim the “geek factor” will ruin the romance of cycling are missing the point . It’s simply the foundation of facts, which should then be translated (by good journalists) into great storytelling and drama. Without it, cycling coverage will continue to be very pretty aerial pictures of Europe.

    • I had to laugh when I read “This allows the commentators to paint a picture of the drama on the road” and then thought of TV’s Heckel and Jeckel or the clown on British Eurosport. Your ideas might work well, but I’m less than thrilled with the thought of making pro cycling interesting to the fans of NFL, NBA, MLB, etc. with all kinds of statistics, comparisons, etc. that encourages “fantasy” sports leagues and such.

      • Larry, for yesterday’s British Eurosport coverage of the women’s TT they had Dr Hutch (cycling weekly and books Faster and The Hour).
        He is a TT specialist so was able to actually add insight, discuss power data, power to weight versus power to drag, and equipment choices. It may be a bit geeky for some but I thought it was excellent and really added to the TT experience. (He had also thoroughly researched his notes on previous performance etc).
        I would accept you don’t want this on every stage, and like you and others, I often just want to see racing and not worry about any stats.
        My point is really that science and technology can be useful to broadcasting but it needs the right people to present it and above all else it still needs good riders in exciting races.

          • Good points Bilmo and Leo.
            I’ve just watched the Elite Mens TT on red button BBC.
            Dr Hutch and Chris Boardman both in favour of releasing all power data, certainly for an in-race experience. And they’re both capable as presenters of making some sense of it for the viewer.
            I think it would definitely add something to the TT viewing experience.

    • Re fans who don’t ride: I am purely a spectator, and I have managed to learn about the importance of the draft and impact of the terrain from being told about it, and I’m sure I’m not unique.

      The TDF coverage this year did have onscreen graphics about who had been on the front for what percentage of time in the breakaway, which was at times interesting. It didn’t get into the work being done in the peloton, which I agree is an area it would be good to show more of.
      However, go too far down the on screen graphics route and you risk obscuring the screen with heaps of numbers, which are a little distracting. They also may annoy the people who do tune in for the pretty aerial pictures of Europe (which is a substantial proportion of the audience, as discussed at length previously so I won’t go into it again here).
      I do think it would be good to have that information available through mobile apps for people who are interested in lots of detail, and then for the commentators to include some information in their coverage for people who only want some detail, supplemented by some limited on screen graphics.

    • “More than other sports the fanbase of cycling is people who actually ride”
      That is only true in the countries where cycling is not a mainstream sport. And it is true to some degree for any sport that is not mainstream. Here in NL I know plenty of pro cycling fans that don’t ride any bike except for means of transport.

      • Exactly, many of the hundreds of thousands who watch in the US, UK, Australia etc are active riders. Most of the millions who watch in France, Italy, Spain, Belgium etc are not. The sport can accommodate both, especially with digital TV and multi-channels which can serve different audiences.

  11. Transparency is a an important attribute that can help us become more comfortable with performances.

    Strava – has become the dominant site for sharing cycle rides – it also offers interactions with us mortals.

    Here is a non-scientific selective look at the number of team members sharing this year’s races/training on Strava (excluding Trainees)

    Etixx Quickstep – 11
    Katusha – 0
    Astana – 4
    Giant Alpecin – 13
    Cannondale Garmin – 15

    One could do a analysis of all the pro riders and find that some teams seem to encourage transparency or inversely discourage it.

    There are riders like Kwiatowski and Laurens Ten Dam that posted pretty much all their rides and races this year – including power data, and have tons of followers. There are others that post their training rides – and others that post their races with times, speed cadence – but with out the power data.

    It is likely we will see more and more from many teams… although I would expect some riders to not share all their data.

    If the Yates brothers, Degencolb, Pinot, Barguil, Pinot, Dombroski and Boswell are…. perhaps the younger generation has a different view going forward.

    • Doping doesn’t work like “I don’t train then I dope and suddendly I’m a great rider” nor “I’m training, here are my *normal* values, but when I go racing I start to dope and my performances go up”.
      People used to dope mainly to train more and better, sometimes they even doped while training and not during races (which, as we know, isn’t meant to say that *in general* people didn’t dope for races or during races). You couldn’t tell a rider is doping by his training or racing data only: an expert trainer could, with a whole set of data (cardio included) – which isn’t generally available. It isn’t that hard to change the shape of your training just not uploading an hard session which will maybe appear like a rest day…
      I feel that Strava is great to, as you say, connect with fans, prompt curiosity and the likes, but… transparency? And inference about a team or a rider’s attitude based on his Strava presence? This is highly questionable.

    • Undetectable modifying of power meter data is simple. Just change the slope and/or zero offset value of your SRM file. Takes 2 seconds.

      That digital epo method must be idiotic to leave such obvious traces of tampering.

      One of the wealthiest teams in the sport uses a single sided power meter. lol Good luck relying on that data for anything other than for general training intensity guidance (which is its limited intended use).

      At end of the day, power guesstimates (or even real power meter numbers) are not going to raise anti-doping flags any more than reading race result sheets will. The targets are already pretty obvious.

      Not once in all the years of polemics of “using power as an indicator of doping” has anyone ever (not even the self appointed poseur of reading power tea leaves himself, A. Vayer) come up with an example of a rider to target that would not already be a target by simple observation of the fact that:
      (a) they are a pro and/or
      (b) their race results have changed and/or
      (c) they are winning / nearly winning.

      The only valid argument is one of providing the appearance of transparency. Can’t say I’m overly convinced by it but perhaps it’s for the mass media audience and less so for those that actually understand power data.

  12. A funny case is when riders make data public for transparency’s sake, assuming nobody will be able to see anything suspicious there (or being cynical enough to know that even if you see something strange, it won’t be enough to be sort of a proof, hence the mere *transparency* effect will bring part of the fans on your side, whatever the experts may suggest).
    A couple of friend with medical expertise were quite surprised by Horner’s blood data from the Vuelta. They were stranger than Kreuziger’s (yeah, I suppose that some elaborate mathematical models explain why what looks strange isn’t at all and the other way around).
    Lance had similar troubles, I think.

  13. About doctoring files: I have one foot in another sport were performances are nowadays assessed by gps. The files are encripted (unlike ‘Strava’ and powermeter files which are editable ascii), but that is the easy part. After all, who can say that you did not swap your logger with someone else? So you end up with a veritable ‘chain of custody’ where official observers ascertain that your logger was indeed with you throughout your performance. Cumbersome but necessary, and the same applies to any other use of tracking or logging devices to ‘authenticate’ performances.

  14. The most interesting thing about power data being made available is that those that clamour for it the loudest in July – such as Antoine Vayer – pay absolutely no attention to it when it is actually released.

    The general response to Dumoulin’s data – that of a man who was a stage away from winning the Vuelta – has basically been ‘Meh’. This should be a goldmine of information. There has been more using Dumoulin’s action to bash others for not doing similar than there has looking at the actual data. So why have those who were so desperate for it from Froome in July so indifferent to Dumoulin in September?

    The answer is that they were never interested in the data in the first place – just the concept of it. They just wanted it as vehicle for manufacturing suspicion and garnering self-publicity. And that’s why riders would be advised to keep data to themselves.

    • “(…) those that clamour for it the loudest in July – such as Antoine Vayer – pay absolutely no attention to it when it is actually released.”
      How can you tell?

      “(…) This should be a goldmine of information. (…) So why have those who were so desperate for it from Froome in July so indifferent to Dumoulin in September?”
      Again, purely personal assumptions.

      “(…) They just wanted it as vehicle for manufacturing suspicion and garnering self-publicity.”
      This is your opinion. It’s by far not a fact.

      Your are talking as if you knew “them” (Vayer?) – show us some facts or have at least the humility to state that this how your are feeling or thinking right now.

        • He did bring up some of his “radars” (French for speed camera by the way) during the Vuelta but the race just attracts a lot less attention, it also had few climbs that have been used before so nobody could make climbing time comparisons to Armstrong/Pantani etc.

  15. Rich (above) is right. Vayer and his ilk use their Twitter feeds, blogs (and also in Vayer’s case) paid gig with Le Monde, non-stop, when it suits them.

    But re Dumoulin’s data…*tumbleweed*

  16. Another funny thing is that the real interest of power data lies right elsewhere from what I’ve see published by Dumoulin (don’t know if there was anything more).
    Why is power data better than climbing times? Well, because they’re authomatically downloaded and not calculated climb by climb (with exact distance, time and altitude references which are often complicated to obtain), and because you can include flat or rolling sections, too!
    The interest of *looking at the final/decisive climb* is understandable but limited: the real difference, at least for me, unleashing my Geek nature, would be to be able to look at the whole stage development. Watts on the previous climbs, average watts on flat and rolling sections… that kind of things (note that I’ll always favour watching the race over watching the digits).
    But the leading model, supported by Vayer and others, is just to look at peak performances. I remember the debate about Contador 2007 which became boring or even annoying for many reader, but which showed quite well how broad must be the scope to understand a race.

    Although for some riders it’s like that, putting out one impressive single performance is not the only way to win (nor the only way to cheat, for what it’s worth).

    PS In Tom’s data there’s an inconsistency in the top W/kg numbers. It looks like he’s weighing some 2 kg more on that occasion, and I doubt it’s possible, even if there had been three easy stages before, since he had already lost them the day after. Top W/kg should read 7.3

    • I understand the point Gabriele, but is whole race data necessary ?
      There is no evidence that I am aware of that substantiates the fact that the peleton are much quicker these days than in the past. Individual stages are not quicker than in days of yore are they ?
      On the contrary, as Inrng points out, this year’s race was slower than last year’s.

      So if we assume that the GC contender’s efforts are on a par with each other, and their strength being sapped at a comparable pace, then the peak performance is the element where possible *false* strength manifests itself.
      This is assuming that the rider has been safely hidden behind their rolleurs and indeed perhaps behind a puncheur on the lower slopes (still drafting).
      So it is all about the peak performance, no ?

      • Despite the “cheat” parenthesis, which was rather accessory, I wasn’t referring mainly to doping, but to a general, ‘sport science’, interest in the data.

        Nevertheless, since most of the people interested in data to discover something about doping are delighted by comparisons across the years, to do something like that properly you need to understand how the race was raced.
        A final average stage speed can be the result of climbing faster, or racing faster on the flat, or both, or racing slowly at the beginning than fast at the end… they’re all different situations which affect differently different riders and may explain (or not) the resulting performance.

        What you seem to imply about comparisons between riders in the same race isn’t really *that* relevant in doping terms: if nobody is doping, someone will probably climb faster (with better W/kg) than the others, won’t he?
        In that sense, climbing times really tell you all you need to know (mechanical doping is another story, but anyone using a motor wouldn’t have troubles fixing the watts before releasing them).
        A single data set like Andorra’s final climb would tell us that Aru had proportionally way better W/kg than Dumoulin, comparing with their relative gap in Sotres.
        Was Aru doping in Andorra? Was Tom in Sotres? ^___^
        Is it all a ‘bad day/good day’ matter? If we had the data of all the stages, we could confirm (or not) that a rider regularly fades less than the competitors in given circumstances, hence any coherent performance would be ‘normal’ (unless he was always doping, but data just detect anomaly), whereas if someone put out an impressive performance in a situation that usually doesn’t favour him, that would be more suspect.

      • I was forgotting – to avoid misunderstanding, it’s better if I reiterate that I don’t believe you can identify dopers only nor mainly through data. Too many variables you can’t grasp.
        People who think they can, usually revel in hindsight. That is, their benchmark is built (afterwards) on people who had been caught – but the ‘scientists’ have no idea about what percentage of their performance was doping related. Besides, they tend to highlight performances which were already suspect – to say the least – for an expert eye… the rest is borderline, yellow lights and the such. Trying to sort out a norm from a bunch of way-over-the standard individuals.
        This perspective is ultimately reinforcing the utterly wrong ‘individual cheater’ model, whereas the doping problem is really such, and really big, when it’s a systemic one – team doping, complicit institutions.

Comments are closed.