Fellow blogger (and, full disclosure, good friend) Cosmo kind of stirred things up a bit with post regarding attrition rates in this year’s Tour. My cycling blogging has been mostly just for fun, as I’m not much of a cycling expert, so I mostly do it to entertain those folks who enjoy mixing sports, numbers and graphs. But statistical commentary on sports draws me like a moth to a flame, so I basically have to weigh in.
At the moment, his commenters are kind of laying into him, and there are some legit criticisms there. But as with anything else on the internet, people are getting considerably more worked up over this than seems reasonable. Would data on attrition due to crashes specifically be more germane? Yep! One of his commenters went out and tried to track down that information on crashes and found that the attrition rate due to crashes does seem considerably higher this year. However, it’s important to note that information on the reason for withdrawals can’t be found in every case, so even that analysis is somewhat incomplete.
That’s not really a criticism, just an observation you can make about any attempt to answer a question using data.
- <Tour de France Bump Plot: Stage 12
- <Tour de France Bumps Chart: Stage 14
- <Tour de France Bump Chart: Stage 16