Statistical Skier

Podcast Shenanigans

So I tune into the FasterSkier podcast tonight during my run, only to hear that I’m apparently the lead topic, by way of Kikkan Randall’s strong classic sprint (and distance) results in Kuusamo. Fair enough. As fun as it would be to turn this into a full on blog war, I actually think we’re all on the same page here.

Colin and Christopher had expressed a fair degree of confidence in their first podcast this season about Randall’s chances of winning the overall sprint WC. I disagreed slightly. I guess I’d just point out that they made those claims before Randall’s good classic results, so my point that she’d have to vastly improve her classic results seems pretty spot on to me. Also, one weekend of strong classic races does not a trend make. (But I’m certainly hopeful!)

And I’m still not sure how Randall can win the sprint WC overall without beating Bjørgen at least once in a classic sprint. But if Bjørgen really does skip two freestyle sprints in December, that will help enormously.

Vibeke Skofterud

This is just going to be a Nordic Commentary Project heavy post. Christopher Tassava lays out a fairly good case arguing that Vibeke Skofterud tends to ski fast early in the season, but fades considerably as the season wears on. I’m certainly glad that my new graph making tools proved useful!

He asked if I’d weigh in on the issue, and I’m glad that he did. You see, statisticians spend a lot of time being Debbie Downers. People come to us with interesting questions and we spend a lot of time, a lot, disappointing them. You don’t have enough data. There’s no clear trend. There’s not much difference. Nothing happened.

So you can imagine how delighted I was when I sat down to look at Skofterud’s distance results condensed into a single “season” so we can see trends by month:

She’s had a handful of strong races in early to mid March, but this is a pretty clear picture, I think. The blue line is the median by month. The situation is only marginally less obvious with sprinting:

That wacky hiccup at the very beginning is because I’m aggregating on month, and there were only those ~3 races in October back in the day. Ignore those for the moment, and you can quite clearly see a distinct shift beginning in January.

I’m not going to speculate on the reasons, as I’m sure I have no idea. I’m just going to bask in the pleasure of confirming someone’s hypothesis rather than refuting it, or worse, saying that we can’t tell for sure.

[ad#AdSenseBanner]