Thursday, January 31, 2008

Brainstorm: Stats in Ultimate

Nation,

The call to action for this post is
  • Help me list key stats on o and d for teams and players
_____________________________________________________________
On "vacation" this week, but I really haven't had power the last few days due to the ice storm that has hit my native province. If interested, this site may sum it up.

Was looking at my UPA ultimate newsletter this week, and I have to say I was mostly impressed with the product. Great photos, complete coverage of the respective divisions, and a lot of great stuff to read. The writing was a little weak in the divisional write ups (right style, just a little less engaging and concise than possible) but I definitely like the overall product. Injury advice, UPA news, spirit awards, and a retrospective by Steve Mooney.


Sockeye Photo Courtesty http://www.seattlesockeye.org/

An interesting part of the newsletter is pages 30-31. Stats are listed for the finals of each division. Team and player totals were collected for the very basic:
  • Assists
  • Goals
  • D's (defensive takeaways)
  • Turnovers
Immediately my friends are thinking that I'm going to attack this because macro level stats can be worse (more misleading) than no stats at all. I admit that's my first thought. But then I realize that this is a good start.

Using stats to indicate true player contribution and value will never be perfect. So why do it? I'll give you the reason through a quote by Billy Beane:

I think the misconception about any statistical analysis is that you’re not going to be 100 percent correct,” Beane said. “What you’re trying to do is create an arbitrage … if you’re right 25 percent versus 20 percent you’ve created a 5 percent arbitrage opportunity. That’s really all you’re trying to do.”

Billy Beane
Source: ESPN/Ehrmann/WireImage


That's why its worth it.

I can tell you there is some observable trends that are of interesting note
  • The masters final was the most efficient. There were 11 turnovers and 6 d's.
  • The men's final in open had 44 turnovers and 15 d's. It is incredible that such talented teams were that inefficient.
  • Highly touted Ben Wiggins of Sockeye had a very inefficient final. He was credited with just one d, no goals, no assists, and 4 turnovers. This might be exactly the type of player that is undervalued by a 4 variable stat package.
  • Johnny Bravo's Parker Krug had 7 turnovers in the open final. Riot's Miranda Roth had 7 turnovers also in the Women's game. Their teams lost, probably because they didn't substitute these players off.
  • Coed had the worst turnover to d ratio of the four finals. This in spite of the fact the newsletter talks constantly of the improved level of the division.

So, I am hoping viewers can help me by listing all key stats on both players and teams for offense and defense.

Post on the site or e-mail me. Forward it on to your stathead friends.

2 comments:

Alistair (Allstar) Boyle said...
This comment has been removed by the author.
Alistair (Allstar) Boyle said...

Playing time is an obvious one that would be interesting to look over.

If you haven't you should plug through parinella's blog. He's pumped out numerous posts detailing historical DoG stats from various tournaments over the years.

http://parinella.blogspot.com/2006/01/whats-good-about-stats.html
http://parinella.blogspot.com/2006/01/perils-of-stats.html
http://parinella.blogspot.com/2005/11/how-would-you-assess-quality-of-play.html
Lots of interesting comments on the posts too.

There was one on playing time I can't seem to dig up. Anyways interesting reading.