The Outer Line: A missed warning signal

During the Tour de France 25 years ago, pro cycling missed an important early warning signal about the escalating impact of EPO on the peloton.

The individual time trial at the Tour de France on the shores of Lac de Vassiviere, in 1995, suggests a simple way to effectively police the peloton and make the sport more exciting at the same time.

The recent ESPN “30 For 30” documentary on Lance Armstrong may have been intended as a character study, but Director Marina Zenovich also provided some new and deeper insights into professional cycling’s doping culture – and how the sport ignored many warning signs and institutional failures. Pro cycling’s reputation remains shaky in the public eye as a result, even though it has arguably innovated and implemented more sweeping measures to control doping than most other sports. These new insights should not be overlooked as we continue to strengthen cycling’s credibility and build a better sports marketing image for its economic future.

Interestingly, the sport totally missed one of the early warning signs about its emerging “EPO era” crisis – one which came 25 years ago near France’s largest artificial reservoir, Lac de Vassiviere, the picturesque backdrop for the 1995 Tour de France’s final time trial. The reputation of the sport might not have been damaged as much if the red flag raised there had been recognized for what it was. Indeed, it might have led to more focused scientific and analytical methods to help build a cleaner sport.

Before we dissect that time trial stage in detail however, it is important to discuss the context of why the data from that race might have represented a milestone turning point. Cycling is a notoriously difficult sport to physiologically analyze, despite all the power meter and heart rate tools we have at our disposal today. Individual athletes can be tested and observed under controlled lab conditions to measure wattage, heart rate, oxygen consumption, and other parameters that are critical for endurance sport success.

In addition, many analysts and critics (such as former Festina trainer Antoine Vayer, and Francaise des Jeux coach Fred Grappe) have painstakingly dissected performance and power metrics which riders have unleashed on famous cycling finishing climbs and timed events to try to analyze what is – or might not be – an authentic effort. Others like Dr. Michael Puchowicz and his colleagues are proposing a critical power model which can help to indicate doped performances. But in most open road racing situations, conditions such as wind, rain, road surface, and inter-team race tactics can significantly distort such observations.

There are other ways to review the data. For example, looking at the Tour de France’s average speed over its history shows that the racing has gradually become faster. However, as shown in the chart below, the race’s total distance has also become progressively shorter. Thus, the higher overall speeds and corresponding average physiological/power output values seen today do not necessarily suggest that riders are cheating.

To really get a handle on potential changes over time, we need to get a complete look at peloton-wide performance for very closely comparable challenges at different points in time – and without the interference of widely varying racing tactics. And no cycling event is as simplified or easy to analyze as when GC contenders race against the clock in an individual time trial to preserve their overall position in the race. From this perspective, Lac de Vassiviere provided cycling with perhaps its best historical and most under-appreciated reference point for assessing the impacts of the EPO era, as described in “30 For 30.”

On three different occasions – 1985, 1990, and 1995 – the town of Lac de Vassiviere-en-Limousin hosted the Tour’s final time trial. Each edition came at the end of the third week of the Tour, placing the athletes at the limits of their fitness and recuperation with respect to rider conditioning. And with respect to the road course, each edition followed an identical loop of the reservoir (with a very slight difference in 1995 owing to the start/finish zone placement). Furthermore, the weather conditions were very similar in each instance (although early starters in 1995 had the benefit of cooler, overcast conditions). And critically, as mentioned, there is very little in the way of team tactics involved in an ITT.

Greg LeMond won the first edition in 1985 by five seconds over Bernard Hinault. Iconic images of LeMond blitzing the hilly and difficult course in the now-retired “combination” classification jersey became the inspiration for many new 1980s American riders. Erik Breukink took the 1990 edition, while LeMond finished fifth, to take the yellow jersey, en route to his third Tour win. Miguel Indurain triumphed in 1995, sealing his final and fifth consecutive Tour victory.

Tour de France time trial comparisons

But comparing the finishing times of these three-time trials tells a much different, and in retrospect a much more insightful history lesson. Over nearly identical distance and in essentially identical weather conditions, LeMond and Breukink covered the course in extraordinarily similar times. Breukink’s 11-second improvement on LeMond’s 1985 winning time had the benefit of 1990s more specialized TT bikes, lighter disc wheels, and critically, the addition of aero handlebars. The close similarity of that winning time, and the comparable spread of the top 10 riders in those first two editions when compared to each other, underscores what is still considered to be an extremely challenging and technical course.

By comparison, Miguel Indurain’s 1995 winning time seems completely out of context. He obliterated Breukink’s 1990 effort by over five minutes – on a version of the course that was actually over 500 meters longer than in 1985, to accommodate better start line logistics. Rolf Aldag’s pedestrian 27th place in 1995 would have tied Breukink’s winning time five years prior. And going deeper into the rankings, Stephen Hodge’s last-place in 1995 was also nearly five minutes faster than the 1990s last-place finisher. (A more detailed comparative analysis of the three editions, including course maps, can be found on a Greg LeMond fan web site. And for those in the Armstrong vs. LeMond mood, Armstrong’s 43rd place was 21 seconds slower than LeMond’s 1990 Tour-clinching effort.) In short, the entire peloton picked up the pace by about five minutes – an approximate and improbable eight percent improvement.

It is conceivable that refinements in body position, bike aerodynamics and better wheel choices could account for perhaps one or two minutes of improvements, though the use of aero bars and some aero helmets by the time of the 1990 edition didn’t appear to have much effect on the winning time. So where did all the other improvements come from, and how did they all add up in 1995? Was it the sum of more scientific training, and many “marginal gains?”

The Lac de Vassiviere time trials do not confirm who doped in any of its three editions, but they do conclusively suggest that a pervasive problem was locked in place by 1995. Every one of the 115 riders who completed that Tour, from top GC contender to lowest domestique, was significantly faster – and that day in Lac de Vassiviere stands out as a data anomaly supernova. “Alien” time trial performances and “angelic” climbing had become “normal” by 1995 – as was the availability of EPO, due to production advances and pervasive ‘gray market’ drug diversion in the U.S. and Europe. But few people really noticed or commented that the sport had changed that much, because everyone was riding faster all at once.

This simple observation goes a long way to confirming the testimonials presented in “30 For 30” and elsewhere – from mountain bike racing legend John Tomac sharing details of his decision to leave the European peloton in 1991, to Derek Bouchard-Hall describing his introduction to European racing, to David Moncoutie’s recent description of trying to race clean at the height of the era. In just five years, the doping culture had reshaped professional cycling.

Lac de Vassiviere’s history lesson 25 years ago is almost unique in cycling and therefore worthy of a closer re-examination. Since then, the Grand Tours have often revisited many of the same finishing climbs – from Alpe d’Huez and the Angliru to Monte Zoncolan and Mont Ventoux – but as pointed out earlier, evaluating road stages over the years is far less conclusive – with so many weather, strategy and team tactical decisions complicating any direct comparison. With the exception of two Tour prologues at Le Puy du Fou in 1993 and 1999, no ITT stages have returned to and duplicated the same long time trial course as closely as Lac de Vassiviere’s three editions. In fact, to a skeptic, it could almost seem suspicious that we have not seen more replicate courses that could provide clear longitudinal markers for the progression of rider abilities in our current era.

The dismissive explanation of “marginal gains” for every new jaw-dropping effort is gradually losing its credibility for sports like cycling. Author David Epstein pointed out in 2013 that sports are reaching an ever wider television audience of younger potential participants, and as a result, there may always be outlier athletes with genetic gifts and greater potential who emerge and rise to the top. But at the same time, biology experts have modeled, and believe that these top athletes are already and asymptotically approaching natural human limits – such that incremental improvements may only come with so-called marginal gains. Summarizing many years of his research for Scientific American, physiologist Peter Weyand said, “I think people will find ways to enhance oxygen delivery through the body and squeeze more performance out of humans. The only question is will these approaches be considered legal.”

Cycling is in the midst of a challenging structural crisis because of the ongoing pandemic, but this must also be viewed as another opportunity for change. Different and innovative anti-doping methods can be part of the plan, but this historical anecdote suggests a very simple way for the sport to recognize and publicly come to terms with its complicated past.

Rather than passing off Lac de Vassiviere as an anomaly or a missed signal, grand tours and other stage races should deliberately schedule such identical, fixed-course stages into future events, to serve as performance reference points for competitors and the sporting public alike. Such events could be exciting gauges against the past, and they could provide a true benchmark for rider fitness in the modern era – perhaps becoming new “sportif” destinations where recreational cyclists can test themselves against verified efforts of the world’s best riders. But most critically, they could help rebuild trust in professional cycling.