That way Jeter would have to give back every bogus Gold Glove award he’s ever won…namely ALL of them.

Jeter = Gold Glove

is like

George Steinbrenner = no commitment to winning

I just don’t get why this award isn’t given more national coverage. You’re the best in your league amongst all the gold glovers. Sick.

A pop up that is misplayed is a hit (often a double). A flyball that is misplayed is an XBH. A ground ball that should be an out becomes a hit because the defender has no range. Any time these things happen (and they happen repeatedly every single game), they affect ever offensive and pitching stat.

In addition, umpires miss a lot of calls on balls and strikes…a missed ball or strike call can have a huge effect on how the rest of the PA progresses, causing a gap between actual pitching performance and recorded pitching performance.

In short, everything people complain about in defensive stats already exist in hitting and pitching stats….people are just accustomed to those “problems” and understand that they are part of the system and generally will “even out” over time….sometimes they do, sometimes they do not.

It’s only one metric, but that’s Fangraphs’ UZR primer. BIS data is charted by interns off video (presumably DVR). While we don’t know all of the inputs, there are things we can reasonably speculate on based off what we know about UZR. For instance, the primer indicates that batted ball type and location are inputs, each composed of component pieces.

“One of the differences between UZR and linear weights is that with UZR the amount of credit that the fielder receives on each play, positive (if he makes an out) or negative (if he allows a hit or an ROE), depends on how often that particular kind of batted ball, in terms of its location, speed and several other factors [...]“.

Without going into the “other factors”, we can talk about batted ball type, location and speed [velocity]. BIS breaks down batted ball type into GB, IFFB, LD, “Fliner”, and FB. These are subjective judgments with a high susceptibility to human error. Further, not all batted balls of each type are created equal. Location is essentially plotted on a graph by the BIS interns with their software. There’s human error here as well. Further, the difficulty of, say, catching a line drive, can’t be accurately deduced from the batted ball input (a line drive is likely plotted where it lands or is caught; either BIS is projecting the landing location on line drives for balls that are caught, or the plotted locations are different based on whether the defender touches the ball or not). For velocity, BIS isn’t taking objective exit velocity to my knowledge. While there are ways of deducing batted ball velocity, they require input closer to what a company like Sportvision (Pitch, Hit, and Field FX using camera systems) or Trackman (cameras, sensor measurements using the Doppler effect) provides. To my knowledge, BIS doesn’t have the hardware, access, or a partnership with a company that has the hardware to give accurate velocity readings on batted balls. While I don’t know enough about their processes to completely rule out accurate velocity readings, it’s strongly in doubt for me given the technical difficulties of getting either the velocities or data from which the velocity can be deduced.

I could continue on the validity and reliability of defensive metrics, but I think I’ve given enough example to explain why defensive metrics are viewed on shakier ground than offensive or pitching metrics. I’m not making the argument that defense metrics are without merit (I don’t believe Ben was either), only that there are significant questions over the validity and reliability of defensive data. Ultimately, I think adding the defensive metrics will help the Gold Glove voting process (they’ve pretty clearly been a joke for a long time), but as Ben said, it is a surprising to see a defensive award be the first to embrace sabermetric methods.

That problem has not yet been solved for defensive data

This isn’t necessarily true. With the introduction of Field F/X data, it’s possible that all the problems for the public defensive stats has been removed, but none of us has access to it.

and most prone to inexplicable year-to-year variation?

Why is this an issue? Barry Bonds one year had a .688 SLG and the next had a .863. No one says that SLG is bad because it’s prone to year to year fluctuation.

It’s such a awesome award, I’m surprised it isn’t more publicized.

Batting average might be a junk stat if we couldn’t count on the official scorer to accurately judge a what’s a hit, and then record it in the official record. We would worry about the veracity of batting average data if we wondered whether pop-outs were being recorded as hits. That’s the problem with defensive data. The fundamental inputs are flawed because they often aren’t or can’t yet be accurately judged. The input data itself is bad. We don’t worry about the veracity of batted ball data because baseball solved that problem decades ago. That problem has not yet been solved for defensive data.

RBI is a junk stat (team dependent)
Batting average is a junk stat (uses wrong denominator and numerator)
Pitcher wins is a junk stat (team dependent)
ERA is a junk stat (ERA belongs to both a pitcher and his fielders, not just the pitcher)

There are plenty of junk stats that people still use to form opinions about pitching and hitting. Imperfect doesn’t mean the same as junk (= useless).

Current fielding stats are still far better than what most managers/coaches relay on (errors and fielding %).

“…the introduction of the Rawlings Platinum Glove Award to honor the best defender in each League among the Rawlings Gold Glove Award winners in each season.”

So essentially, it’s the defensive MVP. Sweet.