I read your most recent request for information on the topic of sabermetrics and am glad to help you further your understanding of the topic. In particular, you seemed interested in the topic of advanced metrics and how they impacted this year's Gold Glove and Silver Slugger Awards. We here on the dark side pride ourselves on excellent customer service, so I'll be happy to help you out with your request.
In particular, your questions seemed to revolve around the new SABR Defensive Index (SDI), which Rawlings commissioned to serve as part of the formula for determining who won this year's Gold Gloves at each position. In your request, you asked for a translation of SDI "into English."
The basics are that there are several systems out there for measuring defensive performance in baseball. (SDI kinda combines them all.) They vary somewhat in the data sources that they use and the technical assumptions that they make, but they generally have the same basic premise. The mark of a good defender is that he catches balls in the air when they are hit near him (or far away from him!) and he picks up ground balls and throws them to first base to record outs without doing silly things like dropping the ball or throwing it into the stands or letting the ball get past his diving body into left field.
Thanks to the wonders of technology, we have decent data on where balls are hit and how hard, and from that we can figure out how easy or hard that play was for the average defender in MLB. A two-hopper to third is easy to convert into an out, and if a third baseman makes a play on that ball, it's not particularly praise-worthy. We certainly don't ding him for making it, but he'll get more credit for the soft groundball to his left where he has to charge, scoop, and throw on the run. Not everyone can do that.
All defensive metrics come down to three questions: "How many balls should he have gotten to? How many did he actually get to and convert into outs? What is the difference between those two numbers?" We start with the assumption that a good fielder is a guy who turns more balls than average into outs (that is, after all, his job description), and work from there. That's the English version.
If you want to know all the gory mathematical details behind it, that'll take longer, but if you accept the logic above, the numbers are simply that idea expressed through mathematical formulae. That's it.
The reason that a fielding metr—right, you're not fond of that word…a fielding gobbledy goop helps the process is that while managers and coaches might see players on a daily basis, this doesn't necessarily make them the best voters. You point out correctly that there have been times when the coaches have gotten the Gold Gloves hilariously wrong (1999 AL first base winner Rafael Palmeiro, who played 28 games at first base that year). Coaches might have two series against a player during a year, and maybe see him make a dozen plays four months apart. And when they see these plays, are they really keeping track so that when Gold Glove time comes, they can get it right? It might be nice to have something a little more comprehensive.
Allow me to relate a somewhat imprecise, but instructive analogy. I went to high school at an all-male Catholic school in Cleveland. For reasons that have never been entirely clear to me, my school, despite having a female enrollment of zero, had a yearly election of a Homecoming Queen. The electors were the senior class (read: a bunch of 17-year-old boys). Now, there were several all-girls Catholic high schools in the city, and students from those schools were welcome to be part of our school marching band, plays, pit orchestra, and cheerleading squad, as well as community service projects that the school sponsored. The Homecoming Queen was supposed to be the young woman who contributed the most to the St. Edward High School community, presumably through these avenues. And if some group of people knew who was helping to contribute to the school, it would be the students, right?
Let's just say I don't believe that "contributed the most to the school" was always the deciding factor. Perhaps other, less noble, concerns were taken into account when voting. Once again, 17-year-old boys.
The moral of the story is that people can be blinded by other concerns that have nothing to do with anything when they vote. I think the meme that coaches might be blinded by plays that look nice on TV is overdone. They are better equipped than most people to understand that a guy who took a good route to the ball and didn't need to dive has a leg up on his diving counterpart. But the brutal truth is that even coaches might not have a lot of information when they vote as to who had a superior defensive year. During the time when they ostensibly were supposed to be collecting this information (during games?), they had other things to do, and they can't watch all of the 14 other games each night.
If our coach-voter does use the highlight shows to bolster his knowledge, then he will be exposed to what looks good on TV. Maybe he can see through it, but he'll hear that guy's name a bunch of times. More importantly, he won't hear the name of the guy who makes everything look so routine, because that kind of dependability is… boring. When people are pressed to come up with a name, they are more likely to go with a name that they have heard more often and with whom they associate positive images. That's just human nature, and probably why there have historically been so many repeat Gold Glove winners. I guess we can't blame the voters for being human.
If, back in high school, we had wanted to get a good idea of which young woman really had contributed the most to the school, we might measure things like how many hours she spent at the school or how many leadership positions she had taken in the past four years. We'd get a much more complete view of her involvement in the school. In the same way, fielding databases are able to see every single play in all 2,430 games, whether it made Baseball Tonight or not. That's something that not even the most baseball-crazed person can (or should) do. Even better, fielding gobbledy goops can compare everyone to each other on a similar scale, enabling one to compare apples to apples. In both fielding and Homecoming Queen voting, neither gobbledy goop is a perfect indicator of quality, but they’re both better than voting on who looked prettiest.
I hope this clears up any questions you had on the basics of how fielding gobbledy goops are calculated. If you have further questions, please feel free to contact us here on the dark side. We're always ready to help.
peace, love, happiness, banana pudding,
Russell A. Carleton
Author, Baseball Prospectus and Proud Nerd