Can we easily find the “bad” teams … ?
Again, in 2008 the most interesting thing for a statistician following the Tour de France is to look for means to detect possible doping cases. Right now we have the total results for the first 11 stages for all teams, and a simple boxplot of total time after stage 11 against team – ranked by the median time – looks like this:
Not that this is a final evidence, but the gap between the 5th and the 6th ranked team makes me a bit suspicious … Let’s see how things will further develop.
If anyone wants to come up with a sophisticated model, I am happy to share the data, we have collected so far.
