I've been wondering about this a bit. What kinds of things could make a test like that come up with results which we interpret incorrectly? Set aside the obvious nonstarters, that either he cheated or that the test was flawed or tampered with. They're all more probable than anything else, but make for boring discussion.

One thought I had goes back to some articles I read a few years back about the role of testosterone in decision reinforcement. If I recall correctly, when you make a good decision -- or are convinced it's a good decision -- this hormone tends to increase, while when you're not comfortable with your decisions it tends to decrease. The theory was that this leads to a cycle where those making the best decisions (at least according to themselves and their peer reinforced opinions) end up with higher confidence and leadership that comes from the increase in the hormone, while those making poor decisions or those not supported by their peers end up with lower levels of the hormone and tend to become more docile and less likely to lead (over time and in general of course).

Apply this to Landis for a minute. He comes off the worst day of the race he's had and ends up with a huge rebound. In the one day he makes up for previous errors and really pulls ahead. Could this be not a result of increased testosterone but rather could the increased levels be the result of this exaltation following a perfect event?

Surely some will say that the great performance is more evidence that he'd been doping. Is that really possible? As I recall, no previous or following tests showed unusual levels of these steroids. Do they work that quickly? I had thought the benefits from drugs like this were cumulative and increased performance over time by increasing the result of exercise and work. If that last bit is the case, any correlation between a one time doping incident and such a performance would have to be coincidental.

I'm not a doctor, obviously. These simple conflicting bits of information however do raise more questions about the testing process and the use of these drugs than the news reports seem able to answer at present.