With the 103rd Tour de France underway, the inevitable questions surrounding performance-enhancing drugs weave their way into just about any conversation about the Tour. Rightfully so; between 1999 and 2005, no official champions are credited with winning the Tour de France after Lance Armstrong’s admissions of doping during his seven wins of that period. While volumes have been written about doping in sports and it makes for fantastic debate, the interesting story from a data analytics perspective, is how the World Anti-Doping Association (WADA) changed their approach to testing athletes to clean up the sport. And, at a broader level, how different approaches to analytics can change the game – or, in this case, race.
For over 15 years, Lance Armstrong proclaimed his innocence to frequent accusations of doping, by citing that he was the most tested athlete in all of sports and he had never failed a blood test. While there were a few discrepancies in a handful of tests during his time as a professional cyclist, this statement is generally true. (Note: many other athletes in and out of cycling have made the same claim. However, as sports stories go, few, if any, can compare to the vast chasm between what Armstrong claimed to be true and reality on such a grand stage).
This raises the question: “How was an athlete that admitted to a highly organized and prolific regiment of doping over 15+ years able to beat the system?” And what changed within the testing procedures that now makes it so much harder for athletes to cheat? As it turns out, athletes that dope are really good at hiding the presence of substances like EPO, Human Growth Hormone (HGH) and steroids from those charged with testing them. For every test that WADA or USADA came up with prior to 2009 to discover banned substances, dopers were already two steps ahead of them. The problem was that these organizations were conducting real-time tests at a single point in time to identify the presence of banned substances. Athletes were far ahead of the testers in masking the presence and effects of the drugs. So what happened?
In 2009, WADA adopted a new method of testing called the Biological Passport. Instead of conducting real-time analytics for the presences of drugs at a single point in time, they began analyzing samples for the effects of drugs over an extended period of time. According to Peter Vigneron’s Outside Magazine article, “What the Heck is the Biological Passport, Anyway?”: “A single screen may not yield much information about doping (real-time analytics), but 10 or 20 (or even 30) can reveal not only what drugs the athlete is taking, but when and for what purpose” (batch analytics). Instead of looking for the presence of drugs, the new method looked for variations in different types of red blood cells over time. Simple version: Clean athlete = no variation. Dirty athlete = significant variation.
In terms of data analytics, this highlights an important distinction between different methodologies. In this instance, batch analytics, which used multiple data points over the course of an extended period of time, yielded better results than real-time analytics, which analyzed a single sample at a single point in time. Each serves a purpose, but the lesson to be gleaned from WADA is that the proper application of analytics can mean the difference between success and failure. More importantly, when the existing process isn’t producing the desired results, reevaluating the analytics processes and methodologies is an important step in reform. Check out ViON’s eBook, “A How to Guide to Predictive Analytics” to learn more about making the right big data analytics decisions for your organization.
The complexion of doping in sports has changed dramatically over the past decade. While the testing process is still not perfect, analytics is bringing a degree of respectability back to a tainted sport. What examples can you think of where data analytics has or is now making an impact?