Friday, April 25, 2008

Floyd Landis, Hypotheticals, and Cheating

In my class Wrong Science, Bad Science, and Pseudoscience, we've been looking at science in society and last week the question turned to sports-based applications and cheating. One can cheat with technology if one gains an unfair advantage. The question then becomes determining what makes an advantage unfair.

There are plenty of fair advantages -- those that come from unequal skill or innate ability, superior training and preparation, more effective execution, to name a few. But what is it that makes an advantage unfair?

Two options that seem obvious, but have flaws are asymmetrical advantage and advantages that break the rules. First, if one competitor has access to something the other doesn't that is not fair and therefore cheating. This does not seem to really be a necessary or sufficient condition because there are fair asymmetric advantages (me and Michael Jordan playing one-on-one) on the one hand, and on the other, it makes sense to say in certain circumstances, "They were both cheating." If both runners leave before the starting gun goes off, it is a double false start not a fair start. But if asymmetry was sufficient for cheating, that would make no sense.

Second, while it may certainly be cheating to break the rules, it doesn't seem necessary. That is, we can think up new ways to cheat that haven't been encoded in the rules of the game. Indeed, there must be a non-rule-based notion of fairness to guide the way we make the rules. Rules are not capricious, we don't just make them up willy-nilly. Rather, we understand that some act violates a pre-existing sense of fairness and then make a reasonable rule to explicitly rule out the act. The unfairness doesn't come from breaking the rule, but the rule comes from recognizing and wanting to call out the unfairness.

We were playing with the case of Floyd Landis, the cyclist who had a sample come back testing positive for steroids during the Tour de France. His previous samples had come back clean, but a sample from late in the course of the race tested positive. His argument in his defense was that he is rational and understands how steroids work. Steroids give an advantage if used while training in order to build muscle mass. They convey no advantage in the middle of the race when one is looking for endurance. Steroid use by cyclists is against the rules and at the time he supposedly took them would convey no advantage, therefore a rational person would not take steroids in the middle of the Tour de France.

I don't know whether he is being truthful or not, but let's work with the hypothetical that he did take the steroids late in the race. Given that they convey no real advantage, would it have been cheating? It did break the rules, but gave no asymmetric advantage. Would it be different if he wrongly believed that he was going to gain an advantage? Suppose someone gave him a placebo which he willingly took thinking it to be steroids which he wrongly believed would give him an advantage?