The Great Glove Debate: How Do You Measure Defense?

I was thinking recently about defense.

Defense seems to be the thing that we know least about in baseball. I have been reading a number of articles on the site recently that discuss the defensive merits of various players.

I’ve have been surprised to see that the numbers quoted almost always begin (and sometimes end), with fielding percentage.

Simply put, fielding percentage is not a valuable way to judge someone's defensive ability.

First off, look at a list of the greatest fielding percentages of all time. You will notice that, both in terms of seasons and careers, the list is heavily, heavily skewed towards contemporary players.

Whenever something like this happens, it requires further investigation. Most of the guys in baseball history with the highest lifetime batting averages died 50+ years ago, so does this mean that they had an inherent ability to hit for higher averages?

No, it means something else. It’s most likely that in their day, the league was weaker in certain areas. This allowed for excellent players to take advantage in a way that isn’t possible when the general quality of the game improves.

Fielding percentage has increased over the years. Obviously it is false to assume that all of baseball's best defenders have played in the past couple of decades.

A big production was made of the Rockies team being "the best defenders ever" because they made the fewest errors. What was not often mentioned is that the records were set by recent incarnations of the Red Sox, Orioles, and Mets.

The most important part of defense is not the avoidance of errors, but the recording of outs. If player A makes 25 errors at short, but gets to 540 balls, he is certainly more valuable than player B, who made 15 errors but got to only 390 balls.

Player A would be significantly more valuable, although B's fielding percentage would be far better than player A.

The next part is a little trickier. There are those players who have a reputation of defensive greatness (Derek Jeter comes to mind, although if you watch him play a lot, you realize that he doesn't even LOOK like he is good out there, other than that one deep-throw-in-the-hole play, which is admittedly awesome looking).

This is why people suggest that Roberto Alomar was a better defender than Ryne Sandberg. He just LOOKS so damn good out there. He may have been better, but most of the metrics don't see players by what we see. This makes no sense if one thinks about it.

Is the man with the prettiest swing necessarily the best hitter? Of course not. Is the man who hit the farthest home run the best home run hitter? Not necessarily.

By the same token, why is the man who makes the prettiest play necessarily the best defensive player? He is not (although it should be noted that perception does often match reality).

When viewing defense, it’s important to pay attention to things like BP's FRAA, John Dewan's +/- numbers, range factor, etc., and to forget about who looks pretty or makes the fewest errors.