MOSCOW (AP) Even as he worked to cover up doping by Russian athletes, Grigory Rodchenkov was developing technology which would help to catch them years later.
The former head of the Moscow anti-doping laboratory is the star witness for the World Anti-Doping Agency investigator Richard McLaren, whose report Friday accused Russia of operating a state-backed doping program which covered up more than 1,000 tainted drug test samples, including for medalists at the 2014 Winter Olympics.
However, Rodchenkov's role in helping catch drug cheats isn't widely known outside a small circle of the world's leading anti-doping scientists.
Methods devised by Rodchenkov and his former assistant at the lab, Timofei Sobolevsky, to detect two common steroids have become a crucial weapon for drug testers in a wave of retesting carried out this year by the International Olympic Committee, though some dispute the Russians' work.
So far, 62 athletes - almost half of them Russians - have been disqualified from the 2008 and 2012 Olympics in IOC retests after testing positive for turinabol, a banned substance which Rodchenkov helped make much easier for labs to find in samples.
There are also six cases involving oxandrolone, another steroid on which Rodchenkov carried out research, though all but one of those also tested positive for turinabol, a black-market steroid developed in the old East Germany which bulks up muscle and has plagued global sport for decades.
''Even if they are old and quite well known substances, there is continuous research on the metabolic behavior of these substances,'' said Tiia Kuuranne, head of the laboratory in Lausanne, Switzerland, which handles retests for the International Olympic Committee.
''These advances are the ones that lead to these kinds of breakthroughs or leaps in doping control.''
Those who tested positive for turinabol in IOC retesting competed for 10 different countries, mostly in the former Soviet Union, and range from Kazakhstan's Ilya Ilyin, who used his record-breaking weightlifting career to build a following of 400,000 on Instagram, to a Belarusian runner and a Russian wrestler. Many of them deny doping, including Ilyin.
The key advance was the discovery of new turinabol metabolites, the chemical traces left when complex steroids break down in the human body. Turinabol produces a wide range of metabolites, some of them quickly flushed out of a doped athlete's system, others which can linger for much longer.
If the steroid can be detected for longer, drug testers can catch dopers who ended their steroid use a few weeks before a major event like the Olympics, expecting they could never be caught.
''It's very non-sexy science to find these metabolites, but it's really cool what it can do, so many dopers. It's just amazing,'' says Marcus Ericsson, director of a WADA-accredited lab in Sweden.
Research conducted in 2011 by Rodchenkov and Sobolevsky found six new metabolites. One referred to as M3 in their research proved to be the key, raising the detection window from a few days of last use to as much as seven weeks, Rodchenkov and Sobolevsky estimated.
Rodchenkov and Sobolevsky ''had excellent scientific knowledge on methods, on steroid metabolism, really excellent,'' says Peter Van Eenoo, who runs a drug testing laboratory in Belgium and frequently met both men at conferences.
''Grigory, I need to say, also had knowledge which was better than most other lab directors, if not all lab directors, on usage, how people were using, the doses they were using ... It did make some people wonder how he knew about this.''
Turinabol, also known as dehydrochlormethyltestosterone, has not been manufactured for legitimate medical purposes for years but is widely sold on the black market without proper quality control. In many countries, that means it's difficult to conduct scientific studies using volunteers without breaking medical ethics rules, so it's hard to tell exactly how long it stays in the human body.
The drug has also been linked to health problems by East German athletes forced to take it during the Cold War.
Before the 2012 London Olympics, McLaren alleges Rodchenkov helped Russian authorities with secret testing of his own before their departure to ensure doped athletes would test clean later. His new discovery about turinabol was published too late to use in testing at the games, but was pioneered by a laboratory in Germany soon after. The new technology began to catch dozens of athletes, though not from the Olympics, because those samples had already been processed.
However, Rodchenkov was playing a more complex game. According to his testimony, he had been doping Russian athletes with turinabol but now, as McLaren writes, ''while appearing to be at the forefront of the development of doping detection science, he was secretly developing a cocktail of drugs with a very short detection window.''
Rodchenkov testified earlier this year that turinabol was replaced in the new ''Duchess'' cocktail with trenbolone, a steroid usually used for muscle growth in farm animals. There is still no new test for trenbolone using long-term metabolites, and there have been no positive cases for it in more than 100 IOC retest cases from 2008 or 2012.
Rodchenkov hasn't explained his motivations for publishing his research into turinabol, but Van Eenoo, the director of the Belgian lab, suspects an attempt to catch out Russia's rivals. Many Russians have been caught but other ex-Soviet countries like Belarus and Kazakhstan have seen their Olympic weightlifting programs almost wiped out by turinabol retests.
''If you're really clever, you'd only do this once you've got something else and you want to hit the competition,'' Van Eenoo said.
Rodchenkov's double life and his lurid testimony of steroids dissolved in whiskey and samples swapped in the middle of the night have led some to doubt the credibility of his turinabol research.
''They can't be described as those turinabol metabolites using this method,'' says sports lawyer Artyom Patsev, who represents some Russians who failed IOC retests. ''What could they be? Maybe they're metabolites of a different steroid. Maybe they're metabolites of orange juice.''
Of several lab directors consulted by the AP, some privately expressed reservations about how the original research may have been conducted in the tainted Moscow lab, but all said its findings had been upheld by subsequent work elsewhere. The metabolite in question, they said, could only come from turinabol or ''designer steroids'' based on turinabol but so close to the original they would also come under the same ban. WADA says it's a ''validated method.''
Rodchenkov left Russia for the U.S. before going public with his revelations. The Russian state's position toward him varies. Some officials have attacked his mental health or claimed he is being paid to lie, while others, particularly in law enforcement, have painted him as the ringleader in a conspiracy to force athletes to dope.
In Switzerland, Rodchenkov's public research and his illicit practices have combined to produce the busiest year of Olympic drug retesting ever. More is to come after the IOC said Friday every Russian sample from 2012 would be examined again.
''It has been an interesting period,'' Lausanne lab director Kuuranne said last week of her ever-changing job handling the IOC's retests. ''This, I think, is the spice of any kind of forensic analysis. It's quite a dynamic field.''