Why does Mary Slaney get a pass????

uranidiot wrote:Her international ban was upheld more because of the need for consistent rules than because it was scientifically correct. IAAF admitted as much.

When did they do this? Puleeze?

Dude, re-read my post. I agreed with you.

Look in the mirror boy wrote:

When did they do this? Puleeze?

Hahahahahaha. Did you even read the stuff at that link you posted? LOL.

Although I'm somewhat inclined to believe that Slaney used exogenous testosterone late in her career, I'm skeptical about allegations that she was a doper in the mid-80s, when she was winning world championships and setting world records. That's a much different situation from that of either Jacobs or Jones.

Undoubtedly, however, there are different standards for different athletes, especially when national loyalties are thrown in the mix. As I recall, Al Oerter admitted using anabolic steroids at some point in his career, yet he has been elevated to sainthood in the sport within the U.S., and is generally regarded as having been an outspoken critic of doping. Conversely, Lasse Viren's record on doping appears spotless, yet many outside Finland have assumed his guilt.

Perhaps Slaney has been given too much of a pass regarding her elevated T/E ratios in the mid-90s, but I tend to believe that many have also been far too harsh on her for the debacle of 1984. In my opinion, she was a clean athlete competing successfully against some very dirty athletes, and one of the most extraordinarily gifted runners ever on the world scene.

Imagine the deification she would have by now if she'd died in a car crash at the end of '83

Avocados Number wrote:

I'm skeptical about allegations that she was a doper in the mid-80s, when she was winning world championships and setting world records. That's a much different situation from that of either Jacobs or Jones.

Why? Because she wasn't caught in the 80s. This is not new speculation, this idea that she was cheating in the 80s.

Oh, and Viren was suspected of blood doping, not taking drugs.

Jay Dellmore wrote:

[quote]Avocados Number wrote:

Oh, and Viren was suspected of blood doping, not taking drugs.

I wonder how much benefit blood doping in the 1970s really provided. They didn't pack the blood as is now done.

There's no good reason to believe that Slaney competed clean at any point in her career. That's the pall of doubt that is rightly cast when one is caught doping, nobody can ever be sure exactly when the doping started. Does anyone really believe that Marion Jones's doping only went as far back as Jones claimed it did? (Remember, Jones was never officially caught and sanctioned for doping by USADA, WADA, USATF, IAAF, USOC, or IOC.) Does anyone really believe that all of Jacobs's performances prior to her bust were clean? I'm not sure why people want to treat Slaney any differently, except that she was a phenomenal, white, American female. I think SFH gets favorable treatment in a similar (though of course not identical since she never tested positive that any of us know of) light. She's a cute, blond, American female yet she was competing with and beating Europeans who were probably doping as well. Russians, Romanians, etc. were and still are evil (cold war remnant prejudice) but our sweet, cute, smiling, bubbly girls hold their own against them with nothing more than apple pie in their systems.

Who was Slaney's coach in 1983? Dick Brown, right? The same Dick Brown who was caught with a $50k slush fund from which drugs and drug testing were being funded.

Jay Dellmore wrote:

Oh, and Viren was suspected of blood doping, not taking drugs.

Oh, gosh! Thanks for pointing that out! For all of you who were misled by my reference to "Viren's record on doping," I should clarify that what I actually meant was . . . well . . . "Viren's record on doping."

White Indignation wrote:

There's no good reason to believe that Slaney competed clean at any point in her career.

One of the sillier statements on this site. I'm guessing that you don't remember, as I do, little Mary Decker when she was fourteen years old.

Like you, I don't know whether Mary was a doper in the '80s. But if anyone on the planet could set world records and win world championships in those days without doping, it was Mary.

Your response was certainly sillier. The point, since you completely missed it, was that you can't point to a moment in time when Slaney went from clean to doper. At any rate, being 14 (like being white, female, and attractive looking) doesn't matter a hell of a lot. It's not clear that Marion Jones wasn't doping in high school. You've only stated a belief based on a gut feeling. Thanks for that?

White Indignation wrote:

The point, since you completely missed it, was that you can't point to a moment in time when Slaney went from clean to doper.

Your statement was that "There's no good reason to believe that Slaney competed clean at any point in her career." That's a really dumb thing to say. If your "point" was merely that I "can't point to a moment in time when Slaney went from clean to doper," then perhaps you should have said so. It wouldn't be a very interesting point, but at least it would be something we could agree about.

I know what you think, but I don't know why you think that way except perhaps out of a strong sense of nostalgia. Because you didn't come up with a "good reason to believe that Slaney competed clean at any point in her career." We know she doped at one point and we don't know when the doping exactly began. You presume that she wasn't doping at a certain point and you support that with fond recollections from your hazy past of only what you could see with your own eyes. That's an irrational bias, no matter how you try to flame away at things you can't agree with or wrap your head around.

Actually, we don't know that she ever doped. At most, we have information that, at some point in her late 30s, tests showed T/E ratios that, under the rules of the day, created a rebuttable presumption that she elected not to contest in subsequent proceedings.

But more to the point, I have very good reason to believe that Mary was competing clean at some point in her career. Mary's first cross-country race -- which she won by a large margin -- was in 1969, when she was eleven years old. It is neither "a strong sense of nostalgia" nor "an irrational bias" that causes me to believe that this 11-year-old girl was competing clean in 1969. It is, rather, a very rational belief based on, among other things, the paucity of evidence indicating that sixth-grade girls in Huntington Beach, California, were engaged in athletic doping back in 1969. Mary's elevated T/E ratios over a quarter of a century later provide approximately zero evidence of any doping when she was a child.

The situation in the early 1980s was much different from the situation in the early 1970s, when Mary first competed internationally. At the world-class level, female middle-distance runners in the 1980s were a suspect group. I would be surprised to discover that any of the world-class female runners from Eastern Europe were clean. I would also be surprised if any of the world-class female runners coached by certain individuals in the U.S. were clean. Mary's associations with some individuals may raise legitimate questions. But if there was one female middle-distance runner who was capable of being the best in the world without doping, it was Mary.

Flip of a Coin wrote:

I don't think so. I am using Bayesian statistics.

http://en.wikipedia.org/wiki/Bayesian_inference

You can use this calculator if you want to test out my maths.

http://psych.fullerton.edu/mbirnbaum/bayes/BayesCalc.htm

Prior probability of being a cheat = .10

probability of positive test given using drugs = .90

probability of positive test given no drug use = .1

Probability of being a drug cheat given a positive test = .50

Actually this is wrong.

The original author gave us the following information:

"The T/E ratio catches people that actually are using testosterone 90% of the time."

This is equal to P(positive test given using) = 0.9, let's call this P(+|U) = 0.9.

"Lets also say that at that time only 10% of athletes were actually actually using testosterone."

This is equal to the prior probability of using, P(using) = 0.10, let's call this P(U) = 0.1.

Bayes' theorem states (with Us and +s):

P(U|+)=P(+|U)*P(U)/P(+)

where P(U|+) is the probability that you were using if you test positive. (This is what we're interested in.)

What we don't know (but must know to evaluate Bayes' theorem) is P(+), the probability of testing positive.

The previous author falsely assumes that P(+) = 0.18. This is based on the assumption that P(+|U) + P(+|C) = 1.0 (thus P(+|C) = 0.1). If this was true, then we could evaluate P(+) as P(+) = P(+|U)*P(U) + P(+|C)*P(C) = 0.9*0.1 + 0.1*0.9 = 0.18. This would make Bayes' theorem P(U|+) evaluate to 0.5, as the author claims.

In fact, conditional probabilities don't need to add to 1.0 like this (rather, P(+|U)*P(U)+P(+|C)*P(C)+P(-|U)*P(U)+P(-|C)*P(C) must equal 1.0, assuming +=positive, -=negative, U=using, and C=clean are all the possible outcomes in the event space). To find P(+) we need P(+|C), which we can't find in the given information. (The current author claims that it is "probability of positive test given no drug use = .1," but as I said, this is built on the false assumptiont that P(+|C)=1.0-P(+|U). In fact, it is easy to imagine a test that provided totally no information about drug use in which P(+|C)=P(+|U); or, alternatively, one that was ineffective, i.e., low P(+|U), but which an innocent person never failed, P(+|C)=0.)

So, although I agree Bayes' theorem is correct to use here, we are missing one vital piece of information that we need to evaluate it - the false positive rate!

If there are any Bayesians out there, please feel free to correct or ammend!

As an addendum to this, any result based on Bayes' theorem relies on the specification of a prior - which we often don't know. For example, the previous author assumes that P(U) = 0.10, but most information about drug use actually comes from testing, which probably vastly underestimates the prevalence of use in the sport.

Consider two examples. First, let's assume that drug use is relatively rare, say P(U) = 0.10. We have a reasonably effective test P(+|U) = 0.9, with a false positive rate, P(+|C)=0.05.

We first calculate P(+) (now based on enough information to do so) as:

P(+)=P(+|U)*P(U)+P(+|C)*P(C)=0.9*0.1+0.05*0.9=0.135

We can then evaluate the posterior probability that someone with a positive test is indeed cheating as:

P(U|+)=0.9*0.1/0.135=0.667.

This means that a full 1/3 of the time someone tests positive based on this test, they have not cheated.

By comparison, if we take the more pessimistic view that cheating is quite prevalent (say 75% of athletes), then:

P(+)=P(+|U)*P(U)+P(+|C)*P(C)=0.9*0.75+0.05*0.25=0.6875

and

P(U|+)=0.9*0.75/0.6875=0.9818.

In this case, because cheating is so prevalent, the probability that someone who tests positive is cheating actually exceeds the "positive rate" of the test. This is because its so likely to begin with that you are cheating, and a positive test only reinforces that probability.

This makes intuitive sense too because P(U|+)->1.0 as P(U)->1.0. (I.e., as the proportion of athletes using goes to 100%, so does the proportion of athletes with a positive drug test who are guilty - everyone is guilty, so how could they not be?!)

LR's posts have great potential in killing this thread...

AN's sentiments, while clearly based on nostalgia, are as good of an explanation of why Mary get's a pass as any other post on this thread.

We know the Russians were drugged. If there was one female on the planet that could compete with them clean it was Mary - she was an undisputable talent. It was pure joy to watch her run. Does that make her any less likely that she was a drug cheat? No.

Now what is the likelihood that Mary raced clean against them; with her failed drug test, her competition and the seemingly unchecked drug culture of the 80’s? It may be so low that even Bayes couldn't quantify it. Having a gut feeling that would purport to overturn these factual and circumstantial pieces of history are woefully inadequate.

As the tribunal said, the burden of proof shifted onto Mary’s back and she chose not to shoulder the load – drug cheat.

A number of the later posts on this thread have been deleted; I'm not sure why. Perhaps they would have helped you to understand my views about these matters.

I don't think that "nostalgia" is the proper term for an assessment based on a recollection of historical events as well as a review of hard data. As I pointed out in one of the deleted posts, Mary was a 4:37 miler when she was fourteen years old, standing five feet tall and weighing 89 pounds. She was already a world-record holder at the age of fifteen. These performances were achieved in 1973-74, and no one has produced the slightest bit of evidence, anecdotal or otherwise, that they were drug-enhanced.

Although I'm not taking a position one way or the other about whether Mary was a doper ten years later, in 1983-84, Mary's extraordinary talent does, in fact, make it less likely (although obviously not dispositive on the issue) that she was a drug cheat when she was winning world championships against drug cheats. One needn't have any formal background in Bayesian probability theory to understand that. And the absence of any extended articulation of that understanding on a message board doesn't transform it into a "gut feeling."

What you've described as a "failed drug test" a dozen or more years after her peak performances doesn't shift the burden of proof back to Mary to demonstrate that her peak performances were clean. That's true as a matter of IAAF and USATF rules, rules of evidence throughout the U.S., and probabilistic analysis. The purported "failed drug test" is simply one additional datum, the significance of which can be disputed. A contrary view is merely a moral position, not a scientific or mathematical one.

These points do not, however, directly address the original question, as to why Mary "gets a pass" -- if, indeed, she does. One could certainly take the view that one doping infraction or admission at any point in one's career should wipe that individual's name from the record books and invalidate that individual's entire athletic career. That view, if widely accepted within our sport, would have some interesting consequences, but it is not the official position of the sport's governing bodies, and is not even a vew that is as widely held as some apparently think. People can, and do, make more complicated assessments of careers and lives. In Mary's case, one or more individuals on this thread have trotted out certain explanations based on racial and sexual assumptions. I don't reject those explanations on grounds of social policy or morality, but I do question their accuracy. (Consider, for example, doping assumptions about world-class runners from Eastern Europe -- largely, if not primarily, blonde white women. Contrast those assumptions with assumptions about Kenyan men. John Ngugi, for example, was suspended for a "failed drug test," but I don't believe that he was a doper, while I operate upon a working presumption that any white female 100-meter runner at the world-class level in the 1980s was a doper, whether or not she failed a drug test or admitted doping.)