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.

Saturday, July 06, 2013

Clean performances to surpass doped performances?

When will clean performances surpass doped performances?

Will clean cycling performances ever match those of the doped generation of the 90s & 2000s?  And when?  Let's look at progression, doping effects and realistic expectations.

Yesterday David Brailsford gave an interview about the release of power data to the public.  In it, he said the following: 
"At some point in time, people have to accept that performances are going to move forward. If we always hold back, and say, here is some data from people who were doping, then if we draw a line, we can then deduce that anyone crossing that line must also be doping. Well, that’s false. They do not have to be doping, because the whole human race moves forward. At some point in time, clean performances will surpass the doped performances in the past."
That's an interesting statement to analyze, so some quick thoughts on it.

Normal progression: The drivers of progress

First, he's not wrong - progression is part of sport, and so it does stand to reason that on the day you remove doping from the sport (ideal case scenario, bear with me), performance will regress back by whatever amount doping improved it, and will then slowly creep back up until it eventually matches and then surpasses the doped performances.

There are two key questions regarding that progression.

First, ask what are the factors that drive progression, and secondly, ask what is a normal rate of progression?  The two are of course interlinked, but dealing with one at a time:
  1. Factors driving progression include:
    1. Physiology, which is an all-encompassing term that includes training induced adaptations, better athletes to begin with (think emergence of East Africans - different athletes), nutrition, psychology - basically, all the things the sports scientists reading this lose sleep over!
    2. Technology, which I'd expand using some creative licence to also include technique related factors.  There's no better illustration of this than swimming, where the invention of the tumble-turn dropped times, then swimming goggles, faster pools and of course swimsuits all shift performances dramatically
    3. Commercial factors - show me the money, and I'll show you improvements in performance!  Marathon running has for instance benefitted from increased sponsorship, which then attracts better athletes (see point 1) and greater levels of competition.
    4. Doping - and this is the crux. If we can understand the impact made by doping on performance, then it opens a world of insights regarding sudden advances, regressions in performance with better doping controls, and the present discussion.  Unfortunately, this is not too easily known, because controlled studies have never really been done to quantify the magnitude of the doping advantage, as I'll explain shortly.
The point is that it's very difficult to tease these factors apart - commercial factors drive technology, which drives physiology as a result of increased exposure, participation, focus etc.  Doping affects physiology too, not just acutely, and so isolating one is impossible.

The technology impact

The linkage between the above factors complicates the discussion around a realistic progression.  To illustrate, this consider technology, which has an instant, and sometimes dramatic effect on performance.  The introduction of Speedo's LZR Racer has such an enormous effect on performance that the 2008 Beijing Olympic Games that 94 out of 96 available medals were won by swimmers wearing it!  World and Olympic records were broken in all but two events, and by the time the sport's governing body ruled the suits illegal, 178 world records had been set across all the events (I think...I may have lost count at some stage).

Point is, in the absence of technological controls and rules, as was the case for swimming pre-2010, advances in equipment can dramatically affect performance progressions.  The same is happening with carbon fiber prosthetic limbs for amputees, incidentally, though that's also related to better athletes emerging.

Cycling has some guidelines in place with respects to equipment.  There is still some 'play' within those guidelines, but revolutionary breakthroughs are no longer common place.  Look at the evolution (and then devolution) of the world one-hour record to see this in action.

So what this means that over the last few generations, where the guidelines have existed, the impact made by equipment is much smaller than in swimming, but probably larger than in athletics, where equipment matters less (javelin's issues being the exception - I'll stick to track events for this discussion).  

Certainly, comparisons in performance between the 1980s and present day are grossly inaccurate, affected enormously by the advances in bikes - lighter frames, stiffer, gearing, brakes etc.  So the best way to understand performance progression is to try to rule out major technological advances as much as possible, and this is done by narrowing the time-frame as much as possible.  I accept that bicycles improve all the time, but these are incremental improvements, and the magnitude of advantage they confer is now small enough that I think it is negligible within the time-frames being discussed here.

Similarly, commercial factors have changed enormously since the days of self-seconding, small prizes and little media and sponsor involvement.  But they haven't changed too much (relative to doping, that is) since the EPO era of the early 90s.  We can thus consider any improvements as a function of physiology and doping.

The doping impact

Aside from the occasional study that has typically used recreational or well-trained, but not elite athletes, relatively little is known about the magnitude of performance enhancement due to doping.  A recent paper by Pitsiladis found a 6% improvement in running performance over about 9 minutes in good athletes who took EPO.  Is the effect larger or smaller for elites? Is it larger or smaller for very long duration exercise?  You can actually argue those questions either way, so we just don't know.

We also know that a historical comparison between the 1990s, 2000s and 2010s (call them the EPO era, the sophisticated doping era and the biological passport era) have revealed a slowing down in the peloton of between 5% and 10%.  What used to win Alp d'Huez (37 to 38 mins) is around 3 to 4 minutes faster than what was winning it since 2008.

Finally, some leaked documents from the East German system suggested that they expected performance improvements in the range of 5 to 10% from their athletes when using anabolic steroids.  It was higher for women (implications discussed below).  That's a typical or average improvement - there would undoubtedly be high and low responders to doping, just as there are to any medication or treatment.

All told, 5% seems a safe value to illustrate the concept with, though I stress it is illustrative.

The example of Track and Field athletics

So, the question is, in response to Brailsford's statement that "At some point in time, clean performances will surpass the doped performances in the past.", is when do we expect this?

If the effect of technology and commercial factors is considered negligible (I know, it's not, but bear with me for the example), then we can get some insight from track and field athletes.

I drew up the table below in 2009 to compare male and female athletes.  It shows the current world record, the age of that record, and also the gap between the best performance over the last four years and the world record.

It's pretty clear that women athletes deserve some sympathy - the world records are simply out of reach.  Driven by rampant doping in the 1980s, most of those records have been pushed so far beyond the reach of physiology, technology and commercial factors that they are simply untouchable.  

The men's side is a little different - the average age of records is 12 compared to 21 years, but that's skewed by those men's field events, which drive the average up.  So, what is interesting is that doping happened for BOTH men and women in the 1980s, but the women's times have survived, not the men's.
That is a function of physiology, because it reveals that doping pushed women's performances outside of the normal curve, as illustrated by the figure below:

What this shows is that doping has moved the top performances, the world records, into a space that now lies between men and women, and the passage of time, plus technology, plus physiology, plus financial incentives, has not been sufficient to help the best women close that gap.  In other words, the magnitude of doping improvements is greater than physiology, technology and commercial factors combined.

It may yet happen that the women are able to close that gap, but it's been 30 years and counting.  In some events, the gap is narrowing (the 100m hurdles seems imminent, for example), but for the rest, the 1 to 2% differences you see in the table seem, for this generation, anyway, insurmountable.

Cycling's case

Let's now go back to cycling, and ask when Brailsford's prediction might be borne out.

First, consider that in men's athletics (we use men's because as shown above, women's progressions are dramatically skewed by doping), the typical rate of improvement in the 100m and marathon has been around 1% every ten years.  That means 0.1 s or 90 seconds in the 100m and marathon, respectively, every decade.

If we apply this to cycling, once again setting aside technology and commercial factors, then we can begin to see where current performances should be, assuming they are non-doped.

I previously used three separate markers to suggest that doping might improve performance by approximately 5%.  If that were true, and if the expected progression was 1% per decade as for running, then doping is worth 50 years of progression.  I'd argue that because cycling has always had testing, it constrains the degree of doping, thus lessening the effect, and so a conservative assumption here would be that doping is worth 3%, and thus worth 30 years at that example rate of progression.

Therefore, remove doping and performance should drop back to where it was 30 years before, assuming no impact of technology.  This means the times we saw in 2010 would be comparable to those of the 1980s.  I think that is a reasonable assumption when you now factor in bike technology, and then begin to compare times of the late 80s and early 90s to what we have seen since the biological passport was introduced.

Therefore, it would seem reasonable to suggest that 20 to 30 years would be required to bridge the "doping gap" and return to the 37 minute ascents.

That's the starting suggestion - if you believe that it will take less time, then that's because you place more value in technology and training sophistication.  If you believe longer, it's because you believe in a physiological limitation that only doping can overcome, as is the case for female track and field athletes.  I'm in the latter group, myself, but cycling's technology means it's unlikely we'll find ourselves in quite that situation, where performances seem 'stuck' for decades.  Expect improvement, but slowly.

Thoughts are, as always, welcome!