A few odds and ends by request
July 7, 2007 5 Comments
A few things that people have asked for that I got to. After I posted my study on whether baseball players were getting bigger and slower (using Body-Mass Index as a measure of how big a player is), StatSpeak reader and frequent commenter Guy asked whether there were any patterns by position. Interesting question.
I classified players by the position at which they played the most games in the year in question as given the Lahman database. My biggest regret is that all outfielders had to be lumped together (the Lahman database doesn’t begin separating them until 2000). I’m guessing that the average center fielder looks a little different than the average left fielder in terms of BMI, but the picutre is interesting nonetheless.
- Well now, there are clearly three bands of positions. Catchers, first basemen, and designated hitters are the biggest players. Pitchers, third basemen, and outfielders are below them, and middle infielders are the skinniest of all. In the early 80s, DH’s were slowly working their way down into the middle band, and then went back to being comparable to catchers and first basemen. Then, in the mid-90s, something happened to DHs. I believe the yellow line says it all.
- Outfielders have also been on an upward trend from the early 80s onward. I’m not sure if this has something to do with the stadium building boom in the 90s with tiny ballparks and tiny outfields, but smaller outfields mean that a team doesn’t need as mobile a fleet to patrol it.
- The other interesting trend line is that second basemen apparently don’t want to be seen with the shortstops any more and are moving slowly up the ladder toward the middle band.
Update!: Thanks to a tip from Tango Tiger on my last post on BMI, I was able to solve the problem of all outfielders being lumped together. Here, then, are the seasonal trends of outfielder BMI broken down by the outfield position in which the player appeared most in that year.
Center fielders have gotten a bit bigger over the years, but corner guys have seen a huge growth spurt. Notice that in the early 1980′s, all three outfielders were roughly the same size. By the late 80′s, the only one staying on the slim side was the center fielder, while his corner patrolling brethren got huge. Thanks, Tom.
Back to your regularly schedule post:
A little more on closers. Commenter DanC on my previous post on the worth of closers asked whether a closer has some worth in that he is perceived to be good, and in a close game, he has an advantage for that reason. It’s hard to measure what a batter thinks of the pitcher standing 60 feet away from him, but the idea here is whether a closer (or any other pitcher) has some sort of advantage with the game more on the line. Sounds like… clutch hitting… just without the hitting.
Are there really clutch relievers out there? Are there guys who turn it on when the game is on the line? I looked at this one using a methodology similar to what I used in a piece I did on clutch hitting a while back. Basically, I’ve outright stolen the definition of clutch from David Appleman of FanGraphs (go there if you haven’t yet!) and Tom Tango. The measure takes advantage of the concepts of win probability added (how much a single play or event changes the probabilities that one team or the other will win) and leverage index (how much this particular play) matters in determining the outcome of the game.
I isolated all plate appearances in which the starting pitcher was not on the mound at the time (in other words, a relief pitcher) from 2000-2006. Then, I restricted the sample to those with 100 batters faced (in relief) for the year in question. I assigned all blame or credit for what happened while the pitcher was on the mound to the pitcher. That’s a dubious assumption, but in science, we sometimes have to sacrifice precision for direction. As I am wont to do, I then looked at the intraclass correlation over the seven years for reliever clutch performance. The result: .055
Performance in relief is thus about 3/10ths of one percent clutch ability and 99.7% other things… like general talent for pitching mixed with a little bit of luck. There’s very little evidence that there are relievers who turn it up a notch when the game is on the line. Closers are generally very good relief pitchers, and like “clutch” hitters, their good results with the game on the line are actually much more of a function of their being pretty good pitchers than having some special ability (whether intimidation factor or an extra gear) when the game is on the line.