Good Morning
Good Morning

The continuing discussion of defense measurement

Houston's Jason Bourgeois makes a sliding catch against

Houston's Jason Bourgeois makes a sliding catch against the Braves during a spring training game in Kissimmee, Fla. (Mar. 1, 2011) Credit: Getty/Mike Ehrmann

 The Batting Stance Guy is always awesome, but this is his best work yet.

(Thanks, Twitter. You ain't too bad yourself.)

--I got my copy of Baseball Prospectus 2011 in the mail earlier this week, and it caught my attention before I even completed the "Statistical Introduction." The BP folks made a very significant decision when it came to evaluating defense:

"The general trend in the sabermetric community has been toward stats based on zone data, where human stringers record the type of batted ball (grounder, liner, fly ball) and its presumed landing location," the BP book reads. "The trouble is that this zone data, unlike the sorts of data that we use in the calculation of the statistics you see in this book, was never made publicly available; it was recorded by commercial data providers who kept the raw figures private, only disclosing them to a select few who paid larger sumes for the privilege. As additional data has come to light, we've found that the zone-based defensive metrics don't hold up especially well to outside scrutiny.

"Different data providers can come to very different conclusions about the same events, depending on the providers' recording practices and vantage points. Moreover, two metrics based on the same data can come to very different conclusions if the starting assumptions - assumptions that haven't been tested and that are derived from methods that can't be duplicated or verified by outside analysts - are different."

With such a write-up, BP casts doubt on the group of defensive measures that includes Defensive Runs Saved, UZR and Total Zone.

BP has decided to essentially take a step back and go with something "more granular," as they write. Fielding Runs Above Average (FRAA) utilizes play-by-play data. To cite once more from the book:

"Simply put, we count how many plays each player made, as well as the expected number of plays for the average player at each position, based upon the pitcher's estimated ground-ball tendencies and the handedness of the batter. There are also adjustments for park and the base-out situations: depending on whether there are runners on base and the number of outs, the shortstop may position himself differently, and we account for that in the average baselines."

Basically, BP is now working off results, rather than collected, first-hand observations. According to these measurements, Derek Jeter has been even worse defensively these past few years than measures like DRS, UZR and TZ have asserted. Here's the 2010 breakdown on Jeter.

DRS: -13 (13 runs worse than the average shortstop)

UZR: -4.7

TZ: -10

FRAA: -22

I was curious what Mitchel Lichtman, who came up with UZR, thought of this. So I e-mailed him. He wrote back:

"They make some good points and in the long run it probably avoids the biases associated with the zone data. In the short run, the zone stuff is probably better. Somewhere in the middle, they are probably equal. I don’t know where that point is, but if I had to guess, I would say 4 or 5 seasons of full-time data (depending on the position of course) might be the equilibrium point."

And so the challenge of measuring defense continues. It's by far the most difficult aspect of baseball to nail down.

But I love that people keep trying to nail it down. To constantly reassess the process.

Yes, the human eye has its value as we try to determine a player's defensive skill. It's one data point. It certainly shouldn't be our only data point, as the Gold Glove Awards show us each and every year.

Let the discussion continue. It's all driven, after all, by a love of baseball and a desire to better appreciate it.

--While Oliver Perez is approaching the end of his Mets career, Luis Castillo just might be buying some time with his play at second base.

As David Lennon reports, a scout who follows the team regularly opined, "Castillo by default," when asked whom he'd name as the starter.

This is a negative development for the Mets, as they knew what Castillo could do and hoped that Brad Eamus or Daniel Murphy could exceed that. If Castillo goes ahead and duplicates his 2009 season,...well, that wouldn't be the worst thing. But it's fair to wonder whether Castillo can reach even that modest level again.

--Joe Girardi said that the Yankees' bullpen could be superior to the 2009 crew. Agreed.

--Bartolo Colon pitched well for the Yankees, matching his competitors in the race for the final two spots in the Yankees' starting rotation.

--Miguel Cabrera allegedly threatened a restaurant manager and off-duty fish and wildlife officer before his DUI last month.

--Have a great day.

New York Sports