Wednesday, March 09, 2005

Baseball Defensive Statistics ... Graphically

David Pinto of Baseball Musings is one of the better baseball bloggers out there. He's a smart guy with access to a ton of data (something not all of us are privvy to).

Over the offseason, he's been running a bunch of tables that show what he calls the Probablisitic Model of Range. Basically, he's got the direction and type of hit of every play in baseball in 2004 and runs it through a model to come up with the number of expected plays by each player versus the number of plays they actually made. Very cool stuff and also very similar to the UZR stat that MGL of the St. Louis Cardinals by way of Baseball Think Factory.

It's not a perfect stat -- there's some weirdness with certain positions seemingly being manned only by below average players and some only by above average players, but it's a worthwhile number to look at when looking at defense.

That's just it. Before today, it was just a number.

Today David broke that barrier. He's graphically represented the defensive capabilities of players. If you're a baseball fan/stathead, this is amazing stuff. He graphed both David "People Love Me Because I'm Scrappy and Suck" Eckstein and Cristian "You Want to Pay Me *How* Much, Mr. Bowden?" Guzman.

The graphs show it all. Eckstein isn't great moving to either side, but he makes the routine plays. He's awful catching foul pops on the 3rd base side. He's bad on liners. Basically, he's a pretty poor defensive SS (and this doesn't even address his arm). Guzman's graphs show a player who's pretty average on grounders to SS, but pretty great at snagging anything in the air, whether a pop or a liner.

Just awesome, awesome stuff. I'm slowly getting ready for baseball season. In fact, you should go read my analysis of the Milwaukee Brewers over at The House that Dewey Built. And check again later this week when I tear the Arizona Diamondbacks a new one.


Post a Comment

<< Home