November 17, 2006
Minor League Defense

For all the giant steps that statistical analysis has taken in the last decade, prospect evaluation remains largely the domain of scouts. This is particularly true when considering defensive contributions; not only are defensive stats less established than offensive measures at any level, but the data available for the minor leagues is scant.

That is, until now. Using the 2006 play-by-play logs that power, I’ve begun to tackle this longstanding problem. I highly doubt that this project is going to put any professional scouts out of work, but it may open the door to new ways of looking at prospects, and it will certainly give fans more insight into an area that the scouts formerly monopolized.

For my first foray into defensive stats, I’ve used David Gassko’s Range. Range is based on a simple premise: the number of batted balls that are hit to each fielder depends on the handedness of the batters that each team faces. The math gets a little hairy once you delve into the details, but the concept is simple. Given a few basic stats about a team–the percentage of RHBs and LHBs faced, the number of grounders hit to second, third and shortstop, and the number of balls hit in the air to left, center, and right field–you can predict how many plays an average fielder would make.

From there, you can compare the actual number of plays made. For infielders (not including first basemen), Range looks only at groundball assists. For outfielders, it counts putouts. For the time being, I’m only doing those six positions–there’s plenty more work to be done before I can expand to minor league first basemen and catchers.

Without further ado, here’s a look at the Mets system at each of these six positions.

Second Base

In about 30 games at Triple-A, Anderson Hernandez was great, making 11 plays above average–one of the best performances in the high minors on a per-game basis. Jeff Keppinger was excellent before he was traded: +10 in about 50 games. Wilson Batista was nearly as good playing full-time in Binghmaton: 17 plays above average.

After that, it goes downhill fast. Both Ryan Coultas and Enrique Cruz were more than 10 plays below average in St. Lucie; Hector Pellot allegedly made 191 plays in Hagerstown, but should’ve made 39 more; and in Brooklyn, Jon Schemmel was five plays short of average. On the assumption that average defense is stronger in the higher-level minors, these guys have a hard road ahead of them.

Third Base

It probably doesn’t matter whether the Mets have anything in the system at third base, but there are a slew of guys at the position who may have more value that is commonly thought. Chase Lambin may have a Double-A bat, but he appears to have a Major League glove, making 16 plays above average in about 100 games between AA and AAA. Chris Basak has an equally low ceiling, but was +12 in only 200 innings, a rate that would make him among the best in the minors if he kept it up for a full season.

The low minors hold more promise: Jonathan Malo was +25 a half-season, and in Kingsport, Alejandro Zuaznabar was +29. I have no sense yet whether that means that Zuaznabar has a MLB-quality glove, or just that Appy league third basemen are generally awful. But if you’re going to have a little-touted third baseman in rookie ball, why not have one with gaudy defensive numbers?


Anderson Hernandez was almost as good at short as he was at second, making 17 plays above average in 600 innings. The Mets organizational depth may be just as meaningless here as it is at third, but it’s similarly strong: every SS in the system who played at 200 innings was above average for their level.

Notable among them were the two rookie ball shortstops. In Kingsport, Emmanuel Garcia was +15 in 420 innings, and in only 317 innings in the Gulf Coast League, Jonathan Santos came in at +11.

Left Field

I suspect that Range isn’t terribly worthwhile for corner outfielders, at least until it’s park-adjusted. (I haven’t done that yet.) It seems implausible that anyone could be quite so good as Jonel Pacheco, who apparently made 31 plays more than average in Hagerstown. 31!

Dustin Martin, Corey Coles, and Brahiam Maldonado all came in with positive double digits, while Jorge Padilla made 14 fewer plays than average.

Center Field

Like Carlos Gomez? Try liking him more. His 24 plays above average tied him for the best among Double-A centerfielders. (Though, to be fair, Jacoby Ellsbury came close in fewer than half as many innings.) Lastings Milledge also supplemented his bat with some nice glovework, rating a +5 in 464 innings.

After those two, there’s a whole bunch of mediocrity. Corey Coles managed a +6 in limited FSL action, but everyone else was below average. Fernando Martinez was +2 in St. Lucie but -8 over more innings in Hagerstown.

Right Field

Victor Diaz doesn’t appear to be a MLB-caliber defender in right: he was -17 in about 700 innings for the Tides. Jorge Padilla, on the other hand, turned in a much better performance in RF than he did in left, performing right at league average. Could be a quirk of the Binghamton park or it could just be the limitations of a small sample size.

Jonathan Sanchez is another study in contradictions: in Brooklyn, he was +13; in Hagerstown, he was -14. “Average” may be better in the Sally League than it is in the New York-Penn (though probably not by that much), or it could be due to the park.

In Conclusion…

As with any statistics, offensive or defensive, it would be great to have more data to go on, especially for guys like Anderson Hernandez who played multiple positions. The same problems appear with major league defensive stats. But as a glimpse into how some Mets prospects performed in 2006, these numbers suggest a few guys who are underrated, and a few others who can be completely removed from your radar screen.

12 Responses to “Minor League Defense”

  1. Comment posted by Jeff Sackmann on November 17, 2006 at 12:07 am (#168632)

    I meant to include this in the article, but forgot: Ben Johnson was just a smidge (less than a full play; it’s not a technical term) above average in 300 innings in CF last year. I’d imagine, then, he’ll be quite good playing the corners in Shea.

  2. Comment posted by tom totem on November 17, 2006 at 12:08 am (#168634)

    Boo yah!

  3. Gravatar
  4. Comment posted by Eric Simon on November 17, 2006 at 12:22 am (#168660)

    I’d imagine, then, he’ll be quite good playing the corners in Shea.

    He’ll undoubtedly make Shawn Green look like dogshit by comparison.

  5. Comment posted by Jeff Sackmann on November 17, 2006 at 12:34 am (#168671)

    He’ll undoubtedly make Shawn Green look like dog#### by comparison.

    Yes, well. I didn’t want to be trolling in the comments of my own article, but Julio Franco in right field would make Shawn Green look like dog#### by comparison.

  6. Comment posted by tom totem on November 17, 2006 at 12:37 am (#168673)

    I hear the Mets are looking at Rob Deer as a late-inning defensive sub in RF.

  7. Comment posted by The Real Marty on November 17, 2006 at 8:42 am (#168802)

    F-Mart is hurt. He was taken out of the game on Sat. after 1 AB and hasn’t played since. Any idea what’s wrong with him?

  8. Comment posted by BryanB on November 17, 2006 at 9:14 am (#168805)

    F-Mart is hurt. He was taken out of the game on Sat. after 1 AB and hasn’t played since. Any idea what’s wrong with him?

    Before being taken out he made two feilding errors. Don’t know if that’s related or not. He did have some injuries during the season, also.

  9. Comment posted by george on November 17, 2006 at 4:39 pm (#169279)

    victor diaz got traded.

  10. Comment posted by Udamn on November 17, 2006 at 9:46 pm (#169411)

    This is a really intersting area to look into stats wise. Maybe I am incorrect but couldn’t quality of fields and thing of that nature have a large effect on defensive statiscal analysis in the minors.

  11. Comment posted by n in NJ on November 18, 2006 at 11:40 am (#169593)

    hmmm… very interesting: ‘defensive sabremetrics’, I am not clear If the formula accounts for difference in ball speed/spin/difficulty to field from say a Hard hitter vs. soft hitter? using say the players batting average? or success with same hit against other fielders?

  12. Comment posted by Jose Reyes, RBI Machine on November 18, 2006 at 4:19 pm (#169725)

    Just spitballin’ here:

    Chase Lambin is excellent defensively, as is Corey Ragsdale. Just… neither of them can hit a lick. I saw them both in ST last year and they were great.

    I still think that Lastings Milledge is a great defensive player too. Big Monster aside, he was good last season and should get better- he’s got speed, makes apparently good reads, and isn’t afraid to dive.

    Carlos Gomez… yes. Awesome.

    The fact that FMart is basically average defensively already too is fantastic news. Man, he’s so good and so young.

  13. Comment posted by Sean Smith on November 22, 2006 at 11:06 am (#173166)

    I take that each player is compared to league average at his position?

    In that case we can’t assume that a plus defender in the minors will be even average in the majors. Though once we get more data on that, MLE’s for defense can eventually be calculated.

    Jeff, are you going to publish just the final result for these players or will you give us the whole stat line – innings, PO, AST, E, etc.?

    This is really cool. Probably on the level of Sean Forman’s splits on B-Ref. The only place I have found defensive stats from the minors is Baseball America’s book, 2 months after the season, and they only give you games, not innings. And they also don’t give you catcher SB/CS – is that something to look forward to on your site?