Welcome to the Science of Sport, where we bring you the second, third, and fourth level of analysis you will not find anywhere else.

Be it doping in sport, hot topics like Caster Semenya or Oscar Pistorius, or the dehydration myth, we try to translate the science behind sports and sports performance.

Consider a donation if you like what you see here!

Did you know?
We published The Runner's Body in May 2009. With an average 4.4/5 stars on Amazon.com, it has been receiving positive reviews from runners and non-runners alike.

Available for the Kindle and also in the traditional paper back. It will make a great gift for the runners you know, and helps support our work here on The Science of Sport.

Wednesday, July 30, 2008

The Olympics on the horizon

Dust settles on le Tour, and the Olympic Games loom on the horizon

Well, it's now 9 days to go before the world's sporting attention turns to Beijing, and the 2008 Olympic Games. As we've tried to do all year, we'll be keeping a close eye on proceedings both on and off the track/pool/fields to bring some insight and analysis of the Games.

You can read all about Beijing's pollution, prospects of doping at the Games, Speedo and Arena's swimsuit wars, and other previews of the Olympics in our "Beijing Build-up" tab above. From now on though, with the Games so close, we'll switch to a specific, theme-by-theme preview, and try to spend the next week or so looking at issues that will be relevant to performance in Beijing.

They will almost certainly include doping, with much going on in the battle against drugs in sport, so I expect a great deal of coverage of this controversial topic. The swimsuit war will finally come to a head, when Speedo's LZR meets Arena's Powerskin Revolution Mach-2 in the pools, and the result should be a slew of world records. We're obviously particularly fond of the endurance events, and so the battles between Kenya and Ethiopia on the track should deliver some fireworks. Expect Ethiopia to dominate, winning at least 8 out of the 12 medals in the men's and women's 5000 and 10000m events (and all the golds).

As for the marathon, can Martin Lel, the greatest marathon runner in the world at the moment, deliver Kenya it's first gold in that event? If he fails, there's a better than good chance that it's another Kenyan who beats him anyway, with Cheruiyot and Wanjiru particularly posing the greatest challenge. We'll look at all these events and more in the weeks to come, and we'll post those articles at this link for those who want to jump straight there.

The world of sport will be going into something of a "hiding" phase for the next week as most of the teams will be taking their athletes to pre-Olympic training camps. Last night, for example, saw the last major track and field meeting before the Olympics, and many of the great athletes were not there, having finished their preparations a week or so earlier. So actually knowing who is taking what into the Games will be difficult, but we'll use the "gap" to build up to what will hopefully be the greatest Olympics yet!

Cycling: The Tour is over, looking back

The Tour de France finished on Sunday, and not surprisingly, it was a procession into Paris that saw Carlos Sastre become Spain's third consecutive champion. The Spaniard won the race with his solo attack on Alp d'Huez, and then held on by riding a great time-trial over 53km on Saturday.

The one man who was favoured to beat Sastre (experts gave him a 60-40 chance, or more) was Cadel Evans, who simply didn't have the legs in the final ride to bridge the 1:34 gap. Whether it was Sastre who rode out of his skin or Evans having a bad day, I guess we'll never know, unless we can get hold of the power output data from the two riders in both their Tour time-trials.

However, interesting observations from Wayne, one of our readers, suggests that if we use Stefan Shumacher as a "benchmark" (since he won both Tour time-trials this year, incredible performance), then it appears that Evans did slide off his performance in that second time-trial.

Looking at the performance, it turns out that Evans was 1.26% slower than Shumacher in the first, 29km time-trial, and fell to 3.26% slower in the final time-trial. Sastre, on the other hand, was 4.8% slower than Shumacher in TT1 and improved to 4.02% slower in TT2. Sastre was therefore better in TT2 (relative to Shumacher), but not by a huge amount, whereas Evans was considerably worse. Bernard Kohl, who surprised all with his final time-trial (including us), improved from 5% slower to 3.7% slower.

There are of course important considerations here, like the fact that the second time-trial, being hillier and longer, might favour guys like Sastre and Kohl, but it's interesting to note that if Evans had even kept the same gap between himself and Shumacher (1.26%), then he would have beaten Sastre by 1:44 and claimed yellow...but it wasn't so much the spectacular time-trialling of Sastre that held the yellow, it was a combination of his moderate improvement and Evans' fatigue after three weeks of hard racing that may have denied him yellow. If we could just get the power data...but thanks Wayne for that insight!

As for the overall summary of the race, that stage on Alp d'Huez was crucial, and I really can't help feeling that if Evans hadn't been caught up by the CSC tactics behind Sastre, and just ridden his own climb, the result would have been different. I am quite sure that by himself, Evans could have ridden the climb in 40:30, rather than the 41:45 he did it in. Had he done this, he'd have gone into the final TT with a 34 second deficit and who knows...?

Ifs and buts get sportspeople nowhere of course, and so we're not suggesting anything should have changed - the strongest man in the final week won the Tour in the end.

The 2008 Tour: A turnaround for cycling?

As for the Tour, it's being hailed as "turnaround" for cycling. Apparently viewership figures were up, spectatorship was higher, and of course, the authorities feel that they are getting to grips with the doping problem. The big feather in the cap of the doping authorities was that they were able to catch guys using a third-generation EPO, which those riders must have felt was not detectable. For once, it turned out, the testers were a step ahead (or level with) the dopers, and that can only send a positive message to all sports.

Four positive tests were eventually returned, and it's been hailed as a success because that is fewer than in previous years (and at least we didn't lose the yellow jersey in the cloud of controversy this time around!). I'm not sure I agree. To me, the number of positive tests tells you nothing about the state of the doping battle, because one could just as easily return not a single positive test because the dopers are that far ahead in the race. So I wouldn't say it's time for high fives and pats on the back just yet - for every MICERA, 3rd generation EPO that can now be caught, there may be 5 drugs that can't, and so the fight must be intensified.

Whether the race was clean this year is difficult to say. Certainly, the days of dominant riders going off the front are over. The jostling for position, the small time gaps, the brutal racing all suggest less doping, which would narrow the physiological gaps. The days of single attacks producing time-gaps of minutes between riders are now over. Sastre succeeded because of the tactical play behind him, but no other attack was able to create that much time in this year's race, which is interesting. So I think it's a step in the right direction, but wouldn't get too carried away just yet!

Thanks for joining us in our coverage of the 2008 Tour, and let's hope Beijing brings the same kind of action!



Andrew said...

Great stuff - I'm really looking forward to your Olympic coverage! Three questions on the TdF:

1) I read that the final TT course was full of twists and turns through small towns. Would the frequent speed changes cater to the strengths of a climber?

2) Stage 10 saw Evans fall victim to cat-and-mouse games with Menchov on Hautacam. Do you think CSC learned their stage 17 tactics from that episode?

Andrew said...

Rats, I forgot the 3rd question:

3) As a fan of team GC, did Vandevelde really have a chance to win? A tactical error on stage 10 and a fall on stage 16 appear to have cost him ~4 minutes...

(Sorry about my own "tactical errors" here!)

nickgavey said...

You say that lack of dominant cyclist in the Tour is evident of less doping. No argument from me on that front.

The question that is rather left begging is how does one explain Lance Armstrong's seven victories? Like most cycling fans, I would love definitive proof that Lance was clean for his entire career. His battle with cancer makes for one of the most inspirational sports stories of all time.

So what is your opinion - is it possible for a vlean athlete to be so dominant for so long over a field that certainly contained athletes who were doped?

Larry said...

I frequently read your blog. Two scientists, writing openly and honestly about doping! It’s terrific, especially for someone like me. I am a non-scientist who has spent an insane amount of my free time trying to understand some of the science of doping, and of anti-doping.

I understand that not everything you post here is intended to stand up to scientific scrutiny. Sometimes, you’re just posting your thoughts and musings like any other reasonably intelligent sports fan might do. And in most cases, you’re careful to indicate when you’re just shooting the breeze, and when you’re giving us scientific information. However, I don’t think you were always careful to draw this distinction in your recent Tour coverage.

Let me point to one example. You stated that “[t]he jostling for position, the small time gaps, the brutal racing all suggest less doping, which would narrow the physiological gaps.” THAT sounds like a statement intended to express a morsel of science, but you haven’t backed up this statement with anything resembling scientific proof, and I don’t think there’s anything in the science that could back up your statement.

What makes this worse is that you’re repeating an oft-stated canard, that when someone performs in a sport like cycling in a way that is extraordinary or exceptional (the word in common use today is “unbelievable”), then the performance is probably fueled by dope. For example, we’ve heard a lot of nonsense that Floyd Landis’ 2006 Tour de France Stage 17 victory was fueled by dope (witness Dick Pound’s famous statement about how he looked like he was on a motorcycle). Putting aside the issue of how Landis’ alleged doping with testosterone could possible propel him up mountains, there is the obvious issue of what Landis’ competitors might have been taking.

If doping is common in the peloton, then I would presume that doping alone is not the explanation for why one cyclist would excel above all others. Looking at the question from the simplest point of view, if they’re all doping, then they’re all getting whatever performance-enhancement is available from doping, and the differences in rider performance must be due to other factors. Yes, it is possible that some riders dope more effectively than others, so perhaps the argument goes that the most effective dopers are the ones who finish first and the least effective dopers are the ones who finish last. Of course, we have no data whatsoever to support the idea that there are wide variations in the effectiveness of the doping programs used by different doping cyclists, and given the highly competitive nature of the sport, it seems unlikely that a rider would dope in a careless or indifferent manner. But to explain a rider’s extraordinary performance as being fueled by dope, one would have to assume that this rider is able to dope more effectively than any of his peers, and it’s difficult to explain how this could be.

It is also possible that some humans benefit physiologically more from doping than others. I’ve seen nothing in the scientific studies to indicate that this is the case, but it’s at least possible in theory. Of course, it’s possible that the people who would benefit most from doping are not the very top athletes in a given sport. It may be that doping would be most helpful to riders in the middle of the pack. If you take seriously the law of diminishing returns, it might be the case that the very top riders have already done so much to reach the top, that further efforts at improvement can only have small effects … while the guys in the middle of the pack might have more room for improvement and could benefit more from doping. Of course, this is rank speculation … but I see no reason why doping might not serve to “narrow the physiological gaps”, to quote you.

In any event, your statement quoted above does not betray any serious consideration of this issue. Maybe I’m failing to understand, but you seem to be saying that the variations in rider performance within a doping peloton would be greater than in a non-doping peloton. Why would that be the case? If you’re considering average performance over a reasonably large group, you would not expect that a particular performance-enhancer made available to the entire group would result in greater performance differences between individual members of the group. Has this happened in golf or tennis, where we’ve seen dramatic improvements in equipment over the last 20 years? Have you seen sharper distinctions in track-and-field performances (some of which can be measured with considerable scientific accuracy), now that doping products are available to these athletes?

Please, either show me some science here, or back off and state that you’re just shooting the breeze, like any of us might do around the internet.

Ross Tucker and Jonathan Dugas said...

Hi Andrew

Good questions, none are especially obvious to answer, but here goes:

1) Not sure. I didn't get the impression there were a great deal of changes in the rider's rhythm watching it on television. That would be the only basis for an advantage to the climbers. The problem for a climber in a time-trial is that because they're so small (usually), it's not as simple to find the absolute power to accelerate if they have to slow down in the twisty sections. On the climbs, it's another story, because relative power counts for a lot. So I suspect it makes little difference in a time-trial, actually.

2) Could well be. Here in SA, there was some decent Tour coverage, and reading on the internet also kind of suggested that Evans has this reputation as a bit of a follower. That pattern emerged in the weeks leading up to the Tour, and he did little on that Hautacam stage to convince anyone otherwise. So I guess his reputation as a rider who doesn't really take the initiative was all the incentive CSC needed to mess the rhythm around and see what happened. Turns out they were 100% correct.

I think Evans is a guy who really needs a strong team. A lot of people wrote in and said "watch Evans when he gets a strong squad". And they're right, he'll be better, but he could still have won this Tour simply by seizing half chances and not allowing others to dictate. I think CSC played this perfectly.

3) Very difficult to say. I don't think he was ever a realistic winner of the race, because he would have needed a spectacular performance somewhere along the way to bridge gaps that were created by the first TT and the Hautacam stage, tactical error notwithstanding. That spectacular performance is ultimately what denied a guy like Kohl the win too, even Evans, one might say, and so I'm not sure that Vandevelde was ever a likely winner. That said, of course he had a chance, because in his first big Tour in overall contention, he gave an excellent account of himself, and the two big time losses are things that one might excuse in a first-time challenger.

So perhaps 2008 was his apprenticeship, and he'll come back with more nous and experience next time, and the same form, and avoid those errors. I just got the impression that this year, he was there as a contender, but not a guy who carried that lethal, "knock-out" punch which he will need to really be a threat. Perhaps next year that will be part of his armour...

That's all speculation though, who knows?

Thanks for the questions, and let's hope Beijing is equally interesting!


Barrld said...

Don't count Ryan Hall out on the marathon podium!


Anonymous said...

"if Evans had even kept the same gap between himself and Shumacher (1.26%), then he would have beaten Sastre by 1:44 and claimed yellow..."

You would also need to know if Schumacher improved his time over TT1. One of the cycling forums did have kph comparisons for Evans and Sastre.

Anonymous said...

"For example, we’ve heard a lot of nonsense that Floyd Landis’ 2006 Tour de France Stage 17 victory was fueled by dope . . . Putting aside the issue of how Landis’ alleged doping with testosterone could possible propel him up mountains, there is the obvious issue of what Landis’ competitors might have been taking."

I've seen a study that suggests EPO raises androgen levels. I've never seen it mentioned by anyone, but could be a clue.

I picked Landis' ride as suspicious because, from memory, the day before he had bonked big time. That sort of recovery is not usually possible -- for anyone.

Guess who else I picked early on as sus last year . . .? Team didn't get invited this year.


Larry said...

Aussie, I'm trying to get Ross and Jonathan to talk about whether a doping peloton should exhibit greater or smaller variations in individual rider performance than a clean peloton ... so I don't want to get too distracted.

But yes, I've heard the argument that Landis could not have naturally recovered to have had such a strong stage 17, after "bonking" in stage 16 ... ergo Landis must have doped after stage 16 to aid in his recovery. I'm not a scientist, but it's at least possible that something like this could have taken place. You could imagine a situation where a cyclist might be using small amounts of dope during a race, then bonked, then used a little extra of the doping products to aid in recovery.

There are problems with this argument. First, I've seen nothing in the science to suggest that a top athlete could not naturally recover from a bonk in a short period of time (say, overnight). I've learned to be suspicious of any explanation of human physiology that sounds too simple ... but the simple explanation of a bonk is that during an athletic performance, a person can use up his stored glycogen, forcing a switch-over to stored fat and blood sugar to derive the energy needed to complete the athletic performance. If this is all there is to a bonk, then once the performance is over, the athlete can eat the right foods to restore his glycogen and accomplish a rapid recovery. I've read that the period right after the performance has ended is supposed to be a good time to do this. OK, I acknowledge that the athlete's performance during the bonk might create a kind of shock to his system that would itself require recovery, and maybe dope could help with this recovery. Of course, there might be more than one kind of bonk, with different recovery rates attendant to each type of bonk.

Let's look at the arguments on the other side. First, we commonly see athletes have bad days followed by good days, in cycling and in other sports. So, an overnight recovery should not be an automatic cause for suspicion. Second, even for an athlete who has not bonked, there is little evidence to suggest that doping products like testosterone offer any performance-enhancement during the course of a race. So, even if Landis might have wanted something to aid his recovery post stage 16, it's not clear that testosterone (or for that matter any other doping product) would have done the trick. Sure, I acknowledge the performance-enhancing potential of drugs like EPO, but how would EPO help an athlete recover from a bonk? There's nothing I've seen to suggest that a bonk depletes an athlete's red blood cells. Also, I don't believe that EPO works overnight.

There is one substantial problem with your argument that we haven't discussed yet. Let's say that you're right, and that the only way that Landis could have recovered so quickly from his bonk on stage 16 was to have doped. But this only explains Landis' recovery. It does not explain his extraordinary performance. In essence, your argument can only explain why Landis was not spit out the back of the peloton in stage 17. It does not explain how he outperformed the peloton.

I acknowledge that cycling has a doping problem, but I don't agree that everything you see in cycling can be explained by dope.

Ross Tucker and Jonathan Dugas said...

Hi Larry

Thanks for the considered question and comments. We really value your kind of input, it's thoughtful and thought-provoking and so contributes a great deal to the overall understanding of whatever problem might exist (doping, in this case).

To begin to respond, I will point out that our goal here is not to be a scientific journal - if you want facts referenced scientifically, that's where to look. Having said that, we do pride ourselves on the fact that we attempt to provide a form of "truth" (if such a thing can be pinned down) when it comes to our interpretations of sports events. So our staple "diet" here is the sports news, with a scientific angle.

Looking back on the Tour posts, I don't believe we took any liberties or ever crossed the boundaries of opinion vs. fact, however, for reasons I'll get into shortly.

First, I just want to draw an analogy here with the Oscar Pistorius issue. That is a sports news story which is very obviously heavily biased towards science. It presents us with an ideal opportunity to provide commentary and insight on the matter, and I believe that what I have written is the nearest thing to "fact" that exists on the matter (I'm not sure if you've followed that particular debate). However, I'm also well aware that there is a differing view-point, and some people have been heavily critical of me for projecting what they say is "opinion". I was even censored by a former employer who was concerned that my views would be seen as theirs.

The point is, science leaves great room for debate and opinion. I'm not sure what your background is, your arguments are excellent and well put, so I suspect you have a great deal of education and reading behind you - the thing to realise, however, particularly with the physiological sciences, is that nothing is clear cut - it's never a case of simply providing evidence to make the answer obvious, which I get the feeling is what you're looking for...

So for example, does dehydration cause muscle cramp? Does Oscar Pistorius have an advantage? Does altitude training really work? Who knows? Science goes both ways on all these issues! So let's now get to your question: Does doping increase or decrease variation?

I stand by my contention that a doping-free peloton is a more competitive one over three weeks of racing. There are many reasons for this, but primary among them is that the doping products improve recovery, which means that the distibution of effort over three weeks of the Tour can be much more liberal.

When I wrote that we're seeing more jostling, smaller time gaps, more brutal racing, my reference was the three weeks of riding in the Tour. I believe (backed up by science) that it's the pacing stategy of the riders that is most affected by drugs. A rider who is "clean" cannot afford the kind of effort that would provide him with 2 minute time gaps because the next day they would pay for it as a result of the impaired (though normal) recovery on the doping products.

I believe that in the past, you saw guys who could afford to put it all out on the road because they knew that they would recover by the next, thanks to the wonders of blood doping and EPO. There is evidence that the drugs work this way, though of course no clinical trials have ever been done in a Tour-type research study,for obvious reasons.

There are studies that have examined the distribution of effort over the course of a 3 week Tour, and they have found that riders do pace themselves on each individual stage in order to preserve energy for the three week race - that is obvious (Lucia et al. is the reference, if you're interested).

So, the removal of drugs from the sport (albeit partly) will serve to alter that pacing strategy, such that riders cannot make the same efforts day in day out. This will only serve to narrow the gaps, because suddenly you are dealing with athlete's natural variations in physiology and performance, and so it makes sense that the turnover and 'jostling' for position will be more frequent and more competitive. It's simply not possible to ride at 90% of your ability any more, day after day. One day at 90% will be followed by 80%, perhaps even 70%, as we have seen in this Tour. Only Ricco was able to attack consistently and feature on all the big climbs.

Your point about doping affecting all riders equally is another interesting one. I think that it is safe to say that this is not the case - you need only to look at the studies in exercise physiology to observe enormous variability in any physiological response/system to apprectiate that it's highly unlikely that people respond to doping the same way. There are "responders" and "non-responders". As an aside, this is true for altitude training as well (where "non-responders" is a documented term, by the way) and for the supplement creatine.

Few studies exist on EPO, however. The one that we did discuss last year (November 7th) did show some quite large variability between individuals, so it's pretty certain that you can't expect doping to narrow the gap.

The other point that follows on from this is that the doping riders (presumably of the past, though the problem still exists) are all on different drug cycles, including time, dose, maybe even type. I don't know the ins and outs of the doping practices. The Tour is not a giant laboratory research study, where you can give all the riders the same product, in the same dose, on the same day, and then control their responses to it. That would be the argument you're making for why doping would produce uniform changes in performance, which I don't believe it would.

The point that must be made is that at this level of performance, the gaps between the riders SHOULD BE very small. You are taking a sample of the very best riders in the world, and statistically, working on a "normal distribution" where the Tour represents the far right extreme of human performance. When a couple of guys then stand out from the rest by 3 or 4%, that should send alarm bells ringing, because it simply shouldn't be the case. I do recognize that great performances happen, and outliers exist, but the margins are so tiny, that when dominance comes along, it should serve as a "flag". Not proof, mind you, but certainly a flag.

Therefore, the fact that this year's Tour is more competitive would mean the absence of such flags, which again doesn't mean the race is clean, but that it stands a better chance of being clean than when the same guys were flying off the front EVERY DAY, with little regard for pacing themselves during the three-week race.

As for Landis, all I'll say to that one is that two different courts, plus the initial testing lab have heard the case, seen the samples, provided the evidence, and he's three strikes now. Eventually, as Oscar Pistorius showed, if you fight the science with enough legal muscle, you can cast doubt over anything. But it's striking to me that so many people have heard the case and dismissed it, apart from one person on the US panel.

As for why he'd take testosterone, the benefits of that particular drug would be quite large during training or heavy racing as a means to delay fatigue and improve recovery. Studies have shown that riders experience a fall in testosterone over the course of a 3-week Tour, and if this fall could be prevented, the recovery would be improved.

Have you considered the possibility that Landis was using testosterone all along, but in small enough amounts to avoid being "flagged" for its use (the tests only pick it up if a ratio exceeds a minimum threshold, remember)? Perhaps after the "bonk", he either made a mistake with dosage, or some physiological response to that bonk (the production of more natural testosterone) caused his level to rise above the threshold?

Or even more provactive, perhaps he was using testosterone during the training period earlier in the year (April/May), to help his training, and then maybe he took blood out, and re-infused it for that now infamous day. Perhaps it was blood doping with tainted blood that caught him out? All conjecture, sure, and this last part is opinion (so no science, I'm afraid).

But the first part, about pacing and the fact that a dope-free peloton would be more competitive, tighter and slower (that data remains to be collected) is, I believe true, and based on scientific theory. That is the best we can offer - scientific theory, because as I said, the world is not a giant lab which offers up beautifully controlled scientific studies with black and white answers.

I stand by the theory though.

Thanks for the debate!

Ross Tucker and Jonathan Dugas said...

Hi Barrld

Yes, indeed, it will be a huge question to see whether Hall has what it takes. He is certainly a big chance, and we won't forget him, don't worry!


Larry said...

Hey! Your reply today, THAT's what I expect from you guys. And all you had to do was replace one sentence from your earlier post with about three pages of closely reasoned analysis. Was that so hard? ;^)

Seriously. I appreciate what you guys do here. I understand that the world is not a lab, and I don't want to take the position that science can speak only when there's close to absolute certainty. (That would put me in the same camp with those who argued, once upon a time, that the proof was not in on the dangers of smoking or the fact of climate change.) What I do expect is for you guys to state a disprovable hypothesis, and give me reasons reasonably based in fact to support your hypothesis. Clearly, you did that in your recent post, and I appreciate it.

(I'm going to skip past the parts of your analysis addressing the Floyd Landis situation. I'm happy to discuss Floyd Landis with you guys at a later point, but that's potentially a lengthy discussion, and we have enough of a challenge on our hands just focusing on the 2008 Tour.)

So ... let me see if I can summarize, and of course correct me if I summarize incorrectly. You guys hypothesize that in a clean peloton, on average we will see smaller variations between individual rider performances than we would see in a doping peloton. (Actually, you used the word "competitive", but "competitive" is a subjective word -- there are too many ways for a race to be "competitive".) That strikes me as a proper, potentially disprovable kind of hypothesis, worthy of scientific inquiry.

The main argument you advance to support your hypothesis is that with doping, the distribution of effort (in other words, the pacing strategy) over three weeks of the Tour can be much more liberal if riders are doping, because doping products aid in recovery. I agree that pacing strategy is critical to success in the Tour, and is foremost on the minds of the riders and the teams. Any outside factor (like doping) that could affect pacing strategy is going to have a large effect on the Tour.

Your argument starts to fray a bit (IMHO) when you start to talk about riding at percentage levels of ability. You state that without dope, a rider cannot ride consistently at 90% of his ability any more -- he'll have some days riding at 90%, but some at 80% and maybe even a few at 70%. With this argument, you're taking a position that I'd call "old school"! It dates back to the days of Lance Armstrong, where the argument was that no one could ride as consistently as Lance without doping, that "natural" cycling necessarily involves good days and bad days. This "old school" argument is reasonable enough, but it has been replaced in recent years by the "believability" argument, where it's the exceptional performance that indicates doping. Of course, an exceptional performance by nature is not "consistent" with other performances. (This is one reason why I try to insist on a disprovable hypothesis. If riding consistently means that you're doping, and riding inconsistently means that you're doping, then we've departed from science into the realm of the self-fulfilling prophecy.)

But I digress! The problem I have with your argument about percentage levels of effort is that this argument (standing alone) does not explain differentiation between individual rider performances. If doping allows the doper to ride at a consistently high level of effort (say, 90%), then this benefit is available to everyone who dopes. It does not (standing alone) explain how doping allows one rider to exploit that 90% effort to ride rings around other doping riders who are also able to ride at 90%.

You may be venturing into a fascinating discussion of game theory if you are arguing in essence that a group of riders all able to ride at a consistent 90% level will behave differently than a group of riders who ride at less consistent and less predictable levels of effort (say, 70% to 90%). There may be something to that argument. Personally, I would think that you'd see MORE attacking and aggression in a peloton with greater day-to-day variations in individual rider's abilities ... but this post is already plenty long, so let's shy away from the game theory discussion for the moment.

Since I've conveniently decided to ignore game theory, so far we have not touched upon anything that would allow one rider to exploit doping more than his peers in the peloton.

This brings us to what I consider to be your best argument, which puts aside questions of consistency of effort and directly addresses the question. I argued that doping should help all doping cyclists equally, and you argue that this most certainly would not be the case. I think this is much better proof for your hypothesis than a discussion of level of effort and rider consistency.

If doping does not help all riders equally, then that pretty much blows out of the water any discussion we might have about strategy, and doping producing consistent versus exceptional results. If doping helps some riders more than others, then it doesn't really matter how that "help" expresses itself: it could help a Lance ride with great consistency, and a Vinokourov put in exceptional stages. Help is help.

I'll freely admit that when I've read the scientific studies on doping, I’ve focused exclusively on whether doping really enhances performance, and to what extent, and I did not focus on whether some were helped more than others. But now that you mention it, the Landis case DOES feature testimony to the effect that there are significant differences in individual reactions to exogenous testosterone administration. I'm not sure about the significance of individual reactions to EPO or altitude training -- I'd expect that there would be differences, I don't know how significant they might be.

For purposes of discussion, I'll concede this point: let's assume that there are significant differences in the performance-enhancing effects of doping products on different people. So, we can imagine a peloton of 200 clean riders, and among these riders we can imagine a 1% difference in performance ability from top to bottom. If we then allowed each rider to dope, we'd expect that individual differences in response to these doping products would widen the gaps in performance abilities.

It would be as if professional cycling had never seen a time trial before, and you put together the top 200 cyclists based on their demonstrated abilities to perform well in the usual flat and mountain stages. They'd be bunched together pretty closely, in that 1% range you've described. If without warning the Tour then introduced a time trial, that would tend to blow apart that 1% range. It would turn out that some of those 200 cyclists were a lot better than others at time trialing -- the guys like Evans would finish way ahead of the guys like Frank Schleck. I'd figure that the same thing occurs with dope -- if you've assembled the top 200 clean cyclists and then introduce dope into the group, you'd see much larger variations in performance as a result.

But this imaginary experiment is useful to our understanding of pro cycling only if we imagine that riders are not introduced to doping until they reach the very top level of the sport. Let's say that the 200 riders in the Tour de France peloton are at the very top of a pyramid of professional and young amateur cyclists, perhaps numbering 20,000. If those 20,000 cyclists are doping in significant numbers, then the riders who rise to the top of the pyramid are likely to be those cyclists who respond well to doping products. (And by the way, there is every indication that doping occurs throughout all of professional and serious amateur cycling.)

Actually, we can consider this in a slightly more sophisticated way. Let's assume for the moment that the 2008 Tour was significantly cleaner than the 2007 Tour. Let's also assume that for the most part, the riders in 2008 were substantially the same group as the riders in 2007 (I know that this can be disputed). If this is the case, then what we have is really the opposite of my thought experiment, where we added dope to a group of 200 previously clean riders. What's ACTUALLY happened is a lot more like a thought experiment where we deny dope of a group of 200 previously dirty riders. The dirty riders rose up to the top of a dirty pyramid, based in part on their ability to take great advantage of their physiological reaction to doping products. So, consider the opposite thought experiment, where we deny doping products to a previously doping peloton. What would be the effect? It would tend to increase the performance differences between riders -- the riders in the group with the best reactions to dope would fall well behind the riders in the group who derived the smallest advantages from doping.

Note that in this year's Tour, we did not see any dramatic shake-up in the composition of the riders at the top of the GC classification. Of course, we saw a few young riders (Kohl, Andy Schleck) make dramatic improvements, but we'd expect to see young stars emerge in any Tour. The question is, did you see dramatic relative changes in position among the established stars of the Tour? Did the few guys in the peloton who rose to the top because they could best utilize dope, fall to the bottom of the peloton in the absence of dope? No. For the most part, the top riders stayed at the top.

Let's try coming at this from a slightly different perspective. If you juxtapose the two thought experiments: the effect of doping on a clean peloton versus the effect of going "cold turkey" on a doping peloton, I think a larger truth emerges. If we assume that the 2008 Tour contained a different level of doping than previous Tours (and without this assumption, there's no reason to look at differences in the Tours to derive conclusions about differences in doping), then calling 2008 "cleaner" or "dirtier" is way, WAY too static a viewpoint. You also have to consider the dynamic viewpoint that doping conditions in the peloton were CHANGING.

What tends to happen in a social system if you introduce change? OK, granted, that's a wildly general question, but the general answer is that a changing social environment should create greater variations in human behavior than would have been evident when the social environment is more static. As an example, compare and contrast driving habits before and after the recent spike in gasoline prices. You can figure that before the price spike, people’s driving habits were relatively consistent from year to year, and that the driving habits of similar situated people weren't all that different. Now with gasoline prices through the ceiling, people are reacting in all sorts of different ways. Some give up their cars for mass transit, some buy more efficient cars, some drive less ... but some stubbornly refuse to change, and some are probably driving even more (it’s time to squeeze in that last road trip before prices get any higher!).

Eventually over time, a social system absorbs change and returns to "normal". Most of the older guys who are now riding their bicycles to work will get tired of the cold and the rain, and join the rest of us on mass transit. The people who sold their homes to move closer to their places of work will not continue to sell their homes to move even closer to work. Things will settle down. The pressure to conform will take hold.

So. Here's my counter-hypothesis: if the Tour is really getting changing for the better, the changing environment in cycling should evidence itself in greater variations in performance between the individual cyclists, and also in a greater unpredictability in the race as a whole. I did not see this in 2008, and I will conclude that the race did not get any cleaner.

There is a lot more to say on this subject. One factor we have not discussed is the fact that, regardless of whether cycling got any cleaner in 2008, cycling sure TRIED to get cleaner! The increased efforts in enforcement (the biological passport, the targeted testing) DOUBTLESS had an effect on the behavior of individual cyclists.

I'd say more, but real life intervenes and my fingers are tired.

Anonymous said...

My english are limited so i would just write few points:

About difference response of athletes to doping.
Vayer ex-Festina doctor has stated that his best rider was Brochard but Virenque was the rider who have the best response to many different kind of PED.

Some times PED uses gives some strange results like what happened to Zulle( Festina rider) who have a strange bonk, in fact something named a blocato , followed by an excellent day. The famous Willy Voet explained that Zulle body didn't react well to a different mixture of PED,(corticoïd especially) they were trying to boost him in that stage... Landis' failure was more probably similar than a bonk followed by a blood doping transfusion.
When a rider (or an athlete, it's true for many sport) has a bad day, he can feel it. So the advice is to do the less effort and not to cross the limit. If the athlete cross that limit, he will destroy his body capacity for the next days.

About the recent TDF, we have seen some riders like Popovich be less strong than precedent years despite his priviligied role as Evans' support, a lot of spanish riders were not able to stay in first peloton as usual.
Many EPO are still undetectable, windows detection,... or don't give an AAF as was stated by Damsgaard.
And the effects of the doping past can give an advantage :persistent strength of the "hard work" which has been done with the use of PED.
I think even if we have seen some positive change we need at least more one year to see a more significant change on performance.

Jean C

Larry said...

Jean C -

I know and admire the depth of the analysis you provide, here and elsewhere. And I wish that my French was anywhere near as strong as your English!

I have no doubt that different athletes react differently to different doping regimens. So, when a new doping regimen is introduced into cycling, I agree that some riders will benefit more from the regimen than others. However, I also suspect that this may be more true for certain doping methods than others; my guess is that the more direct and straightforward the method of doping, the more uniform the effect should be on the performance of different riders who utilize the method. For example, if you're trying to raise a rider's hematocrit levels, it may be the case that certain riders do not react as well or as quickly to EPO than others, and you may have some riders who cannot get the desired effect from taking EPO without risking detection by taking too much EPO (more on this later). But a "direct" method of raising hematocrit levels, like blood doping, should work for any rider.

One of my main hypotheses here is that to whatever extent doping is prevelant in cycling, it's been prevelant in cycling for a long time, at all levels of the sport, so that the riders who react best to doping have already risen to the top of the sport. This means that I disagree with Ross and Jonathan that doping would tend to increase the differentiation of rider performances at the top of the sport. Being "good at doping" simply becomes one of the physiological traits that a rider needs in order to be able to compete at the highest level of the sport.

Of course, as radically new doping methods emerge and are introduced into cycling, this SHOULD have the effect described here by Ross and Jonathan. If you take an EXISTING population of top riders and introduce some new physiological trait that the riders must have in order to excel, then only certain of the top riders will have this trait in sufficient quantity to win. Let's try an imaginary example. Let's say that tomorrow the top drug companies come up with drug XYZ, a drug that cures wax build-up in the ears, but also has the side effect of causing cyclists to fly up mountains like they were "on a goddamn Harley." And let's suppose that this drug is taken by 80% of the riders in the peloton. The effect will be to produce greater differentiations in rider performance than we've seen before. Obviously, the 20% of the riders who don't take XYZ are going to lag behind. But even among the 80% of the riders who take XYZ, some will react to it more strongly than others. If XYZ is a powerful enough performance-enhancing drug, then after the introduction of XYZ, the riders who finish at the front of the peloton will be the riders who reacted best to XYZ.

Now in our imaginary example, let's flash forward ten years. Ten years after the introduction of XYZ, we can imagine that XYZ is a known performance-enhancing drug, and that it is available to anyone serious about becoming a pro cyclist. If XYZ remains a powerful performance-enhancing drug, then the guys who rise to the top of the sport will tend to be those guys who react the best to XYZ. Of course, they'll ALSO be the guys who have high power-to-weight ratios, and have the skills to ride at high speeds down hills, and have a high tolerance for pain, and possess all of the many skills necessary to win a three week Tour. Good reaction to XYZ would always be just one of many skills a rider might need to succeed. But reaction to XYZ would no longer be a factor that would tend to produce increased differences in the performance abilities of riders at the top of the sport, because reaction to XYZ would have become one of the established physiological traits required to rise to the top of the sport in the first place.

I think much of what you describe in your post is consistent with my hypothesis, as you're talking about guys riding during an era when there was wide experimentation with different doping techniques plus introduction of many novel doping techniques. My suspicion is that doping is more stable and conservative today, in part because ADA testing has gotten better and riders want to stick to doping methods that have proven to be undetectable in the past. This is what got Ricco in trouble in this year's Tour: rather than sticking to tried-and-true second generation EPO (which appears to be undectable), he switched to new and improved third generation CERA EPO (which unfortunately for him proved to be detectable).

Regarding Popyvich and other Spanish riders ... actually on this point I strongly agree with you, though my opinion is not terribly scientific. I DO think that the performance of certain riders in the Tour may have suffered because they did not dope as they had done in the past, or that they did not dope at the levels they had done in the past. This is consistent with my hypothesis that in a sport in the process of becoming cleaner, you will see greater variations in performance than before, as those riders who rose to the top because they reacted most positively to doping products suffer proportionally greater losses in performance as these doping products are taken away. On this count, I note a feature of this year's race that we have not yet discussed here: except for team CSC, none of the top teams were able to provide any kind of consistent support to their GC candidates in the tough mountain stages. This wasn't only true for Evans: Kohl, Menchov and Vande Velde had no consistent support from teammates in the mountains. I think this indicates a performance fallout from the kind of support rider (like Popyvich) hired for the primary purpose of providing this support. And this is the sort of performance fallout (i.e., GREATER differentiation of performance) I'd expect to see in a peloton changing for the better.

Unfortunately I also agree with your other conclusions. Despite the bold and confident pronouncements during the Tour from the race organizers and the drug testers, there appears to be many ways for a rider to cheat and get away with it. The EPO you mention is one way. We can also look at the "menu" of doping items listed by Victor Conte in his BALCO testimony that are supposedly performance-enhancing and impossible to detect. Given the amount of different PEDs that are undetectable, I don't share your optimism that we can expect further positive change.

I don't think that any reasonably sophisticated doper is afraid of in-competition testing. They MAY fear the biological passport approach to doping testing. That's another topic for another time.