New Biathlon Individual Graphs

Biathlon’s cool in part because there’s so much more data to play with. One seemingly minor thing I’ve been struggling with is how best to graph an athlete’s overall race results and their shooting statistics on the same graph.

The obvious answer of just overlaying them and including a second y axis (since shooting accuracy and percent back are on very different scales) is a big no-no in the statistical graphics world. Dual y-axis graphs are very bad news, almost as bad as the hated pie graphs. In fact, they are so bad that the statistical graphing package I use literally can’t do them!

It took a little tinkering, but I think I like this solution just fine. Basically, I just put the two time series (percent back, shooting accuracy) in vertically grouped panels for each skier. The one drawback (so far) is that this means I can’t only extend this to multiple athletes by making the graph wider. It’s probably possible for me to “wrap” this around, but, well, you don’t care about the details. It’s complicated.

Anyway, as an example here are the graphs for the top five finishers from each of yesterday’s mass start World Cup races in Fort Kent. They’re big, so click on them for the full versions:

The top panels are actually percent back from the median skier (rather than from the winner) which I’ve learned to prefer. The blue line tracks the median result over time. Lower values are better. The bottom panels are shooting accuracy for each season, where higher values are obviously better.

One important piece of information that I decided not to squeeze on here is the number of shots, which early on for each skier can be very low. So, beware small sample size.

Related posts:

  1. Biathlon Briefs: Lowell Bailey and Laura Spector
  2. Is Biathlon More Volatile Than Cross Country?
  3. New Biathlon Race Snapshot Graphs
  4. US Biathlon Preview: Men
  5. US Biathlon Preview: Women

About Joran


3 Responses to “New Biathlon Individual Graphs”
  1. I like them! Although they are a bit difficult to read. Interesting development by Tora Berger…

    As always I’d like to see the 3-5 best compared in one graph if possible, but I guess it would probably become too cluttered.

    • Joran says:

      Do you mean having separate panels for each athlete makes it hard to read? Because I agree it’s more natural for many people to put that kind of data all on the same graph.

      I admit that I’m strongly influenced by the statistical graphics world, which has spent a lot of time figuring out the “right” way to display data; but the “right” way according to the experts doesn’t always seem natural to people. Splitting things into multiple panels is a very popular technique in my world, so I’m really used to it, but I forget that it can seem unwieldy if you’re not used to it.

      • WorldofXC says:

        What I mean is that I am often interested in comparing say the top 5 athletes to see their development over time. Or somebody new coming up compared to the established top athletes – to see what his/her strength is compared to the established top athletes. Or just see how Svendsen compares to Bøe directly in one graph. I know that I can find the data by comparing the different panels (and I often do), but getting the answers right in the face often makes it easier to understand what is behind ; and it makes it more available without having to think so much (OK, I’m lazy:). I know it is more work as you’ve got to first make the figures, then analyze them, and then make new figures, but it makes it easier to highlight specifics. But it is terrific as it is already – just a tip:)

Speak Your Mind

Tell us what you're thinking...
and oh, if you want a pic to show with your comment, go get a gravatar!