Sunday, July 20, 2008

Fallout 21

Questions of the Day

Various sources pass on the El Pais report that Saunier-Duval DS Joxean Matxinas said Ricco confessed EPO use to him. This is unconfirmed. He is also quoted as saying Piepoli told him he'd "done the same thing". Matxinas also said Ricco's control went a tour of its own to Lausanne, LNND, and Barcelona.

Now Ricco is denying having done EPO and says he will defend himself against the charge.

We do not yet have a reported test result for Piepoli, whose controls should be blinded amongst the perhaps 200+ samples that have been taken for the Tour.

  1. Is there pressure to find a positive for Piepoli?
  2. What would failure to find a Piepoli AAF say?
  3. Might there be some bending of sample blinding?

It looks like Landis didn't go to Breckenridge for the NUE event yesterday. What's up? Hard to blame him for being unmotivated.

Oscar Peirero crashed badly on a descent in the Tour and is seriously injured. AP story. You can't say he wasn't trying.

Reuters says Ricco is going to dispute his test results:
He has denied any wrongdoing.

"It's necessary to wait for the counter-analysis, then see if the method they used to do the test is valid," the 24-year-old was quoted as saying in Sunday's La Gazzetta dello Sport. "I don't think it's 100 percent certain."

Good luck with that.

Guardian/Fotheringham thinks the "Drug-Busters" are winning, finally.

IHT runs an AP story under the headline, "A lot of dopes racing in the Tour de Farce", filled with equally original insight.

The LA Times runs a Sunday feature on the reliability of DNA tests, which aren't looking as good as some people assume. It also shows courts having difficulty evaluating scientific evidence.

Steroid Nation points out reports that Spanish Doctor Jesus Losa is emerging as a potential EPO source. He'd been fingered by Millar way back, with no action taken. SN also thinks Ricco is channeling the Landis defense.

We went for a ride, and caught the race later on the Tivo. It was a barn burner, beating last week's 218 miles, 1007 TSS points and 7924 kj with 237 miles, 1065 TSS and 8617 kj. This left us wiped, and in a stupor to watch the bang-up stage. Too bad about Peirero, and Menchov's fall on a move. We'll all enjoy the rest day tomorrow, and look forward to Tuesday and Wednesday.


wschart said...

Part of the CAS decision was that Landis did in fact compete in a US Cycling sanctioned event when he rode at Leadville, resulting in his ban being extended beyond what it might otherwise have been. So perhaps he is not taking any chances and skipped the B100.

Larry said...
This comment has been removed by the author.
Larry said...

A MUST read article for those of you interested in the dependency of the law on scientific tests, as well as the ability and willingness of courts to understand and challenge these tests:

LA Times: How Reliable Is N=DNA?.

This is yet another indication that we're way, WAY to0 willing to trust the guys in the lab coats. Also, despite our wish that Landis had received a hearing in a courtroom, it does not appear that courts are much better than arbitrators when it comes to evaluating scientific testimony.

Davis Straub said...

The comments on the LA Time article point out what a confusing mess the reporter has made of the issues by not having a statistician near at hand to help out with the interpretation. I think it says nothing about Floyd's case. It's just a complete mess.

dailbob said...

Hey Larry,
I think the guys in the white lab coats you refer to are not the guys doing the actual DNA analysis, but the Statisticians who made assumptions about the probability of loci matches. I agree with Mr. Straub that the reporter would have benefitted from having a statistician help him understand where the FBI is coming from. After reading the article, I'm hoping the FBI has validated their probabilities against their database.

On a different subject, I went on vacation for two weeks the day before the CAS decision, so I wasn't around to participate in the discussions. After reading through most of the posts with the associated comments, I wanted to let you know I enjoyed reading your (as well as TBV's, Thomas' and M's) discussion around the emailer/IRMS peak identification. By far, the most interesting to me.


Larry said...

d-Bob -

I thought that the LA times did a very good job with the statistical analysis. I thought they nailed the statistical concepts by comparing the odds of going to a party where someone else has your birthday, and going to a party where there are two people who have the same birthday. The people who are questioning the DNA matches are doing the second kind of comparison, and of course it's more likely that there are going to be matches. There's also the matter of looking at There's also the question of matches between related people, etc., etc. It's not clear from the article that there is any fatal flaw with DNA matching, and I don't think the article is trying to say that there IS a fatal flaw.

The article is, first and foremost, reporting a news story. DNA analysis has become important in crime lab work. Some people are questioning it. Some courts are cooperating with the questioning. That's news.

The article cited some experts who think that the analysis might mean something.

But I think there are two bigger points, which is why I cited the article here. The first point has to do with the public's uncritical acceptance of scientific testimony. I can speak for myself, if the police say that they have a DNA match and there's a 100 billion to 1 shot that the match is just a coincidence, well! That's been good enough for me. I certainly never focused on the fact that they are doing the DNA match on a sliver of the DNA information, and that even that sliver is not 100% available in every case where we given those impossible odds.

The second point has to do with the knee-jerk reaction of the labs, and the police-FBI guys who control the labs, to keep us as far away from an understanding of what they do as they possibly can. Why should the FBI block people from analyzing the DNA data banks for matches? Why should the feds try to intimidate the states from cooperating with these analyses? If it can all be explained by a qualified statistician, then what is there to fear? Why should the FBI threaten to expel states from the CODIS system for cooperating with these investigations?

The FBI even argued that doing these investigations could result in the data being corrupted. How could that happen? Strangely, it's the same excuse that the LNDD gave the Landis team, when the Landis team asked for access to a copy of the Landis test data.

I'm not saying that the science is faulty. I'm saying that we should try to understand the science, and that the people in charge should try to explain it.

m said...

Agree with Straub the stats explanation was largely incoherent except to point out the general areas of contention.

A priori 99.999999% of reporters are statistically illiterate. This guy just regurgitated what he was told without understanding it, and therefore only gave an impressionistic explanation, that may or may not be numerically accurate.

Larry's larger point about most folks uncritically believing scientific experts is probably true, but only if it doesn't run up against their prior beliefs. Think global warming or the OJ trial.