Biathlon Shooting Accuracy

As I pore over my biathlon data, I learn new things all the time.  For instance, I really had no idea what a typical biathlete manages for shooting accuracy.  For some reason, I would have guessed something in the 60-70% range.  But I was way off…

I stand corrected:

Here’s another version with contour lines:

This is based upon the previous three seasons of international biathlon races, both men and women.  Each dot represent the prone and standing accuracy of a single athlete over that time period.  Skiers with fewer than 100 shots were omitted.

Interestingly, you don’t see much difference if you split this up by gender:

I expanded my criteria for this last graph to include results from all season (~1992-present).  I wonder who the people are who have shot better standing than prone?  I’ll have to check that out in more detail at some point…


Related posts:

  1. Athlete Profile: Tim Burke
  2. Gadflies & Punching Bags
  3. Introducing: Biathlete Rankings
  4. More Ranking Graphs…
  5. Victims & Nemeses

About Joran


6 Responses to “Biathlon Shooting Accuracy”
  1. PAC says:

    “For instance, I really had no idea what a typical biathlete manages for shooting accuracy”.

    – I don’t understand why you show a scatterplot of prone versus standing shooting. If you want to learn about the shooting accuracy of a typical biathletes, an histogram would be more informative.

    – What is it exactly ? Do you have the average shooting accuracy for each biathlete over one race ? one year ? the whole career ?

    – It would also be nice to look at the individual variance

    • Joran says:

      The contour lines are a histogram, just a two-dimensional one. Histograms show probability distributions in one dimension, the contour lines show the joint distribution for both prone and standing together.

      I certainly could have done each separately (or shown a histogram of the marginal distributions) but the contour lines give you essentially the same information.

      Sorry for not being more clear about what I was plotting. It’s the shooting averages for each individual skier (since ~1992) with a minimum number of shots. For most skiers this represents their career averages, but for some skiers who retired in the early 90’s I probably only have part of their career.

      By individual variance, I presume you mean the variance within an individual (as opposed to between individuals, which is what I’ve shown). I certain can do that…I’ll put it on the schedule…

  2. PAC says:

    Ok. that’s nice.

  3. Mountainmums says:

    I was wondering how you could combine the shooting stats with overall race perf in order to plot something like average shooting accuracy against average race rank . I’d guess the problem would be controlling outliers within each racer’s data, especially in biathlon where results can vary from race to race a lot more than in XC.
    Food for thought I guess. BTW, I’m so jealous of your R / ggplot skills… :-)

    • Joran says:

      I’ve been wondering about that too. The more sophisticated race reports in biathlon do break down the skiing time separately, but those are usually buried in PDFs, which are a file format invented by an evil, evil person. So I’m usually stuck scraping html. In certain race types (only the individual, I believe) you can simply subtract the time penalties, since they don’t ski penalty loops. Otherwise, I’m not sure how you’d separate the two.

      The question of variability is a good one too. I have a post in the works that touches on that.

      R/ggplot are definitely worth learning. I’m relatively new to ggplot (usually in the past I’ve used lattice) but I’m really liking it so far.

  4. Roman says:

    Hi, could you gve me the link with the data for the graphs you’ve made.

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!