For those who requested a blog, I will just post a summary here, and then address some of "rjm"s points.
I'll post a longish summary now, and then, by popular demand, I will not respond to anything that is not new.
My posts in this thread are about one thing: is doping the ONLY plausible possibility?
To claim that doping is the ONLY possibility is an amazing claim, and showing it is a strong burden requiring "debunking" or at least bounding EVERY possible source of error.
My claim is not radical or heretical: doping always remains a possibility, but simply, with the current knowledge and data, it is not the ONLY plausible one.
I'm not trying to plant new seeds of doubt for an uninformed public looking to chose which leader to follow, or trying to prove Paula is clean, but talking about probabilities, and trying to help put the leaked data in some perspective.
This helps understand why the experts were not concerned, rather than a worldwide collusion between IAAF, UKAD, and WADA, to coverup Paula.
I'm suggesting that we SHOULD EXPECT more fluctations before 2009, i.e. more "false suspicions" and "bigger point swings".
This is based on several rule changes in 2007 and 2009, and a strong presumption on my part that these changes were necessary, and a reaction to observed errors in the past, and that these changes were effective at reducing the probability of false positives.
To suggest that these rule changes were not necessary, and didn't impact the probabilities of false positives is, for me, another amazing claim (without proof).
For the sake of my discussion, I presume the changes were both necessary and effective. If you cannot or have not accepted that, this fundamental difference explains the opposite conclusions.
Can errors happen since 2000? After all, wouldn't want to offend any testers.
When Ashenden studied how to correct for "machine brand" errors, he had to discard 6 weeks of measurements due to a bad machine calibration. Another machine gave completely unusable results for the whole study.
But wait a minute, didn't we leave in a perfect world in 2001-2008, where errors can't happen?
Wasn't Ashenden, the "science at work" methodologist, aware of these issues before the measurements?
What about my claim of 6.0 g/dl?
"rjm" says transfusion is the ONLY explanation.
Yet, in the real world, measurement errors have been observed and reported, due to several factors.
I just picked 3 "actual errors observed" to show that only these factors can explain more than double the swing.
I didn't even pick machine calibration. There are about 4 or 5 more that I could pick.
"rjm" hasn't done his homework. We cannot simply take it for granted, without doing the work.
Again, notice the subtle difference:
I'm saying that multiple errors CAN explain large swings. We need to rule out or bound ALL sources of error to earn the right to say ONLY.
We can do that with data packages required post 2009.
"rjm" tries to do it with ridicule and innuendo -- that's also a tactic.
In Paula's case we have two separate issues:
1) The 3 "suspicious" values
2) The "wild swings" between pre and post race values
Regarding the suspicious values, Paula's 3 values are around the threshold of "altitude" suspicion. Here, small errors due to process can artificially create false suspicion. ("preciously jaded" likes to point out it goes both ways. True "105" scores can slide below suspicion and go unnnoticed. Paula might have up to 11 more "suspicious" scores. Yes it's possible in a micro-dosing scenario targeting a limit. This target didn't exist until 2009. All the more reason to understand the context of the measurements.) Can't we use this argument for the Russians? Yes we can. It's a question of probabilities. Small errors, as well as the known effects of altitude training, can easily push a 102 score up to 109, but is much less likely to push "sea-level clean" scores up to 130-150. We don't need to stretch our minds too far, to see that her marginal scores can be explained by altitude, or that they are too small to be the primary concern -- the priority should have been putting a system in place to catch the Russians.
I won't dwell too much on the suspicious values, as the "real" issue for "rjm" seems to be the wild 33 point fluctuations. (But this was the ONLY issue for the UK media and most of the general public). In any non-doping scenario, I agree it's not plausible that ONLY ONE thing explains the big difference, but (I say) it remains plausible that A COMBINATION of errors in the two measurements, such as a bad measurement pre-race, and plasma loss post-race explains it. To increase the certainty requires external information, not in Paula's possession, and not in the spreadsheet. "rjm": in your 7000 blood reports, were all of them perfectly accurate? How would you know? Did you send anyone out for a 70 minute half marathon in the heat before drawing blood?
The first question, not yet answered, is "is the data real". In a post-2009 ABP world, this is also the first question that experts address. They look at a data package regarding the athlete's recent history, and the data regarding collection, storage, transportation, and data regarding the accuracy of the equipment.
This was not done by anyone for the samples betwen 2001-2008. So conclusions based on these samples must be taken with a grain of salt, always keeping in mind process errors have not been considered, estimated, bounded, when publishing the data, and drawing conclusions. Not by Ashenden, Parisotto, Seppelt, Walsh, or rjm.
"rjm" seems to think that they've already done all the work over and over and over to debunk all sources of error, in there series of rebuttals, and by process of elimination, only transfusion is plausible.
No one in these 1200 posts so far has acknowledged or debunked errors due to sources like machine brand or machine calibration.
No one has showed any data about the collection, transport, storage, or analysis of the blood samples.
To earn the right to say ONLY, you must debunk ALL. This could be done by a statistical analysis, but hasn't been done yet.
The strongest argument here is of the type "don't insult the testers" or "scientific consensus". It's not an insult if the machine brand causes a large error.
What about altitude? Don't low pre-race scores debunk altitude? Maybe, if the values are real.
Yeah but Paula raced for 4 weeks -- she wasn't at altitude. Maybe. Maybe she went back during the week, or maybe she stayed in her altitude tent. The status is "need more information", not "completely debunked".
What about hemo-concentration? Surely that's "completely debunked" now.
If I have to be careful to not commit "bad science" projecting results to Paula, so does everyone else, including "rjm". It is correct to ask if a Florida marathon in February with 4 hour volunteers applies to Paula's 67 minute semi-marathon in hot conditions.
My original claim, to explain the 2-hour rule was 1) Schumacher tells us 2 hours, and 2) Ashenden/Parisotto says "in the order of 10-20%"
After "Plasma Rubbish"'s excellent explanations, both of these claims survived.
Didn't I lie by omission, by not highlighting 0.45 g/dl?
I don't see what this has to do with 2 hours. Schumacher found some plasma loss, not the maximum, and confirmed it came back in two hours. During exercise, we see from the few studies highlighted here, that plasma expansion, and plasma loss are possible -- everything is possible, but it depends on the circumstances.
A finding of 0.45 g/dl in the Schumacher study doesn't "debunk" that Paula's case could have been up to 10-20%.
And I was not using Schumacher to state a claim about the magnitude of Paula's hemoconcentration.
And it is not relevant to the reason for the 2-hour rule - the only point which I bring up Schumacher.
And it is not relevant since you've seemed to accept that you, and science, and everyone agrees that a reasonable maximum is 10-20%, despite other studies in different circumstances not reaching this maximum, and some finding plasma expansion.
To argue 10-20%, I gave a quote from Ashenden.
Plasma Rubbish says, yeah, for cycling but not running.
But then how to explain the running study?
Oh, that's for men, not women. And it was caused by standing up and not running.
Oh and Paula drank, but the men didn't, plus scientific consensus, and women sweat less.
This could go a few more rounds, but no one seems to have the appetite:
Plasma Rubbish previously said fluid intake didn't impact hemo-concentration.
Isn't hemoconcenration about more things than sweating?
How much do you sweat when you stand up? Or when pumping your fist? Or when wearing a tourniquet?
Don't women stand up too, when they run?
It seems like the only way out is a study on elite women in the heat.
If you want to rely on the leaked data to confirm suspicions, you need to show that reliability of comparisons is valid.