Last summer, I wrote a post for ultimaterob.com about a system of ultimate frisbee statistics that I had devised. I used this system to score both an elite ultimate game between NexGen and Furious George, as well as a Calgary Ultimate men’s league game I had played in last spring. Some salient differences emerged between the two levels of play, but in both cases, it was pretty clear to me that what I had come up with was something that looked pretty damn cool. I wanted to see more.
Over the winter, I got together with the directors of the Calgary Ultimate Association and pitched them the idea of tabulating my stats for an entire league’s worth of games. They were keen on the idea, and we decided we would try it out for the CUA men’s spring league this year. This league is basically envisioned as a developmental/tune-up league to help guys either get in shape for the summer season or introduce them to competitive ultimate if they have never played in a men’s league before. The idea of quantifying what everybody was doing on the field thus seemed to fit in well with the “how am I doing and what can I do better?” vision for this league. The season is also relatively short (6 games) and the league is relatively small (10 teams), which meant that the amount of statistical data to be processed would not be too overwhelming. But most crucially, all of the spring league games are played on (Canadian) football fields with yardlines and stands for excellent filming angles. This meant that I could add yardage stats to every single throw I tabulated for my stats, and thereby add a dimension of distance to the analysis of every qualitative event which affects a game of ultimate.
We’re now two weeks into the spring league season, and I’m happy with the way that things are turning out so far. I have been posting game-by-game stats to a website located here: http://www.ultisheet.org/
It is interesting to scroll through the results as I work through the data each week. To a great extent, the top-level club guys dominate the “leaders” pages–as expected–but it is interesting to see how some players find their own niches. My friend Michael has been playing ultimate for all of a year, but nonetheless leads the league (by a lot) in yards per reception. (Hint: he’s fast.) Ultimate Rob took one look at the average pull distance stats and immediately identified which guys he ought to start working with to improve their distance. Stumbling upon statistical surprises like these is a bit like finding bounces in the wind–you never know exactly where they’re going to be until you throw a disc through it.
The response from the Calgary Ultimate community to this project has been fantastic, as well. None of this would have been possible without all sorts of help from a small army of volunteers who have been willing to film games, lend cameras, process data and even just help me identify players. There’s a great quote from an old Canadian writer that goes: be bold, and mighty forces will come to your aid. It has warmed my heart to find out over the past couple of weeks just how true this is.
With that in mind, I would also like to add that my vision for this project goes beyond just the small but friendly world of Calgary Ultimate. Before turning my attention to ultimate stats, I worked for many years in the realm of baseball statistics (check out www.basesproduced.com, if you’re a fan). One of the most important lessons that baseball statisticians have learned over the years is the value of sharing data.
Projects like Retrosheet and PitchFX have dramatically changed the way people understand baseball, and they have done it by simply putting as much game data as possible out there, in relatively unanalyzed form, for anybody with a good idea for a new statistic or analytical approach to work with. My own ultimate stats website tries to follow this example by posting the coded data from each game along with some basic processing scripts to help any fellow enthusiasts out there to get started in figuring things out about these games that I might never have even thought of in isolation.
The fact that we now live in a world where we have so much information about everything makes it even more remarkable that, in ultimate frisbee, we seem to have so little information about anything. I think that if we work together, we can change that, and maybe even become better ultimate players for it. And either way, I’m certain that we can make it a lot more fun to be a fan of the game. 🙂
Now that I’ve succeeded in putting together a basic interface for these stats, my next goal is to work on some analysis of what all these numbers seem to be telling us. I plan on writing up the basics of what I’ve found out so far in my next post for ultimaterob.com. In the meantime, I welcome you to go check out the stats that I’ve posted so far and let me know what you think.