Interesting points. If all you are saying is true, I wonder why there has not been more of a backlash from other athletes and coaches. I agree she can't be banned because of the rules as I read them. The only way her records change is if so much evidence comes forward that there is ultimately an admission. Still, it is good to ask for transparency. I want to believe her but the last few months have created doubt. I would encourage journalists to ask for all of the tests and a full explanation. I would politely note that iAAF and others have had some bad moments and it would help the integrity of the sport if the record of a high profile,anti-doping leader, was fully understood.xxxxxxxxxxxxxxxxxxx The records are important. She remains a big figure in the sport. She held herself out as an advocate for clean running. Yes of course. You are 100% correct. Still, she can't be banned imho because of the 2003 rules (and because that was 13 years ago), which is why I made the comparison to the EPO tests of the frozen samples from the 2001 TdF.
rjm33 wrote:
It is very simple, if we use the Sunday times numbers. At the World Championships HalfMarathon in 2003, Paula's hemoglobin level went from 12.0g/dl prerace to 15.6g/dl postrace within a time span of 2 days. Do the math. That is exactly a 30% increase in hemoglobin values in 2 days. The most plausible explanation for this massive, rapidly changing hemoglobin value is a blood transfusion. Variations from body position of blood draw, or timing of test would be small variations of about 0.1-0.4 g/dl, probably something like plus or minus 3% variation , or it wouldn't be useful as a lab test that is ordered everyday by thousands of doctors across the world. These are common, routine lab tests that are done all the time. They are not fancy, complicated lab tests.
[/quote]
roelants wrote:
Interesting points. If all you are saying is true, I wonder why there has not been more of a backlash from other athletes and coaches.
If I can tell that the IAAF is unconditionally supporting her never testing positive, then I'm certain people inside the sport know that they cannot offend the IAAF.
For me, it has nothing to do with if she can be banned. That is over and done, and simply doesn't matter.
The fact that almost nothing fits other facts outside her statement, not even her biography, just makes it clear she's as never tested positive as so many other podiums.
If Paula did the unthinkable and let Seppelt's data be discussed in the media, AND THEN, Ashenden and Parisotto find high probabilities of not-doping, then she is the GOAT, no question about it. But, none of the facts fit together and Seppelt's data set is overloaded with podium dopers.
EPO cheats out Paula
Hey Paula, transfusion cheats out.
rekrunner wrote:
You cannot combine the probabilities like that, as the tests are not completely independent.
rjm33 wrote:...
As another poster pointed out, if these are 1/1000 cutoff numbers (I am not sure that they are), then the odds that a clean athlete could randomly give these lab results on 3 occasions is 1/1000 x 1/1000 x 1/1000 = 1 in a BILLION chance. If we use the 1/100 cutoff, then the odds are only 1 in a MILLION probability.
Why do you say "the tests are not completely independent"?
These probabilities are for a clean athlete who has "normal" blood values. If these tests aren't independent, then a high score on one test makes it more likely there will be a high score on the next test? Why do you think that would be case for a clean athlete with normal blood values?
Remember, the tests are still independent, even for athletes who have higher-than-normal baseline blood values. Having higher baseline values would simply raise her 1/1000 cutoff numbers. (To see this, consider what would happen if we flipped a coin that has a 70% probability of coming up heads. We'd expect to see more heads than if the coin was fair, but that does not mean the coin flips are not independent.)
PS In fairness, rjm33 did point out that that the cutoff numbers could be wrong.
Because it's the same runner, and because some of the same non-doping reasons are common to multiple samples.
If you want to adjust these probabilities for Paula, then the numers are not 1/100 or 1/1000 anymore.
So maybe one.Strange how cyclists died in their sleep, but he died in a race.Liggett said 100s of cyclists. How many runners? 100s?
fred wrote:
Tarno
fred wrote:
Most chilling of all was the comment that cyclists would "live to ride during the day, and then ride to stay alive at night".
Yeah, there are stories of riders suffering mini-heart attacks at night and being forced onto their bikes by team-mates to get the sludge in their veins moving.
rjm33 wrote:
As another poster pointed out, if these are 1/1000 cutoff numbers( I am not sure that they are), then the odds that a clean athlete could randomly give these lab results on 3 occasions is 1/1000 x 1/1000 x 1/1000 = 1 in a BILLION chance. If we use the 1/100 cutoff, then the odds are only 1 in a MILLION probability. WOW! That's a real confidence builder, isn't it Paula and Wejo..
There is some sort of a graph of the blood values at the bottom of this article with the threshold lines superimposed on the plots:
http://sportsscientists.com/2015/09/paula-radcliffe-off-scores-and-transparency/I think these lines are also including some sort of altitude correction. Even with this correction, they are all close to the 1/100 threshold mark.
Why doesn't she just do the lie detector?
Or release the data?
The famous frozen samples? Where are they now?
doping expert wrote:
Why doesn't she just do the lie detector?
Or release the data?
The famous frozen samples? Where are they now?
Paula is great at talking a big game and then not following through. This is a good enough tactic to appease the masses.
That's the thing. If Paula were innocent, and if Coe would really believe it, the data would be made public. Then one could, unofficially following today's IAAF rules, have three independent experts (ideally without anyone involved in the IAAF corruption) discuss the whole set including Paula's (varying) excuses and vote, whether or not she would have deserved a ban.
doping expert wrote:
Why doesn't she just do the lie detector?
Or release the data?
The famous frozen samples? Where are they now?
As high school maths clearly isn't a strong point on this board, let's just deal with the 1/1000 point.
That probability conveys that, of 1000 normal, undoped, healthy people 1 of them would have that OFF score with no drugs involved.
It's a little like a probability of 1/1000 people having both ginger hair and blue eyes. Unusual, but still natural (I am of course making up the stat about 1/1000 people having blue eyes and ginger hair).
Measure that probability on day 1 and that person has 1/1000 chance of having that natural characteristic. Measure that 365 days later and they still have that characteristic. So you can't simply multiply the 1/1000 together.
In relation to the OFF scores, it is the probability that the results obtained on that day are natural. As you are sampling from a single human being, and as the purpose of the OFF score is to establish some physiological consistency, when that score is repeated after period of time it would still get the 1/1000 tag. You cannot multiply the 3 instances of the probability to turn into anything meaningful.
From an anti-doping perspective, all you can say of every single 1/1000 result is that it is a result which is worthy of further investigation as the stats say that it is unusual.
My take on the Ashenden / Parisotto stuff.
I think we can take it for granted that they have seen the data and concluded for themselves about names that jump off the list. So they will have an opinion about Paula. They are choosing not to share that opinion. That may be because they don't want to point a finger at any specific athlete. Or it may be because they are convinced by the IAAF's rebuttal and follow up testing in Paula's specific case. Or there may be other reasons.
But what Ashenden and Parisotto are focussing on (and criticising the IAAF for) is anomalous results not showing sufficient evidence of follow up by the IAAF either from a doping perspective or from a "duty of care" to athletes.
As I've made clear, I see no *evidence* that Paula was anything other than clean. Fast times evidence. Clean tests evidence. Plausibly explained OFF scores evidence.
But equally there are questions which could be answered better.
My point is, however, that the guys behind this investigation appear to deliberately not pointing the finger at Paula. They are staying entirely silent on the Paula issue (I'd obviously prefer them to state if they do accept Paula and IAAF's explanations about her results).
But if they really wanted to bring the IAAF to its knees *assuming they believe the IAAF are covering up Paula's position* they have the data and info to make that public case. The fact that this doesn't appear to be the way they are operating makes me thing that they don't believe Paula is a doper and that the IAAF are covering it up - because what a fabulous story they would have if they could make that case.
I assume you have seen this in response from Ashenden to the IAAF's Paula document:
" "Anti-doping expert Michael Ashenden has accused the International Association of Athletics Federations (IAAF) of attempting to "muddy the waters" by criticising his analysis of leaked data belonging to the world governing body."
and
"The IAAF has released a statement that comprises 25 pages of hair-splitting, plus 13 pages of appendices," responded Ashenden, who assisted the Sunday Times with their findings."
"The irony of a disgraced federation casting aspersions is not lost on me.
"But their deliberately bloated document, no doubt intended to muddy the waters, cannot go unanswered." "
and at the very end:
"Pound has said there will be a "wow factor" when the next part about the depth of corruption within the IAAF is released."
Yes, I've seen and read that. And was struck by the fact that they do not go toe to toe with the IAAF on Paula's case specifically. If you read the IAAF's document too it is clear that most of it is dealing with the more generalised position.
I think the IAAF missed a trick by not separating Paula from the wider case.
You and I may have interpreted the IAAF report and the responses differently, but that is my take on things.
larkimm wrote:
As high school maths clearly isn't a strong point on this board, let's just deal with the 1/1000 point.
That probability conveys that, of 1000 normal, undoped, healthy people 1 of them would have that OFF score with no drugs involved.
It's a little like a probability of 1/1000 people having both ginger hair and blue eyes. Unusual, but still natural (I am of course making up the stat about 1/1000 people having blue eyes and ginger hair).
Measure that probability on day 1 and that person has 1/1000 chance of having that natural characteristic. Measure that 365 days later and they still have that characteristic. So you can't simply multiply the 1/1000 together.
In relation to the OFF scores, it is the probability that the results obtained on that day are natural. As you are sampling from a single human being, and as the purpose of the OFF score is to establish some physiological consistency, when that score is repeated after period of time it would still get the 1/1000 tag. You cannot multiply the 3 instances of the probability to turn into anything meaningful.
From an anti-doping perspective, all you can say of every single 1/1000 result is that it is a result which is worthy of further investigation as the stats say that it is unusual.
Eeek! It's one thing to not know what in the heck you're talking about, it's quite another to broadcast it. The 1/1000 is a function of the equipment and protocol, not the test subjects. Otherwise, why did Paula have only 3 aberrant scores? If her scores were an immutable consequence of her DNA, then all her tests would have been flagged. But, I think you know this.....nobody who can write as well as you can possibly be as ignorant about basic math and science as you are pretending to be.
But just in case, and for the rest of you out there, as long as the tests are done on different machines or on the same machine after it has been recalibrated, the tests are indeed independent. To argue otherwise is akin to saying that multiple rolls of a die are not independent because the same person was doing the rolling.
Bollocks.
The off score is a physiological test. Its a measure of something about that person. Now, I do agree that its not entirely a precise science (i.e. measure me on 3 separate occasions a month apart and you will get different OFF scores). But the variability (adjusted for any environmental factors such as recent health etc) is expected to be low. That's the whole point of the OFF score threshold.
1/1000 is not a function of the equipment at all.
Its the probability that the off scores obtained were obtained from a subject with no reasons for their blood results to deviate from recorded norms. Once the absolute off score is obtained, there are then qualitative or environmental factors which should be taken into account which may explain one or more factors which contribute to the absolute score being obtained. The sorts of factors which are accepted as having SOME impact on OFF scores are things like general health, recent altitude training (possibly), hydration status, recent activity levels etc etc.
Let's hypothesize for a moment: I am a drug free, healthy individual. On a bog standard Monday afternoon, feeling well hydrated and having spent a sedentary day at home, I have my bloods taken. The off score falls within a range which triggers no suspicion. Next week, I'm suffering from dehydration as I've deliberately not drunk anything since Sunday morning, I've just been for a run, and I've also had a stomach complaint over the weekend. I now get an off score which is beyond the 1/1000 threshold.
Next week I am fit and healthy again, and I am tested, back within normal thresholds. The following week, I've had a blood transfusion, taken EPO, etc and now I get an off score beyond the 1/1000 threshold.
Was the second result a 1/1000000 chance occurrence, or a 1/1000 chance occurrence? Its the latter - the circumstances leading to the first test are independent of the circumstances leading to the other 1/1000 test result.
Now let's apply that logic to the Paula situation. She is tested and most of her test results have "normal" range off scores. But three times in her career she off scores in the 1/1000 range. Each of those scores is completely independent of the other, and if there are plausible and verifiable explanations for those 1/1000 scores, then it is simply mathematically incorrect to conflate those three instances into a 1/1000000000 chance of happening.
My grasp of the maths is fine, thanks.
preciously jaded wrote:
rjm33 wrote:As another poster pointed out, if these are 1/1000 cutoff numbers( I am not sure that they are), then the odds that a clean athlete could randomly give these lab results on 3 occasions is 1/1000 x 1/1000 x 1/1000 = 1 in a BILLION chance. If we use the 1/100 cutoff, then the odds are only 1 in a MILLION probability. WOW! That's a real confidence builder, isn't it Paula and Wejo..
There is some sort of a graph of the blood values at the bottom of this article with the threshold lines superimposed on the plots:
http://sportsscientists.com/2015/09/paula-radcliffe-off-scores-and-transparency/I think these lines are also including some sort of altitude correction. Even with this correction, they are all close to the 1/100 threshold mark.
So dehydration is not an acceptable reason?
Tucker says that a hemoglobin count of 12 is remarkable. No it's not. Not in that kind of weather. Plasma volume increase before the race and decrease during the race are completely normal.
larkimm wrote:
Each of those scores is completely independent of the other, and if there are plausible and verifiable explanations for those 1/1000 scores, then it is simply mathematically incorrect to conflate those three instances into a 1/1000000000 chance of happening.
My grasp of the maths is fine, thanks.
You might want to rethink what you wrote here.
Link wrote:
larkimm wrote:As high school maths clearly isn't a strong point on this board, let's just deal with the 1/1000 point.
That probability conveys that, of 1000 normal, undoped, healthy people 1 of them would have that OFF score with no drugs involved.
It's a little like a probability of 1/1000 people having both ginger hair and blue eyes. Unusual, but still natural (I am of course making up the stat about 1/1000 people having blue eyes and ginger hair).
Measure that probability on day 1 and that person has 1/1000 chance of having that natural characteristic. Measure that 365 days later and they still have that characteristic. So you can't simply multiply the 1/1000 together.
In relation to the OFF scores, it is the probability that the results obtained on that day are natural. As you are sampling from a single human being, and as the purpose of the OFF score is to establish some physiological consistency, when that score is repeated after period of time it would still get the 1/1000 tag. You cannot multiply the 3 instances of the probability to turn into anything meaningful.
From an anti-doping perspective, all you can say of every single 1/1000 result is that it is a result which is worthy of further investigation as the stats say that it is unusual.
Eeek! It's one thing to not know what in the heck you're talking about, it's quite another to broadcast it. The 1/1000 is a function of the equipment and protocol, not the test subjects. Otherwise, why did Paula have only 3 aberrant scores? If her scores were an immutable consequence of her DNA, then all her tests would have been flagged. But, I think you know this.....nobody who can write as well as you can possibly be as ignorant about basic math and science as you are pretending to be.
But just in case, and for the rest of you out there, as long as the tests are done on different machines or on the same machine after it has been recalibrated, the tests are indeed independent. To argue otherwise is akin to saying that multiple rolls of a die are not independent because the same person was doing the rolling.
Fine, but do you actually know how much plasma volume varies pre and post race, and how much this affects hematocrit/hemoglobin levels?
