Odd Thoughts

On Defensive Statistics

Obviously, there are a lot of problems, and the first one is a problem with the rules and the natural human error by score keepers across the game. If a routine fly ball is hit to the outfield, and the fielder just can't see it, and he doesn't touch the ball, it's not an error. If Jorge Posada hits a ball right at the second baseman in a double play situation, and the 2B gets the out and second and the SS throws to 1B, the ball beats Posada by ten steps, and the 1B drops the ball, there is no error because a double play cannot be assumed.

On the other hand, if Omar Vizquel makes a sliding glad on the second base side a of the bag, a ball no actual human can catch, and attempts to make a throw from his knees but can't cleanly get the ball out of his glove, that's an error. So right off the bat, there is a big practical problem concerning what is called or not called an error.

This plays some funny tricks onthe defensive statistics. Often an average fielder at SS will make fewer errors than another SS that is reputed to be great. This is simply because the better SS gets to more balls and has more chances to make an error. And then on top of the, the extra chances tend to come on particullarly difficult plays. So for Omar Vizquel to have more errors and possibly a worse fielding percentage than David Eckstien is not conclusive poof that Eckstien is the better fielder. If a person could regularly watch both players, Vizquel would quickly become their man regardless of what the flawed defensive statistics suggest.

Here's another example of a statistical problem: double plays. If you were to look right now at the MLB leaders in double plays turned, you might be allarmed to find Kansas City, Pittsburgh, and Arizona among the league leaders. Now, some good defensive pairs are up there to, but should we assume that Jack Wilson and Freddy Guzman are the second best defensive middle infield in baseball? I don't think so. The Twins are 4th and they're doing it with a hodge-podge of Nick Punto, Juan Castro, Jason Bartlett, Luis Rivas, and Michael Cuddyer splitting the MI duties. Are Angel Berroa and Ruben Gotay actually better defenders than Vizquel and Ray Durham? The notion is laughable, and yet the KC pair has turned 8 more DP's than the guys in SF. There's no way to truly explain every angle of this, but one reason you find teams like the Royals and Pirates at the top of the list is that their teams just aren't that good. This leads to having a lot more guys standing around on first base, which leads to a lot more DP opportunities.

Considering the problems caused by the recorded data, the best solution then, is to allow the people in the games to hand out Gold Gloves, and in fact, that's what they do. Gold Gloves are voted on by the managers in each league, but this creates another problem. The managers see one SS everyday, and 15 other SS some. If Jimmy Rollins makes one spectacular play in one particular game against the Padres, Bruce Bochy might vote Rollins for gold glove based on that one play. If Cesar Izturis makes one bonehead play against the Astros, but is brilliant the rest of the time, that one play could cost him the vote from Phil Garner.

The fact that managers do a bad job at selecting Gold Gloves is clear from the way that winning Gold Glove becomes a club that is hard to get out of. Josh mentioned that Edmonds had won a Gold Glove in the NL every year he's been there, and that's true. He won 2 in the AL, though his streak was interupted by an injury in 1999. Edmonds is in the club. He'll probably keep winning them until he gets to the point where he's not an everyday player or misses a big part of a season due to injury. Considering the stage of Edmonds career, if he misses half a season, he probably won't win another gold glove - he'll be out of the club, and Juan Pierre or somebody else will take over.

Barry Bonds won Gold Gloves 8 out of 9 years - as a left fielder; he was in the club. Ozzie Smith won 13 staright GG's from 1980 to 1992. Barry Larkin bore the GG torch at short after smith really slowed down. But in 1992, do you really think Smith, at age 37, was a better defender that Larkin, who was 28 that same year? No. Smith was in the club. Sinec Larkin's career began to decline, the NL has not had a regular SS Gold Glove candidate. Since Larkin won his last GG in 1996, the NL has had 5 different GG winners at SS.

Anyway, this is not to say that Gold Gloves are meaningless, they're just not as meaningful as they're made out to be. When you hear that a player like Jim Edmonds won 7 Gold Gloves in his career, it is enough to say, well Edmonds must have been one of the best defenders of his day. That is true. That doesn't mean he was the best defender every year (Andrus Jones has won 7 straight himself), or any year, but he might have been. There's just no statistical way to really measure it. And on top of that, MLB has the wrong group of guys voting on the award.

All that leaves you with in determining the best defensive players in your own impression, the comments of various experts, and some good common sense.

Posted by chefchuddy at September 9, 2005 03:13 PM | TrackBack