A new bit of information has surfaced over at Trust But Verify (tip o’ the hat to regular reader Larry, who pointed out this article in a comment a short while ago) related to how France’s anti-doping laboratory, LNDD, conducted the tests which are at issue in the doping case against Floyd Landis. “Mr. Idiot” posted a comment there earlier today, alluding to the fact that the lab used two different machines to perform the GC/MS and the GC/IRMS tests that, according to the majority of the arbitration panel, proved Landis’ guilt.
As TBV points out, in paragraph 188 of the majority’s opinion, they state:
The GC column is, of course, the same in both instruments.
Apparently, that is not the case. It turns out that the GC/MS was performed on an Agilent 19091S-433, while the GC/IRMS tests were performed on a DB-17. Both units are manufactured by the same company (Agilent, formerly a part of HP), but there are differences in the experimental conditions between the two. The DB-17 has a different polarity than the 19091S-433, which has the potential of causing the various ions/metabolites being measured to move through the column in a different sequence for the carbon isotope ratio tests as for the gas chromatography/mass spectrometry tests.
What’s not known, at this point, is whether that would be the case with the metabolites of testosterone which are at issue here. On examination of the A and B sample data from Landis’ lab documentation package, it appears that LNDD performed the tests the same way for both samples. But there are also indications that even LNDD’s own standard operating procedures required the tests to be conducted using the same type of column, under identical (or as near as possible) experimental conditions.
What does “Mr. Idiot’s” revelation mean? Well, for starters, it turns out that there were greater differences in the experimental conditions than previously thought. This casts a greater degree of uncertainty over the results that LNDD reported. Especially if it’s true that their own standard operating procedure required the use of the same column for both types of tests.
What is true, from what Mr. Idiot points out, is that the majority’s assertion about the same instrument is incorrect. Even after four months of studying the testimony and exhibits, even with the help of their science adviser, this fact escaped their notice. It may have even escaped the notice of Landis’ defense team before and during the hearing. Certainly, this is not a line of defense that was mentioned in May. And it’s hard for me to believe that it wouldn’t have been, had they known about it.
Whether or not this development will be instrumental in Landis’ appeal to the CAS is yet to be seen. And whether or not this kind of difference can be explained away by the lab is yet to be seen. In the meantime, it raises a number of interesting questions about LNDD’s reported results and what the truth behind those lab results really is.
Take the time to read what’s posted at TBV. Surprising that such a detail escaped notice for so long. And it was hiding in plain sight all along.
Rant,
– Great detective work – KUDOS to EVERYONE – Mr Idiot, Larry, TbV and every other person that was involved – no sarcasm intended — My belief in sincerely involved good minds at work — can be a devastatingly good weapon — for truth.
– Have read “required reading” – I realize this has to be “deconstructed” analyzed, etc. My limited understanding of the science end of things, gives me a “mere glimmer” of understanding. I admit it — the data being discussed is beyond me — what I do see is that “someone” knew this in the opposition camp — and it was “intentionally buried,” merely to “justify” the LNDD findings, to “get” a “high-profile” rider for doping. This seems to also point to how the “good guys” under the justification of fighting “doping” have lost their way in that struggle, by applying fuzzy “questionable methods!
– How ironic! No wonder there was such a strong “spin” with the trigger as. “Getting off on a technicality” being used by the accusers – THEY KNEW of this! And did everything to try and keep it buried — expect an onslaught of attacks — with CIRCULAR THINKING against Mr Landis. (Personal comment: Floyd — Can’t wait to see you racing again!)
– IF THIS METHODOLOGY is being applied at LNDD — EVERY test that this lab has produced is now mute — Not just the Landis case. I hope people realize the implications here. This little “technical” FUDGING — Doesn’t matter if done by “accident” or by “intent” to get – “right results” as a certain Australian WADA rep has recently publicly put it — they have shot themselves right in the head!
– I don’t think my response here should give anyone the impression that I am gloating — I am not.
– What I am saying is that “applying” the fast and loose” rules and practices” in fighting CHEATING & DOPING — WADA itself has screwed all their own efforts, their results, their “successes” to now! My hope is that — Dick Pound’s Legacy will finally be shown for what it has been — dishonest.
– I think about all those honest individuals who have put in monumental efforts at combating doping and cheating — this new finding — will blow all of that out of the water! No one should feel, I don’t, happy about this new development in the Landis Case. I am very glad for Floyd — I have always maintained that he is a real champion. But this new finding is catastrophic in the fight to “clean up doping in sports.”
– What it also VALIDATES and points to “clearly” is what people on TbV and the Rant Line have been taking apart in a public forum and saying– THERE IS VERY GOOD REASON AND LOGIC behind the concept of INNOCENT TILL PROVEN GUILTY!
– Now is the time, for me at least, to be patient — waiting on better minds then mine to “look” at what this new “bombshell” development actually indicates — my average guy understanding should not be used as valid — So – Mr Idiot, Larry, TBV and every other person that is involved — I am waiting on you all to show me logically, scientifically and clearly — what this means in layman’s terms. You have until the end of December — all papers and findings must be in CAS by January. Go get them — my sincerest hope is going your way, always. Thank you.
Mr. Idiot here. I just want to make sure that everybody knows that this information about the different columns comes from Arnie Baker. He is the one who gave me the information. I deserve no credit.
About whether there was any ill intent by USADA and LNDD on this specific issue: I’m not sure about that. While I do fault LNDD for some intentional misconduct (I can’t imagine any legitimate reason for erasing the hard drive just before the reprocessing was done), I suspect that this use of different columns was a reasonable mistake based on someone’s inadequate understanding of how the whole process fit together. Simon Davis was amazed by how little the technicians understood about the big picture, and I suspect this use of the wrong column in the GCMS is another example of that.
Rest assured that Landis’ team is working on this for the CAS appeal. It came from Arnie Baker. I do think that they just missed it before the first hearing, probably because they really didn’t have that much time to examine the documents of the reprocessing, which is one of the links in the chain that proves the violation of their own SOP.
syi
Thanks Swimyouidiot.
What about the retesting of other samples, was it a bad column used too?
Yes, same setup of columns in all tests. It was a document in the results of the other tests that revealed that the SOP requires them to be the same column.
See today’s TBV discussion “Different Columns – detailed” at http://trustbut.blogspot.com/2007/12/different-columns-detailed.html
TBV
Mr. Idiot,
So noted. But you were the person who gets credit for making the first posted reference to it — at least in my book.
Everyone,
The new article at TBV is very much worth your time. It goes into the issue in more detail than the original post, and highlights a number of questions that this discovery raises. Take the time to look it over.
Morgan, here’s my best and simplest understanding of the science, and I’m happy to provide additional details:
To understand the science here, you need to consider three different machines used at testing labs like LNDD: there’s a gas chromatography machine (GC), a mass spectrometer machine (MS), and an isotope ratio mass spectrometer machine (IRMS). It was LNDD’s job to use these machines to analyze testosterone “metabolites” present in FL’s urine. (The human body metabolizes (breaks down) testosterone molecules before these molecules are peed out of our system, and these break-down molecules are called testosterone “metabolites”.)
Let’s discuss the GC first. The purpose of the GC is to take a complex mixture such as urine, and separate out the substances found in the mixture, so that each substance can be separately analyzed. The process begins when an athlete’s urine sample is injected into the GC machine. The GC machine does its magic (more on this “magic” in a minute), and if all goes well, the GC will “spit out” a blob (or “peak”) of each substance in the mixture, one substance at a time. The peaks we care most about are the peaks containing FL’s testosterone metabolites — this is the stuff that LNDD needed to analyze. If all goes well, the GC will spit out a peak for each of these metabolites. If all goes well, each of these peaks will be separate, so that they can be separately analyzed. And if all goes well, each of these peaks will be “pure”, meaning that each peak will contain only the metabolite we want to measure.
Notice I said that the GC will do its job “if all goes well”. The GC cannot always separate out substances the way I’ve described. Some mixtures, like urine, are too complicated — there are too many different substances in urine for the GC to sort out. So, the lab has to take two steps to help the GC do its job. The first is called “sample preparation” — the lab performs some chemistry on the urine to simplify it, to remove substances from the urine that the lab is not interested in analyzing. Second, the lab sets up the GC to make it most sensitive to separating and analyzing the substances that it IS interested in analyzing.
The most important aspect of this GC setup is the selection of the GC “column”. The GC column is a long, thin tube that is coiled inside of the GC. The injected mixture (in this case, the athlete’s urine) is heated and pushed through this tube. The inside of the tube is coated with a substance that slows down certain molecules in the mix more than others. This “slowing down” process is what allows the GC to separate the substances jumbled together in our mixture into peaks that each contain a single substance. The GC can be set up with hundreds of different kinds of columns, and these columns can use different kinds of coatings designed to separate out different kinds of molecules. In the FL case, the lab would want to select a column that’s best at separating out molecules of testosterone metabolites.
OK. With this description of the GC, we’re now ready to look at the two OTHER machines I mentioned in the beginning of this post, the MS and the IRMS. Why do we need these two other machines? Because the GC (as used at testing labs like LNDD) does not actually ANALYZE the stuff in an athlete’s urine sample. The GC’s job is to separate the stuff of interest in the sample into pure peaks containing a single substance. These peaks are then “spit out” of the GC machine into a second machine for analysis. The MS and the IRMS machines are the two machines used by LNDD for this analysis.
Let’s consider the IRMS machine first. To use the IRMS machine, the lab technician connects the IRMS machine to the GC machine, and then injects the sample into the GC machine. If all goes well, the GC machine will separate out the substances in the sample, and spit out each substance as a pure peak into the IRMS machine. If all goes well, the IRMS machine will then analyze these peaks to see if the peaks appear to come from an “exogenous” (artiificial) source.
But there’s one problem with IRMS testing: it can tell us whether a peak is exogenous, but it can’t identify the peak! It’s a little bit like you’ve opened a bank near a mine, and the local miners have deposited with you the bags of the stuff they’ve mined. Your rule is that you’ll only accept bags containing a single substance, that you need overnight to analyze each bag, and you’ll pay the miners $800 an ounce for each bag of gold, $40 an ounce for each bag of silver, and $200 an ounce for each bag of platinum. Then you hire a guy to analyze the bags for you. The next morning, the guy gives you the results of his analysis: “bag 1: 40 ounces, bag 2: 85 ounces, bag 3: 70 ounces …”. That’s great information, you DO need that information. Unfortunately, you don’t know what’s inside of each of these bags, or whether the content of each bag is pure. This is the problem with the IRMS: it measures whether each GC peak appeared to come from an exogenous source, but it does not tell us what is inside each peak or whether the content of each peak is pure.
So, a lab like LNDD needs to conduct a test BEFORE it does the IRMS test, to determine the content and purity of each GC peak. This pre-test involves the use of a third machine, the MS machine we mentioned earlier. The lab hooks up the GC to the MS machine, injects the sample into the GC, and (if all goes well) the GC spits out pure peaks of testosterone metabolites for the MS to analyze. The MS can be used to identify the peaks and to determine that the peaks are pure. (Going back to our analogy, the MS can determine what’s in each of the miner’s bags and that only a single substance is contained in each bag.) The lab then CAREFULLY compares the peaks in its GC-MS tests to the peaks it obtains for its GC-IRMS tests. If the comparison is accurate, then the lab has all of the information it needs to know for each peak: it knows (1) what metabolite is contained in each peak, (2) whether the contents of each peak are pure and (3) whether the peak appeared to come from an exogenous source.
Again, notice I said that the lab needs to CAREFULLY compare the results of the GC-MS test and the GC-IRMS test, in order to correctly combine the information learned in both tests. Going back to our analogy about the bank, you had one of your employees weigh each miner’s bag, and the employee gave you a sheet reading: “bag 1: 40 ounces, bag 2: 85 ounces, bag 3: 70 ounces …”. Then a different employee analyzed the contents of each bag, and gave you a sheet reading “bag A: pure gold, bag B: pure silver, bag C: pure platinum …”. You have all the information you need EXCEPT that you need to figure out a way to match up the lists you received from your two employees. Is bag 1 the same as bag A? Maybe. (It would help to know if your two employees analyzed each bag in the same order!)
A testing lab has the same problem as the banker in our example: he has information for peaks analyzed in the GC-MS test, and for peaks analyzed in the GC-IRMS test, but now the lab has to match the peaks in the GC-MS test to the peaks in the GC-IRMS test. In many ways, the match is more difficult for the lab than it is for the banker. The peaks in one test may not look like the peaks in the other test. Worse, there might be more peaks visible in one test than in the other!
How is a testing lab supposed to do this match up? This was one of the most controversial issues in the FL arbitration. For the moment at least, I’ll duck this issue, and say simply that the experts called by FL disagreed with USADA’s experts on this point.
However, regardless of whose experts you might believe on this point, everyone would agree that it’s important for the lab to carefull compare the results of the GC-MS test to the results of the GC-IRMS test. Given the importance of this comparison, a testing lab should do all it can to make certain that the two tests CAN be compared. In other words, you want to do everything you can to set up identical conditions for the two tests. In particular, you want the setup of the GC in the GC-MS test to be as close as possible to the setup of the GC in the GC-IRMS test.
We can now finally discuss the information disclosed here under “Even More Required Reading”. What we’ve learned from the information revealed by Dr. Baker is that LNDD did not set up identical conditions for the GC-MS test and the GC-IRMS test. They didn’t even come close. They used two different GC machines for these tests, and the GC machines each used a different column. As a result, there’s even more doubt than before that the GC-MS test can be compared to the GC-IRMS test. Going back to the example, it’s as if our banker received the reports from his two employees and realized that the report of one employee could not be matched up to the report of the other.
Morgan, I’d like to know if this all makes sense to you. I’m working on a long article to publish at TBV, so if you can help me polish this description by asking me questions, I’d appreciate it.
Larry:
This is the best explanation in layman’s terms I have seen. I have a much better picture now of how these tests are conducted and what each machine does.
Larry — your explanation is quiet clear.
In essence — the really important “action” with handling the sample is that the tests be recorded. This recording is not optional.
Each aspect of each test must be notated so that in the end — who ever (the tester) is trying to find out the results — can match the bag/sample with the tests run on it. ESPECIALLY when more then one test is being run on the sample bags.
As you explain — the problem arises for the “tester” when he has no way of putting together which test was run on which bag. Also what the sought after substance” was being tested for.
It is not a question that the test was run — but rather that it becomes possible for the “tester” to “identify” and match up the test with the samples, if the procedure is not notated and followed. Am I correct?
Ergo — a problem arises that if the SOP is varied and not notated — than whomever draws conclusions about the result is only guessing — since he/she has no way of knowing which test results go with which bag, not only that — but is the “result” for the gold test or silver test or the titanium test? Therefore trying to “match” against the control chart — numbers or graphs — becomes nothing but “guessing” on the rationale that the numbers “look like” they belong to a number group or the graph figure from the test “matches” closest to what exist on the control chart — the tester is guessing – correct?
This becomes even more lost in confusion when the testing requires 2 separate devices that have different jobs to perform — to get a clear result.
The first test is done to rid any extraneous material (the substance contains a wide variety of content) – the first test is to filter out or “get rid of substances other then what the tester is looking for.
When this is accomplished — and the tester has not lost track of what he/she is testing — gold results, silver results or titanium then –
Then the second test is run to find out “what percentage of the sought after substance” actually exists in the “cleaned sample.” Therefore if multiple tests are being run and the tracking of the steps is not done linearly — one step after the other and notated — then it is unavoidable not to lose what sample is being tested.
It gets even more complicated when the testing is for multiple sought after substances is being performed and notation is not followed (SOP).
From how I understand this — the tester will get results yes — But he/she has no way of knowing whether the test was for gold or silver or titanium.
It is therefore logical to assume that the testers have some kind of “chart” that shows what a certain amount of “sought after substance” should ideally look like, if it is a chart in the form of a “graph” then a visual comparison is made. If the chart is “numbers” then the numbers are compared from the test result with the chart numbers — which seems more direct then a graph form or shape.
But it would rally get crazy at this point.
Whatever chart was being used for comparison — If the tester lost track of which test for which substance is being handled and “looked at” to compare to the control chart. The tester could not know which “sought after substance” was in their hand and being compared. Is he/she comparing the “gold” results or is it the “silver” or is it “titanium?”
At this point — the tester might be tempted to match numbers or graph shapes by comparing their “look” or their numerical similarity to the control chart — but this is not more then “guessing” — since he/she has no idea what the numbers or graphs are from — gold, silver or titanium.
It isn’t a question of the tester being honest or not — it is more a question of did he/she keep track of the individual process so that he/she knows that the test result is for gold, silver or titanium. Therefore if at any point the tracking is lost, or altered and this means he/she CANNOT know what results he is comparing to the “chart.” He/she is merely making matches on the control chart that look similar to what may be the actual test results — of gold, silver or titanium — but since he/she can’t identify is it titanium or gold or silver — then there is no way to be sure of the results.
Sorry it took so many words Larry — but this is how I understand what you wrote — do I understand it correctly? It is not as clear as I would like it — but I figured you were looking for my “instant response” So I have written it as I was understanding what you wrote — without polish. I hope this helps. I am certain I can make the reasoning structure clearer and perhaps less cluttered.
Your explanation is actually very clear Larry – Clearest of all is just at how many points the tester can lose track. All this – even before the “results” would even be handed to the banker.
Thanks Larry, that cleared a lot of things for me at least!!!
Very much appreciated!!
Larry,
I’ve been reading TVB Article II – “Even More Required Reading” What jumped out at me is that – “LNDD SOP required that the DB-17ms be used for both GC/MS and GC/C/IRMS, but for some (undocumented) reason they used the HP-5ms.”
tbv then makes this comment – “Interestingly, in a chart in the FAQ section of the site (regarding choosing the correct column), Agilent lists steroid identification for the DB-17, but not for the HP-5.”
Does this mean that LNDD made the test using the HP-5 which has no column identification for steroids? But then what column would they use to “check against – the only answer has to be that they had to apply DB-17 Column
tbv additionally states – “Also note that USADA 104 identifies the MAN-52 as being a SCAN, where other GC/MS tests are identified as SIM. That implies to me that the test is designed to collect a full-scan, where others are to use selective ion monitoring.
Does this mean that the wrong machine was used to do the “getting rid of” extraneous material – other then what was being tested for?
Bill, Sara, Morgan, thank you.
Morgan, for the moment I’m responding only to your first post. Your comments there are helpful. I can see that my analogy to the miner’s bags is confusing you. The issue is not so much how each test is recorded. The issue is how each test is actually performed.
OK, this is helpful to me. Let me try again.
Let’s say that we own a casino. At the end of every day, we have a huge bag of different colored chips that have been received at the cashier: the red chips may be worth $5, the green chips may be worth $25, the blue chips may be worth $100, and so forth. We have to total the amount of these chips. So we make it the job of one of our night employees to put all the chips through a chip sorter and to report back to us. The chip sorter is similar to the coin sorters you may have seen in stores and banks: you put a pile of chips in the top, and the machine sorts the chips into stacks, and each stack is supposed to contain chips of a single color. The chip sorter is like the GC.
The employee does this job, analyzes each stack of chips, then returns all the chips to the bag. The next morning, we get the bag along with the employee’s report: there were 340 chips in stack 1, 195 chips in stack 2, 450 chips in stack 3, and so forth. But the employee tells us something we did not know: the employee is color-blind. He cannot tell us anything more about each stack of chips than how many chips were in each stack! Well, what can we do? Our employee has given us important information, because we needed to know how many chips were in each stack. But obviously, this is not all we need to know. We need to know the color chip that was in each stack, and we also want to make certain that each stack contained chips of only a single color. This employee’s report is similar to the report the lab gets from the IRMS: it gives important data on each peak that emerges from the GC, but it does not identify what’s in the peak or tell us that the peak is pure.
So we go to a second employee, give her the bag of chips and ask her to analyze the chips and provide us with the information we’re missing. She protests; she’s not good with numbers. We tell her that we already know how many chips were in each stack, we just need to know the color of each chip in the stack. The employee comes back with the bag of chips and says, stack 1 contains only red chips, stack 2 contains only green chips, stack 3 contains only blue chips, and so forth. The second employee’s report is similar to the report the lab gets from the MS. We now have all of the information we need to figure out the value of the chips in the sack, so long as we can match up the reports we’ve received from our two employees.
How will we match up the two reports? We don’t know for certain that the first employee’s “stack 1” is the same stack as the second employee’s “stack 1”. The most logical thing to do would be to go back to our chip sorter and see how it works. If the chip sorter puts all of the blue chips in the first pile, the red chips in the second pile, and the green chips in the third pile, then we might guess that our employees put together their reports using this same order. So, each employee’s stack 1 should match up to the other employee’s stack 1, and so forth. Once we can match up the reports from the two employees, we have the information we need to figure up the value of the chips in the bag: we had 340 red chips, 195 green chips and 450 blue chips.
The testing labs have to do something similar: they have to come up with criteria to match the GC-MS peaks to the GC-IRMS peaks, in order to assemble all of the data they need to determine if an athlete was doping.
Now, if you were running the casino in our analogy, you’d probably want to hire an employee who was good with numbers and was not color-blind! But the labs don’t have an option like this: there is no single machine that can be hooked up to a GC that will both identify the peaks that emerge from the GC, and that will analyze these peaks to see if they contain evidence of “exogenous” testosterone. They have to use the two machines, just like the casino had to use the two employees.
Let’s go back to our casino. Imagine that we’ve been relying on the reports of these two employees for years now to compute the value of the chips in each day’s bag. Then we find out something disturbing: there are actually TWO kinds of chip sorters in the back office of our casino, and each employee has been using a different machine to sort chips! This is very bad news. In fact, we can think of at least 4 reasons why this is very bad news:
1. This is proof that we’ve screwed up. This is no way to count chips.
2. We’ve been relying on the report of our second employee to tell us if each pile of chips contained only a single color chip. But she’s only been able to tell us about the “purity” of the stack of chips that she has seen emerge from the second sorting machine. She’s never SEEN a stack of chips emerge from the first machine, the one we’ve been using to make our chip counts! So now, we have no idea whether our counts are good.
3. Up until now, we’ve figured that each of our employees was looking at the same stacks of chips in the same order — that for each employee, stack 1 was the blue stack, stack 2 was the red stack and stack 3 was the green stack. Now we’re not so sure. We know that different sorting machines sort chips in different orders! Maybe we’ve been applying the wrong counts to the wrong stacks.
4. Up until now, we figured that the chip sorter in the back office was purchased by our casino because it worked best with the chips we use at our casino. We know that there are different kinds of chip counters out there. We figure, ONE of those machines in the back room must be the right machine. But what about the other one? Does it even have the ability to deal with the chips at our casino?
The casino owner can only hope that the two sorting machines are not all that different. If the two sorting machines work in similar ways, then maybe the stacks of chips analyzed by the two employees over the years weren’t all that different. But if the two sorting machines turn out to work in very different ways, then the casino owner is in big, big trouble: with the IRS, the State Gambling Commission, etc.
The LNDD has been using two different GCs to perform their GC-MS and GC-IRMS tests. This is like the casino that uses two different chip sorters. Based on all I’ve read, this is a screw-up, pure and simple. The only hope for the LNDD is that these two GC machines were similar enough to avoid these kinds of mistakes.
The initial information is that these two GC machines were very different.
Morgan, on your 3:05 pm post:
I’ve just posted a new analogy, where I compare LNDD’s testing to the sorting of gambling chips at a casino. For your questions, this new analogy is going to be helpful to us. LNDD used the DB-17 for their GC-IRMS testing and the HP-5ms for their GC-MS testing. This is like using two different chip sorters at our fictional casino.
Going back to my first post: remember that the GC is supposed to sort out the substances found in a complex mixture like urine. The lab does sample preparation to simplify the mix and give the GC a better opportunity to do its job. Then the lab sets the GC chromatographic conditions to best “tune” the GC to deal with the substances the lab wants to analyze. Then if all goes well, the GC will spit out each substance in the mixture, one by one, in separate and pure peaks. The GC will not itself analyze these peaks (not, at least, the way the GC is used by the LNDD) — it is the job of the MS and the IRMS to analyze the peaks.
So, when you ask if the HP-5ms “has no column identification for steroids”, I think what you mean is that this column may not have been designed for steroid testing. It may not be the right column for achieving the proper separation of the testosterone metabolites that’s needed to make pure and separate peaks, like we need to properly analyze these substances.
When you ask what column should the LNDD use to do its checking — my understanding is that they should select the best column for steroid testing (the DB17ms appears to be a good choice) and then use the same column for both their GC-MS and GC-IRMS testing. Kind of like using one chip sorter at the fictional casino from my last post.
TBV’s reference to SIM scans versus full scans is pretty technical stuff, and it’s not directly applicable to what we’re trying to discuss here. For the moment, my advice is to ignore that part of TBV’s message.
You ask if the wrong machine was used to get rid of extraneous material. Again, that’s not really an issue here. When the lab does its sample preparation, it tries to simplify the mixture in the sample, so that the sample is easier to analyze. After that point, the machines have to deal with whatever is left in the sample. The machines don’t exactly “get rid of” the stuff we don’t want to measure. They DO get rid of some stuff you might consider to be “extraneous material” (for example, I think the IRMS takes water out of the sample), and the machines can be tuned to an extent to look for certain kinds of things and not for others. But the issue we’re dealing with here, the differences in GC machines, does not have anything to do with getting rid of stuff. It has to do with how mixtures are separated and sorted. Just like the chip sorting machine.
(NOTE: Rant’s blog may have held up my post where I discussed how the GC is like a chip sorting machine at a casino. If so, then hold on, I’m sure that this post will be on line soon.)
Larry,
Sorry about the delays. Working on figuring out what causes some legitimate posts to get sent off to moderation.
Well, I’m not going let the SCAN vs. SIM point go quite so easily.
In the chip analogy, there’s a possibility that hasn’t been discussed: the presence of black chips, or those from another casino.
If the sorter is in SIM mode, it will only report red, blue and green. The SCAN mode will tell you if you have blacks, whites and chips from other places.
If your employees, for years, haven’t been looking for blacks, whites or others, then your books may be even more screwed up than you thought.
And if you have two sorters, and one puts blacks in the red pile and the whites in the green, but the other sorter does the reverse, you are going to be even more confused.
You may say that it doesn’t matter because the number of these other chips is small, but it doesn’t take too many to skew the value of the $100 chip pile.
Should you find that the next door casino’s red matches yours, but is a $1 value and yours is a $10, and you haven’t been playing close attention to that, things might be very bad for you.
TBV
Larry,
Am not ignoring your request – just got up – am processing your new analogy will respond shortly.
Larry,
I shall respond by paragraph to each of your own — your analogy is great — and I am trying to respond as clearly as I understand them.
“Let’s say that”¦” — The “sorter” is analogous to the GC machine — I get this. It is “supposed” to sort the chips by “color” — the sorter has to count the number of chips in piles blue, red and green. — one should get three piles of chips — Does the sorter tell the employee the number of chips in each pile (?) — and does the employee note this for the record (?).
“The employee does this job”¦” — The employee is unable to identify “color” — So his report is — three stacks — with three quantities — but because he cannot see color — he cannot report the “red has–340.” The “blue has-195.” The “green has-450.” What management winds up with is three stacks of 340, 195, 450 — the sack of chips had a total of 985 chips — but since the employee is color blind he cannot tell his boss exactly what percentage of the total was red, blue or green. The IRMS can sort what it is testing but it cannot identify whether it was red, blue or green. OR how much percentage of each color was contained in each stack (?)
“So we go to a second employee”¦” — numbers are not her strong point — “management reassures her that they have the numbers” – the employee then returns with the information that there are “three stacks — one is red the other is blue and the third is green” — Yes we have information from two different employees — (1) — there are three stacks. (2) – that there is indeed three different colors — but since one employee is “color blind” — his information cannot correctly identify which color goes with which number (?) Second employee is bad with numbers so all she can identify “yes indeed” there are three stacks of different color but she cannot say how many of each color is in the stack, correct. (?) The MS verifies that indeed there are three separate items — but cannot tell how much of each item is in each stack, correct (?)
“How will we match up the two reports”¦” — IF the sorter sorts the same way each time — meaning that it always sorts in the same order — then the assumption can be made that, yes , the sorter always sorts by choosing red, blue and then finally green — but this is not fact — management does not KNOW that the sorter works in this fashion, they assume that this is the way the sorter ALWAYS works, no? So if they assign the collected numbers — this assignment is then based on the supposition that the machine does indeed sort in this particular order red, blue, green.
“The testing labs”¦” — Does the MS machine ALWAYS sort in the order of red, blue, green (?) — Are we certain that the IRMS also separates in the same order (?) — It seems to me then that the “assumption” that this is so is the only thing that allows the management to “match” the numbers to the piles? They are guessing that the information always comes out in the same order — Is this so(?) is it FACT?
“Now, if you were”¦” — Okay — the two machines are required to run the test — but if the employees do not write down (mark) which numbers are red, blue or green — then assigning the values is still not a certainty — it is guessing, based on the assumption that the machines work a particular way and each machine when working always works in the same order — correct (?)
“Let’s go back to our casino”¦” — So management discovers that the employees may have been using two different sorting machines — THIS REALLY COMPLICATES THINGS — Lets say they make their assumptions that the sorter works in a particular order and go with this — but now — we discover that there are two sorters — are the sorters the same? Or are they different? If one employee was using one sorter and the other the second — and if – (1) each machine is the same — then it may be assumed that both will work the same way — red, blue, green — everytime — then this may be okay — (2) — if the machines are different and do not work in the same order then we are sunk. Not only ARE WE GUESSING that machine 1 (employee 1) always performs the same way and that machine 2 (employee 2) used the same machine — we now wind up knowing that we have no way of assigning whatever results we have — because we have found out that the employees may have been using two different sorters.
“1. This is proof”¦” — The management based their counts on a supposition — then discovers that their supposition is wrong — because they made it on “assumptions” not on hard knowledge of what sorters were in their basement.
“2. We’ve been relying”¦” — The report of the second employee was to verify that indeed the stacks were red, blue or green — that only reds were in the red stack, blues in the blue stack and greens in the green stack. BUT HER INFORMATION does not tell if Employee 1 ran the tests and the reds, blues, greens always come out in this order — this was assumed that they do — I am not certain — do you mean that employee 2 was also asked to assign the numbers to the different stacks or did management do this under their assumption that the order is the way they assumed(?) — if the employee was assigned the task of “matching” numbers from employee 1’s test — then she made an assumption — but if management told her “not to worry” this is the way the machine works just apply the numbers and give us your report” then this is a completely different matter, no(?)
“3. Up until now”¦” — Yes — management made “assumption” based on guessing — just about now — the Pepto-Bismol is making its appearance and someone is desperate to not be found out. If he’s smart enough — he knows that there is no way to “go back” and tell what was what, pink slips flying around his thoughts.
“4. Up until now”¦” — This last question — has come too late — it may be a good idea to discover if one sorter is the best and the other is only good or not working right or is the wrong type of sorter for the job — it is irrelevant — it only points out more clearly that management has a history of sloppy work based on assumptions.
“The casino owner”¦” — The second bottle of Pepto-Bismol is half drunk already. Maybe the bottle of Scotch in the bottom drawer is now empty — the man is living a nightmare and he knows he’s not going to be waking up.
“The LNDD has been”¦” — Management is doing it again! Basing their choices on assumptions and not fact — that nightmare where he’s falling — has his heart beating over 230, the sheets are wet and his body is tacking on a rictus of muscular spasm that reminds one of a German statue of a suffering saint. — the guy has gotto wake up soon, if he doesn’t he’s a goner.
“The initial information is that these two GC machines were very different.” — The guy just smashed into the pavement — he wasn’t dreaming.
Respectfully yours,
Guinea Pig
Larry – your post are wonderful. I’ve been struggling with the whole MS/IRMS thing. Maybe I missed it months ago on TBV but now I understand. It almost seems such an extraordinary stuff-up that there must be a simple and logical explanation that people have missed. I wonder what LNND are up to this weekend – crisis meetings or satisfied smirks about how Landis’s team are idiots?
Larry I had answered your casino post — but I am afraid it may be lost so I shall try again — my posting time for this was 2:17 I believe but it seems to have not been picked up, even though it showed on my screen that it was in place — sorry. So here we go — I shall answer you paragraph by paragraph.
“Let’s say that we own a casino”¦” — Every night one employee “sorts the chips” — The sorting is to separate red, blue, green and tell management what the total amount of chips we have in hand – the sorter is like the LNDD GC machine. The machine tells us the numbers in any given “sorting” – % – of Steroids, EPO, Etc. – Got’cha.
“The employee does this job”¦” — The employee — “machine” is color blind — so it can tell us how many total chips were in the sort — but he does not know if it was X-% red, x-%, x-% green — the LNDD GC machine does not identify that “Steroid-red was 340, that EPO-blue was 195 and that ETC-green was 450,” — but the employee machine does tell you the total amount of chips that were sorted but not the specific quantity of each or whether a red got mixed in with the blue or the green? So the results/report may be likened to how the LNDD IRMS machine produces a “sorting count” but it can’t tell management what count was Steroid-red, EPO-blue or ETC-green was present in the whole stack, or that some S-red chip may just have slipped into the E-blue pile and got mixed in with the ETC-green pile. Management has a problem, right?
“So we go to a second employee”¦” — management then sends a second employee/machine to verify that each stack contains only the right color chip — in each stack — but because she’s not good with numbers — she can only verify that indeed there were S-red, E-blue and ETC-green — the LNDD MS machine does this kind of sorting — but it does not tell you the percentage or quantity of what it’s testing. But we have verification that each stack is pure S-red, E-blue or ETC-green. Now — the management must match the numbers from employee-1 with the verification of purity from employee-2, sounds like this should not be difficult.
“How will we match up”¦” — The problem is management doesn’t know if employee 1 and employee 2 – when running their sort/test in the same order — first S-red then E-blue then ETC-green — maybe employee 2 likes blue better and starts with blue — since employee 1 is color blind — he can’t tell the difference — The GC machine can tell you the number / % – of each sort but it can’t tell you the name/color it is sorting while the IRMS machine can check for “purity” but it doesn’t tell you what % or count was in each stack, right?
“The testing labs have to”¦” — So — the problem for management is having a way to match the numbers of the GC machine with the verification of purity of the IRMS — but management must have sure knowledge that the order of the tests were performed in the same manner by both employee machines — otherwise they may assign numbers and percents from S-red to E-green or ETC-green and never know it — This is where the “lab protocol” comes in, right It is by punctilious following of the lab protocol that management ensures that the order of sorting/testing is the same, right? Does it mean anything that my ears are hot and I see smoke collecting in my office?
“Now, if you were running the”¦” — No problem here — the casino must hire the color blind and the dyslexic employees — equal opportunity at its worst. — The lab has no choice — there is not one machine that can do the work of both. so the solution the management comes up with is that both Workers have a very specific protocol to follow — this is not optional or dependent on if you were partying on the night before the test. The test/sort won’t be able to be matched unless the protocol is followed to the letter.
“Let’s go back to our”¦” — I am reaching for my friend “Pepto-Bismol” — this is not news I want to here — WHO WAS THE GENIUS WHO LEFT THAT OTHER SORTER/MACHINE IN THE LAB! — Okay — I’m taking deep breaths, thinking warm fuzzy thoughts — my heart rate is cruising at a manageable 183 — but I can deal with this! Are you telling me that there are TWO sorters in the lab? That my dependable employees have each been using their favored machine/sorter? Is this what you’re telling me? Alice — make an appointment with my physician — I’ll be there at 3 pm.
“1. This is proof that we’ve”¦” — It is at this point that the bottle of “deluxe Scotch, ½ gallon economy size from CASCO comes out of my bottom drawer. The question whams into my brain: “How in tarnation am I supposed to match these numbers!”
“2. We’ve been relying on the”¦” — “Alice! Make out a pink slip! What in heck am I gonna do? Can I fire them? So you’re telling me that MY employee has been feeding me verification but I can’t be certain that her verification matches the colorblind dodo’s numbers? NO I WILL NOT CALM DOWN! Alice! Get me a bottle of Pepto-Bismol!”
“3. Up until now, we’ve”¦” — “How could this happen? How could we be having two sorting machines that may be sorting in different order from another? You realize that we have no way of matching our test/sorts? — No I don’t need a glass — I’m like drinking my scotch straight from the bottle!”
“4. Up until now, we”¦” — Two sorters — one may not even be the right kind of sorter — the heart rate has now reached a flying 200. “I will not calm down! Not before I get answers! Alice! Where is that Pepto-Bismol!”
“The casino owner can”¦” — “Has the air conditioning stopped working? It’s very warm in here Alice — thank you for the Pepto-Bismol — I won’t be needing it, no — I’ve stopped having feeling below my neck Alice! Maybe, nobody will notice! Do you think anybody will notice Alice? Stop looking at the damned floor and answer me!”
“The LNDD has been”¦” — “I’m having a nightmare — that’s it! In a little while I’ll be waking up and everything is going to be all right! Alice — what are you doing in my nightmare?”
“The initial information is”¦” — MEMO: To all employees — Our beloved and very hard working manger is taking a leave of absence. As you may know, Mr Joe has been under a lot of stress, his workload was overwhelming. The good news is that he will be coming back to us as soon as the doctors feel that he can safely be allowed out.
The Management.
Larry,
Your analogy is awesome and great. Thanks.
—
I am a little suprise that a such mistake, using a bad column, would not give stranger results.
With all specialistswho have worked on the case it’s difficult that they didn’t detect it, at least they must have pointed out incoherences on values. It seems for me that they didn’t!
What about the blank testing and so?
Morgan, I will get to you in a later post.
TBV, thinking about the SIM versus full scan, are you saying that LNDD mixed up these two techniques? Oh boy, I have to think about that!
RDK, our little discussion here and the slightly bigger discussion over at TBV seem to have escaped the notice of the rest of the world. Not even a breath of this over at Daily Peloton Forum that I can find.
Jean C, the remainder of this message is addressed to you.
First, you’re welcome, and thanks for the nice words.
I’m also surprised that the results do not seem stranger to us. But perhaps this can be explained by our psychology, and by the ability of human beings to notice patterns, even when the patterns do not exist or are not significant. We were told back in May that the GC-MS graphs and GC-IRMS graphs were produced with the same GC, and we assumed that the LNDD followed a logical and predictable process. Then our psychology went to work, finding patterns and matching them up.
Think about meeting a mother and her new baby for the first time. You look at the baby, and tell the mother, “the baby looks just like you!” Well, in truth, the baby does NOT look just like the mother — the mother is taller, for one thing. But you can see something in the shape of the baby’s face, or in the baby’s eyes … plus you know that the woman is the baby’s mother, and you expect children to look like their parents. So it does not seem crazy to you to say that a baby looks like an adult. Now, imagine that you learn that the baby is adopted. If you had known that the baby was adopted in the first place, it never would have occurred to you to think that the baby looked like the mother. But because of our human nature, even after you learn that the baby was adopted, you may not give up the idea right away that the baby and the mother look alike.
In the FL case, we “knew” that the GC-MS and the GC-IRMS were produced by the same GC. All the scientists and specialists “knew” this, too. And with this “knowledge”, we all went to work finding patterns and similarities. The fact that we were able to find these patterns and similarities is not “proof” that the GCs were the same — it is the product of our false assumption that the two GCs were the same, and the human capacity to find patterns and similarities between any two things.
As far as whether the specialists pointed out that the GC-MS and GC-IRMS tests did not match up all that well — some of the specialists did point this out, and some did not. Unfortunately for the cause of truth and justice, the specialists who pointed this out were paid by FL, and the ones who failed to point this out worked for USADA and LNDD. Again, I think that psychology goes a long way to explain this split … unfortunately, so does economics. Please don’t get me wrong, I’m not saying that anyone was “paid off”, I think that the situation is more subtle than that. If you approach an expert as a potential expert witness for your side of a case, the expert is going to look at the case from your point of view … and the money the expert will receive is additional incentive to see the case from your perspective.
The blank testing is designed to see if a particular machine (GC-MS or GC-IRMS) is working the way it’s worked in the past. The lab has a stockpile of a particular urine that’s supposed to look like urine from a typical athlete. The labs do test runs on the urine from this stockpile, and compare the results of these tests to past results on the same urine. These tests tell the labs whether the machines are working consistently, but they can’t tell if the machines are working consistently wrong.
Now Jean C, a question for you. For the most part, you and I have been on different sides of the debate here. For the most part, I have been on the “fairness to the athlete” side of the debate. For the most part, you have been on the “crack down on dopers” side of the debate. Isn’t it important to BOTH sides that we have labs operating competently and professionally?
I don’t find it so surprising that this was possibly missed before now. As the one post at TBV is says “Hiding in Plain Sight”. The Landis team could have been so focused on the RT issues, and perhaps the CoC issues that they could have overlooked this. I am assuming that these documents were part of those made public by Landis way back when as the basis of this “Wiki” defense strategy, so many people were looking at this stuff and missing it. If using the same column is SOP, not just in LNDD, but as a general principle in all labs, most people would have probably assumed the SOP was followed.
One thought occurs to me one the question of why LNDD used 2 different columns. On at least 2 occasions in the past year and a half we have seen how the general French custom of closing down during August for holiday has been a factor in TdF testing. I wonder to what extent this has more impact on things at LNDD. The TdF is a big event, at least 80 some odd tests have to be done, and the race ends a week before August. ASO and UCI wants test results, leaving not much time to do the tests from the later stages before the lab shuts down. People rushing can cut corners. It seems to me that this is an issue that WADA needs to take up with LNDD (or whatever it is called now). Or perhaps the TdF tests need to go to another lab that remains open during August so that the proper time and attention can be devoted to the testing.
Morgan, I was preparing a long post to you, with SMALL corrections to your understanding of the casino analogy. Then I got to the portion of your memo with the pepto bismol. And the scotch. Now, I don’t think the French are big scotch drinkers. (If I had easy daily access to bottles of Cote de Rhone, I wouldn’t drink scotch either.) But the progression you describe through different “medications” is not only funny, but clearly indicates that you get what I was trying to say!
Ludwig, I’m hoping that you join this discussion, too. I think if I tried hard enough, I can show what LNDD is doing here is likely to produce as many false negative tests as false positive tests. Maybe more false negative tests.
The science that LNDD is using here is terrible. Bad science does not serve anyone’s interests. If it’s true that 80% of the peloton is doping and only 4 guys got caught, this is about the best single explanation I can find for those statistics.
If you want to fight doping, the LNDD is not your friend.
I think we can get together on this.
Larry,
I’m not saying LNDD mixed up SIM and SCAN, except maybe why you used one or the other when you look at the results. I’m saying they only partly got the idea of looking for things that might be wrong, and essentially assumed the sort was going to come out red, green and blue with nothing else present.
People of a certain mindset rarely find what they aren’t looking for, and do find what they are looking for… even if it isn’t there.
In a Russian film called Repentence, a local party boss delivers a speech:
“It’s hard to find
a black cat
in a dark room.
Especially
when the cat’s not there.
But we will!”
They all dope, there’s nothing else in the chromatograms, there is nothing wrong with our technique, and the 5aA CIR is -6 delta units. Here is your cat.
TBV
Well heck, Larry, it’s your own damned fault…My hat off to you – amazing how you seem to be able to produce analogies that allow even a bone head like me to understand technical scientific language. It is really a wonderful ability that you have.
As I was answering you on the second submission – I was a little concerned that you may think I was taking your request lightly – I realized you were serious in asking for my feedback and I was concerned that a bit of levity could be interpreted as me treating it lightly. Am really glad you got it and I am happy that you enjoyed it. I figured answering you in that little last scene would communicate to you better and with some pleasure that you were successful with me. You are an amazingly good teacher Larry – I hope you do teach in your profession. You teach me something about law, science and legal reasoning almost every time I read your comments.
Now – just to be clear on the use of the “Scotch” and Pepto-Bismol – (:-) – that was intentional – if I was communicating with a French audience – I may have used perhaps a Napoleon Brandy (;-)).
tbv – you could have not used a better piece of text – truly. I used to hold a master class for actors here – and one salient rule I rammed down their throats always – “Show me – don’t bother explaining!” Thanks for whipping that black cat out of the bag. May I use it at a later date?
William – a pertinent proposition – and a possible scenario. When one strings together certain words like “scientific,” “SOP,” “Laboratory” – there is an automatic “assumption” that “proper scientific” and “correct legal” procedures are being in use. But this is a great blind spot when applied to situations such as we are facing in cycling today. The fact that it was “missed” till now perhaps is a very good thing.
As tbv says -“They all dope, there’s nothing else in the chromatograms, there is nothing wrong with our technique, and the 5aA CIR is -6 delta units. Here is your cat.” We all needed the time to get deeper into this subject – and you would agree I am certain that our depth of knowledge has greatly benefited from the “miss.”
As to those who think it is logical to immediately turn to find the ones at fault – we will get to that too.
I have a feeling that the methods learned from the experiential knowledge produced by the people in Floyd’s Crew, TbV and the Rant Line and all the others – too many to mention – will not be wasted. The only important thing as I see it, for the moment is not forgetting that this is really all the fault of Mr. Landis and the shameful application of allowing the “public” to be a party to his defense. Floyd and Crew – you have done cycling a great service – not only that – you created the possibility to involve people from all over the world. Thank you.
Okay – one last thing folks – this ain’t over yet. The opposition is not going to lie down and just turn over. Too many have everything to lose, or at least they firmly believe this to be so. The situation is more dangerous and unpredictable now then ever. We need to stay sharp and always watchful. But – as I have been accused of before – I am preaching to the converted. Let me apologize if I waste time stating the obvious.
William, you’ve raised the issue whether LNDD may be rushing some of its tests to get their work done before their August vacation. This appears to be the case; in fact, the director of the LNDD has admitted as much. http://sports.espn.go.com/espn/wire?section=cycling&id=3029499.
But the error with the use of the wrong GC column? It appears that this was not a one-time mistake. It appears that the LNDD used this same (mismatched) combination of GC columns on both the “A” and “B” samples for FL. It may well be the case that LNDD repeated this same error over many tests, for who knows how long.
There’s been an oft-reported statistic that LNDD has a much higher rate of finding doping positives for testosterone than other WADA labs. Up until now, I’ve dismissed this statistic, as the population of athletes tested by LNDD may be different than the population tested at other labs. But now, I am starting to wonder …
Morgan:
I have been thinking for a while that WADA and LNDD are trying to baffle us with “science”. We have some pretty complex tests run on some pretty complex equipment. It seems a first glance to be quite definite. How can they be wrong when their are using all this scientific equipment. But this is not just a case of inputting a sample into a “Black Box”, pushing a button and having the “Black Box” spit out the final answer. But it is not as simple as that, there are many steps and many places that a tech can mess up, from ignorance of proper procedure, carelessness, etc.
It’s like using an electronic calculator to do your math. Many people assume if you use a calculator, you must have gotten the right answer. As a former math teacher, I know from experience this is not the case. You can fail to proper clear a previous problem, you can miss-key an entry, you can put things in the wrong order. You might even fail to realize that different makes and models of calculators operate in different ways. Maybe like failing to realize that using different columns might give different results.
William,
Likewise the powers that be used the same “complexity” to bull their way through – that is inexcusable. You will please excuse me – my poor brain is just now recovering from wrapping itself around red chips, blue chips, and the green ‘uns….I think I may just have pulled something.
Your calculator analogy is perfect William. “Habit” is the bigest problem to conscienscious thinking. Surprisingly – it is also the main cause of boredom. You find me in absolute agreement with you.
Wonder what the darkside of the force is getting ready? Just kidding….about the darkside, I mean.
Rant,
This rather fascinating – not a peep – no “buzz” – no reactions.
Considering that the world has become rather small – and blogs such as TbV and the Rant Line are “public” – the “silence” is deafening. Now it could be that I simply have not found the right sources to read…
Does anyone have any ideas as to what is going on?
Must be time to put out another post. 😉
Just a moment…
Morgan, I am keeping an eye on Daily Peloton Forums, and so far, there’s no reaction there to the column screw-up at LNDD.
“Hello, ESPN? May I speak to Bonnie Ford please? Yes, I can hold …”
“update” – MEMO: “news on progress”
It would appear that our Mr. Joe is doing quiet well, he has stopped attacking anyone in sight. The new combination of meds with Thorazine leading the pack, seems to have brought the desired result. We still feel very confident that our Mr. Joe will return to the fold.
On a completely different subject – MANAGEMENT wishes to welcome Ms: Alice to the management level and convey congratulations for her elevation in status. Enclosed within the sealed envelope Ms. Alice – is your own personal EXEC- washroom key.
Keep up the good work – we think you will be going places.
The Management
Larry,
Your question don’t need an answer, of course everyone prefer a better system, but improvements need work and money! Stupidity or dishonesty is to want to cut the funding of WADA/USADA.
—
the “silence” is deafening because a lot of people have realised that their athletes were cheating them, Landis, OP, Tmob, Astana, Rasmussen,… ‘s cases have opened the mind of people who were just listening the doped athletes’ tales since a while. They have understood that it’s easy to beat the testing, that a lot of athletes doped, and it’s impossible to beat blood doped riders on a GT!
Now incredible performances like Landis’stage 17 after his blocatos would be seen with suspicion.
There isn’t anything over at dpf, because none of ya’ll have posted about it, and the people who read TBV and rant would normally be the ones to alert the forum. I was really surprised Ali didn’t post on the column issue actually, but didn’t think it was my place to start the thread, if s/he wasn’t posting about it.
AJ
The “silence” is still deafening — nothing has been really said.
In my opinion “what was said” only shows how afraid some people are that THE ISSUES are NOT ONLY ABOUT DOPING . Terror that people will actually use their heads and the “only issue that matters to some” – will go puff — because it is discovered that the powers that be — IGNORE — “rights of individuals” — “hard science is not as good as “cargo-science” — if it is in the name of getting dopers — “we can just dispense with legality” — and if that doesn’t work — “we’ll just make up our own rules,” After all — “no priced is too high to pay to get them dopers.”
Perhaps if we were all existing in a Chinese restaurant — then “picking” a rule from Column A — Going with a “scientific test” from – Column B — Mr. Ho — I think I’ll take that “Trial” from Column C”¦
Good thing we are not all living in a Chinese restaurant — and the “end does not justify the means!” or act like ostriches and stick our brains in the sand.