The Floyd Landis Case: Distributed Knowledge, Transparency, and the Wisdom of Crowds

Lynne Kiesling

As KP readers know, I am a cyclist and commented on the innumerate media debacle surrounding the Floyd Landis drug test here and here in July. Between the histrionics of the media, the history of bad behavior on the part of at least one testing lab, and the sheer ignorance of science and statistics, the whole episode leaves a bad taste in my mouth, regardless of Landis’ actions. Nonetheless, I have been keeping up over the past couple of months as a lurker at Trust But Verify (TBV for short) and the Daily Peleton Doping Discussion Forum (DP for short).

On the TBV, there have recently been some extremely knowledgeable and telling forwarding of documents from “Ferret” (all of the support for my brief narrative here is catalogued at TBV). He knew enough, and said stuff in ways that indicated that he might be one F. Landis of Murietta, California. Then over last weekend transparency kicked in, and a new user on the DP Forum called Floyd said that he was indeed Floyd Landis. Independent users have verified the identity of Ferret/Floyd. [Note: edited to correct errors noted by TBV, thanks!]

Furthermore the TBV blogger had done such an outstanding job of evaluating evidence in a balanced way, keeping the information flow large and clear, and stimulating constructive conversation on Floyd’s case that Ferret/Floyd has sent case documents to TBV to archive and make available to all. In addition to the site where Mr. TBV has put the documents in the case, the case files are available at

Stephen Dubner at Freakonomics also noted the 21st-century independent media aspects of this decision.

But the real story here isn’t just in Floyd posting the files on his website. The real story is the conversation that’s been going on in the DP forum and at TBV for the past several weeks, and how Floyd is going to use this community to implement a “wisdom of crowds” approach to their defense files. Floyd’s defense is an open-source project, and is likely to be much better because of the cumulative knowledge of the cycling community. And it’s not just cycling knowledge; it’s chromatography, forensic chemistry, law, statistics, etc.

Floyd and “one-mint-julich” put it succinctly in a couple of posts:

[one-mint-julich] “This is definitely one of the waves of the future, actually of the present. It is getting harder and harder for any one person to know anything that a large number of other people, combined, don’t know better.”

[Floyd] “You called it. Will and I have been reading these from the begining and while you guys are opinionated (as am I) we decided that, if given the info, you would be constructive. We’ll call it the Wikipedia defense, and I would have given everything to you sooner but had a very hard time selling the idea to the lawyers.”

The open source, wisdom of crowds, Wikipedia defense. Bravo. The DP forum and TBV are not necessarily Landis supporters; you have a spectrum of folks in the range of being convinced (or not) by the argument. The wisdom of crowds works best when there’s a diversity of opinion and expertise, buttressed by the ready availability of data that is pervasive in the Internet and in these documents. This process wouldn’t necessarily get us closer to the truth if it were at a fan site; you need the potential for true analysis and criticism. You also need a way to diffuse the fan and the dismissive harsh critic. This distributed intelligence is the key to what makes markets work, and hopefully will make this process converge on truth.

4 thoughts on “The Floyd Landis Case: Distributed Knowledge, Transparency, and the Wisdom of Crowds

  1. Thanks for the flattery, for which you get a link back.

    You have a minor error — I don’t believe Ferret ever appeared at DPF by name. TBV invented the identity out of thin air wanting something other than “anonymous” for a snappy attribution. In retrospect it may have been a mistake, since I doubt Ferret Floyd appreciates the handle.


  2. Ooooh, good catch! I wrote DP when I meant to refer to TBV. This is what happens when I post right before dinner … Post corrected.

  3. Ooooh, good catch! I wrote DP when I meant to refer to TBV. This is what happens when I post right before dinner … Post corrected.

  4. Landis could well have been framed, and not by sample tampering. Testosterone is easily and rapidly absorbed through the skin, especially if combined with DMSO. If Floyd got a hand shake, a pat on the arm, any casual contact with someone who had testosterone cream on their hand (or had wiped it on a handlebar, etc), voila! Testosterone in Floyd’s urine sample. Unfortunately, short of the culprit confessing, proving this theory is next to impossible. But the plausibility is at least as great as Landis intentionally useing testosterone, which would give him no immediate advantage and be readily detected in the urinalysis he knew was coming…

Comments are closed.