Tuesday, June 26, 2007

Zoning In On The Problem

Defensive statistics remain the Holy Grail of sabrmetrics. While any schmo can come up with a way of estimating offensive output, finding a single stat that isolates the play of an individual fielder has proven to be elusive. While they're still crunching, we're left to consider the ones they have, and one of the best -- though far from perfect -- is zone rating.

I won't bore you with all the details. (If you're a sicko who's interested, start here for some background, then follow the links!)

Essentially, numbers-crunching baseball haters have divided the field into distinct regions (zones, duh!), and they track where balls go, giving credit to fielders who make plays in their specified zones. So, if Dmitri Young has 10 balls hit right at him, and 5 of them skip under his glove, he'd have a zone rating of .500. Pretty simple, right?

The wrinkle in the stat, and where THT has improved it a bit (using work and numbers from some other smart people), is that fielders only receive credit for certain areas. So let's say that Nook Logan ranges way over to get a ball in the gap, saving a double, there's a chance that it wouldn't be 'in his zone' under the definition of the stat, and he wouldn't get any credit. Sounds pretty silly, huh?

THT now tells us how many balls players get to that they weren't credited with, so we can see if good ol' Nook is a ball stealin' machine.

Let's start with the big picture first. If you scroll to the bottom here, you'll see the team's defensive stats.

There, you can see that the Nats are actually below the league average in unearned runs. You also see that they're near the top in throwing errors -- thanks, Dmitri!

But it's the set of columns to the right that are interesting. It provides the team's zone rating, and then breaks it further down into infield and outfield.

Team-wise, they're one of the 5 or 6 least efficient defenses in terms of converting their zoned balls into outs. They're also not especially adept at getting to other balls, as their out-of-zone (OOZ) plays is league average.

The breakdown of infield/outfield shows that it's basically on the infield's backs. They're way below average on zone, and they've gotten to about 10 fewer OOZ balls than the average infield. The outfield, as you'd probably expect, does much better here.

So let's break it down a bit further.

Here's first base. Dmitri Young has turned just 67% of the balls hit to him into outs, which is better than only one other fielder (the usually reliable Derrek Lee). You can see there, too, what makes Albert Pujols so special: the man can field!

Here's the list of qualified 2B. Because of the time-sharing on the infield, neither Belliard nor Lopez qualify. But Lopez is at .817 with 11 OOZ and Belliard is at a terrific .851 with 10 OOZ plays. Lopez would be among the trailers, Belliard just on the high side of average.

On its own, Zimmerman's rating isn't special, but then you look at the number of out-of-zone plays he makes, and you see what's terrific about him. He's made 12 errors, but 7 of those have come on throws, some of which a more sure-handed 1B might've been able to save him on. Plus, he's second in the league in double plays, with 19. Despite the errors, he's solid.

Here's the list of qualified shortstops. Guzman and Lopez are virtually identical. Guzman is at .786 with 11 OOZ and Lopez is .779 with 7. Both of those, unfortunately, are at the very bottom of the league, as anyone who's watched the two of them field can attest. They're not making the plays they should make, and their combined 18 OOZ plays would be at the bottom of the table, too.

Thanks to Bowden's transaction finger, LF is a mess, too. Here's the league. Ryan Church is at .880 with 13 OOZ. In very limited opportunities, Ryan Langerhans is at a terrific .941 with four extra plays. Despite reputation and appearances, Church has been quite good. He's not elite out there, and he still makes some ghastly plays, but he's in the top tier of LFers defensively. Langerhans is in a different world, and the decision to bring him in late in games hasn't been a bad one based on the stats -- although, just based on the zone ratings, Langerhans has only "caught" one more ball than Church given the same opportunity.

So what about our friends in center? Here's the league (note the overrated bum at the top). Ryan Church -- would you believe? -- is at .925 with 11 OOZ plays. Nook Logan is right behind with a still terrific .882 and 10 OOZ. Langerhans has a lot less playing time, but he's right there with a .875 and 6. Church's number is terrific, leading the second tier of CFers behind Andruw's results. (The Andruw numbers do seem a bit fishy, but then so do Church's). Logan and Langerhans are still very good, but are kicking around the league average area. By and large, center field defense has been a strength, even if we can all think of 4 or 5 plays that we know should've been made.

Right field is a cinch, thanks to the stalwart, Austin Kearns. There, too, you see his value. He catches every ball he should, and tons that he shouldn't. If he could learn how to call of a back-charing infielder, he'd be the best in the league.

There's no real metric for catchers. But it is interesting to note that for all the troubles that Brian Schneider had throwing runners out early this year, there's only one team that has allowed fewer stolen bases. The team's CS% is right around league average, but they don't just run. Whether that's great work by the pitchers in holding runners on, or just the wariness of opposing batters to run on the arms, something's working, cutting off a part of the game that's more important in RFK than almost anywhere else.

  • So looking back, we've got terrible defense at first and short.

    We've got terrific defense in right and perhaps center and left.

    And we've got an average guy at second.

    Zimmerman's a tough call. Some of his numbers are very impressive, and some are scary. I DO know that he's better than he's shown this year.

    That's not a terrible defense, even if it's not an elite one.

    We're better than last year, and that, as much as anything, accounts for the play of this team and some of the surprising performances by the pitchers. If you took these guys and put them with last year's defense? ouch.

  • 6 Comments:

    • Prospectus ranks the Nats 9th in team defense for what it's worth:

      http://www.baseballprospectus.com/statistics/sortable/index.php?cid=204024

      By Anonymous MF, at 6/26/2007 3:18 PM  

    • Great write up off an article that I was trying to parse thru.

      Help me out - if we had, say, Adam Everett (or insert really good SS here) at SS and, say, healthy and good defensively (2005?) Nick Johnson at 1B, can we figure a way to calculate a number of runs prevented (or, simply the difference in runs allowed between the two different defensive configurations)?

      By Blogger Scott M. Collins, at 6/26/2007 3:45 PM  

    • That's a good idea.

      I'll do that in a separate post, just giving some ballpark estimates for some numbers.

      By Blogger Chris Needham, at 6/26/2007 3:49 PM  

    • Interesting.

      I know it's a somewhat crude measure, but the Nats have been consistently top five in the NL in defensive efficiency (probably what mf is saying MLB-wide, but I didn't follow the broken link).

      By Anonymous Basil, at 6/26/2007 7:46 PM  

    • Some of that's the differences in how they're scored. Plain zone rating only considers plays made in a zone, whereas Def. Eff'y typically looks at all plays.

      That's where the OOZ plays come into play, and if you're bored, you could sum up to see if the OOZ plays added in approximate the DE totals.

      By Blogger Chris Needham, at 6/26/2007 8:43 PM  

    • Thank you for elucidating on Zone Rating and its context and implications. Very informative for me. Thanx.

      By Blogger Bote Man, at 6/28/2007 11:12 PM  

    Post a Comment

    << Home