Statistical Skier

Since I’ve been making some graphs now and then for the folks at skidpepp.se, I’ve been paying more attention to the Swedish skiers.  I noticed that one performance last weekend that turned some heads was Maria Rydqvist’s second place finish in the distance race in Bruksvallarna, behind Charlotte Kalla.  Of course, everyone was quite some distance behind Kalla, but still, it attracted some attention.

This result seemed to surprise a few people (although that could also be Google Translate misleading me…) and led to her being selected to join the Swedish National Team at the World Cup races this weekend, so I thought I’d look a little more closely at her results.

From what I can gather, Rydqvist just recently had a child (as recently as four months, if Google translate isn’t tricking me) and so she didn’t race much last season, as we’ll see in a bit.  At least, I’m assuming it was because of her pregnancy; feel free to correct me in the comments!

Let’s start by looking at just her performance in this one distance race against this particular field.  Using the same style of graph from earlier this week, we get this:

Continue reading ›

So on Monday I posted some graphs that were intended to provide some context for some of the distance race results over the weekend.  Sprinting wasn’t left out in the cold because I dislike it (although, to be fair, I’m sort of a traditionalist in this respect; I will always love distance racing more, sorry!), but because it’s much harder to do meaningful analyses on sprint races.

But it’s just not fair for me to ignore sprinting, so even if it’s not ideal, I should at least try.  What follows is essentially the same analysis I did for the distance events, but for the qualification round of the sprint races only.  So all of this only applies to one’s ability to qualify, relative to the other competitors.  An important aspect of sprinting, to be sure, but only part of the story.

Since the qualification round is really just a mini-interval start race, we can treat it that way and make graphs just like I did on Monday.  I’m going to focus on three people whose performances I found particularly interesting: Kikkan Randall, Ida Sargent and Torin Koos.

First up, the most clear cut case of Kikkan Randall’s sprint qualification result in Muonio: Continue reading ›

This past weekend was chock full of racing in Finland, Norway and Sweden.  There’s way too much for me to analyze, even looking at just the North Americans, so I’m going to focus in on just a few performances that I found particularly interesting.  Even so, this post will be quite long.

Results ranged from Kikkan Randall and Kris Freeman’s victories in the sprint and classic distance events to some rough races for Liz Stephen and Morgan Arritola (who I believe I read somewhere was sick earlier in the week).  The Canadians also had some promising results in Sweden.  It’s easy for people to get carried away, both positively and negatively, after just a few early season races.  In particular, it can be difficult to assess what exactly these performances mean when we have to wait a few hours (or days) for FIS point calculations and even then we have to accept that the penalty calculation is accurately adjusting for the strength of the field.

First, let’s say straight away that Kikkan’s sprint victory was unambiguously a stellar result.  It’s hard to ignore a race like that over skiers like Justyna Kowalczyk and Petra Majdic (among others) even allowing for the possibility that some of them may not have had their best days.  Personally, I’m unsurprised that Randall is capable of a sprint performance that strong; I’m more interested in whether she can perform at that level more than a handful of times a season, which would represent the next obvious jump for her.

In this post, though, I’m going to focus on just the distance races for a few North Americans who did particularly well.

What I end up seeing when people try to make sense of early season distance results are lots of pairwise comparisons:

• She was 30 seconds behind the winner
• He beat Really Famous Skier by 15 seconds
• She was only 8 seconds out of Nth place
• He was only 13 seconds behind Really Famous Skier

This is a reasonable way to think about race results, but we’re assuming that the people we’re comparing ourselves to had “normal” races.  This clearly isn’t always the case and often we don’t even really have a great sense of what’s “normal” for each skier.  Can we do better? Continue reading ›

SkiTrax and FIS are holding a fantasy World Cup contest, roughly modeled after the very popular fantasy baseball or fantasy football competitions that we all know and love.  Basically, you pick four men and four women (plus a “bonus” man and woman, from outside the top 15 rankings from last year), and then your “team” accumulates points over the season according to a fairly involved point system.

This is the first of three such contests, apparently, the other two focusing on the Tour de Ski and the Oslo World Championships.

Fantasy sports competitions are simply awash in numbers.  People can spend hours pouring over stats on their players.  It can get pretty involved.  I’ve decided not to be selfish and use my data and statistical know-how to utterly dominate this contest, which quite clearly, I could.  (Just kidding!)  Instead, I’m going to be generous and provide a little simple analysis that might help you decide how to pick your own team.

I’ll save a more detailed discussion of the the rules for awarding points for later in this post.  The rules themselves can be found at the first link above, and they are fairly complex (and a tad ambiguous, if you ask me).  There are basically two kinds of points,

1. Points awarded to each individual skier based on their race performances
2. Points awarded to your entire team of skiers based on some group characteristic

What I’m going to do here is calculate, as best as I can, the points that each athlete would have contributed to a team based on last season’s results, so I’m ignoring type (2) points.  This requires a few simple modifications,

• Points for OWG will be swapped in for WSC results
• For simplicity, I’m going to ignore the U23 bonus and the “abandonment” penalty for starting but not completing the World Cup Final
• Doping penalties shouldn’t be an issue, since my database has already had those skier’s results removed

There are some complicated issues that come up in order to actually calculate these, involving what to do about ties, and the order in which we apply various point rules.  After a few emails with the SkiTrax folks, though, I think I’ve figured it all out and here’s what you get for the men:

Continue reading ›