Statistical Skier

Yesterday’s post sparked a little bit of discussion about how to measure the “best” performances. That wasn’t what I really intended, I was just trying to give a sense for some of the most extreme performances, relative to the median skier.

No matter what you do, events will conspire to put some “weird” races on lists like these that don’t necessarily match up with our notion of best performances. Frankly, I’m fine with that. Cross-country skiing is, as a coach of mine used to say (quoting another), “an outdoor sport”.

If you’re unsatisfied with my previous measure, we have a few options. We can either concoct some new and fancier way of measuring performance, or we can simply limit our original list in an ad hoc manner. For instance, we could measure the percent back from the median skier, the 25th percentile skier and the 2nd place skier and then take a weighted average of these values. This way we could give greater weight to performances that not only outpaced the field (median) but also outpaced the front of the field (25th percentile) and even the entire field (2nd place).

Or we can simply look at subsets of out original list. For instance, here’s the top thirty restricting ourselves to regular World Cup events:

[table id = 78 /]

Now we’ve eliminated all those WSC/OWG 10/15km races, many of which have weaker fields, and some by chance had bizarre weather. That Smirnov performance stands out to me. If you go back and check the results you’ll see that he won by over a minute. Over second place! Mühlegg’s pre-Olympic WC is still there, of course.

And now we see a bunch of other “unusual” races that we’d probably like to exclude. Six of these races had notably weak fields (the China and Canadian WCs). And so on. But the list is still kind of interesting.

[ad#AdSenseBanner]