More Fun with TotalZone!

by Vuff on Mar 17, 2010 12:18 PM EDT

So, after finding TotalZone when looking for a decent metric to evaluate Cal's defense a couple of days ago, I decided to see what it said about the 2009 O's, since if we look at defensive stats, we usually just use UZR. I was just going to post this in the Open Thread, but it got pretty long. Anyway, some interesting tidbits:

* Brian Roberts doesn't look good via TotalZone, either. I'm not positive, but I think it actually rates him worse than UZR does for 2009, and it definitely does for 2008 (when UZR is mostly unhappy with his double play runs). Between the stats, seeing Abreu make those plays in that ST game, and now the back issue, I'm getting really worried about B-Rob's defense going forward. I had been predicting a rebound, but I'm doubting that now.

* Cesar Izturis was awesome, but we knew that. Though among AL SS with 100+ games, it actually puts Marco Scutaro just in front of him.

* Melvin Mora was, indeed, above-average at the hot corner, if not by all that much.

* It doesn't like Nolan Reimold. Hopefully that'll change.

* It has Adam Jones as above-average, but he'd be below-average without his arm. I don't think the doctor's defensive off-year is anything to worry about, though.

* Nick Markakis has some weird numbers. In 2007 and 2008, he was a bit above average on the road and a bit below average at home. In 2009, he was well above average at home and far below average on the road. I don't really know what to make of that, since anything I come up with is purely conjecture, but it's interesting. Oh, and it actually has his arm a touch below average last year, but I don't expect that to last.

* It doesn't like Luke Scott at 1B, but tiny sample size and all that. He actually was a well-above-average RF with the Astros in 2007. And in LF, he's a touch above average for his career.

* It has Felix Pie way, way above average in CF (for both last year and his career) and still has him well above average in left. Granted, both are small sample sizes.

* It has Robert Andino worse than Izzy but still well above average at SS last year. This contrasts with his career numbers, but I wouldn't be at all surprised if the regular playing time really did help him.

Hopefully some of that was interesting enough to have been worth the read. Unfortunately, like most defensive metrics, it doesn't say anything about catchers or pitchers, so I can't touch on our switch-hitting future or ZAUN. I'm curious if anyone has any thoughts on the weird trend with Nick. Along those lines, I'm developing a theory... does anyone know how much time players usually spend getting acclimated to road ballparks?

FanPosts are user-created content and do not necessarily reflect the views of the editors of Camden Chat or SB Nation. They might, though.

Tweet Comment 4 comments | 2 recs |

Do you like this story?

Comments

The Nick Markakis numbers

may be off somewhat because of statistical errors. Most defensive stats need a fair amount of data, usually close to three years worth to be pretty accurate and therefore have large error bars. If Markakis missed a ball or two that he presumably should have gotten to that can cause a pretty big effect on the stats, especially one with one-sixth the data needed to be very accurate.

by OsandRoyals on Mar 17, 2010 2:29 PM EDT reply actions

Yeah, that's true.

And it’s probably the simplest explanation. It just stood out, because his ’07 and ’08 numbers were practically identical (aside from his greatly improved ’08 arm), and then his ’09 numbers were radically different.

Weaver's Fourth Law: Your most precious possessions on offense are your twenty-seven outs.

by Vuff on Mar 17, 2010 2:42 PM EDT up reply actions

For those of us who know nothing about it,

…what is TotalZone? How is it computed? And where are you getting your numbers? Is it a B-R stat?

"I doubt he could reach [second base]...mostly cuz his fucking arm was in Aybar's nuts." – twistedlogic

by zknower on Mar 18, 2010 12:16 PM EDT reply actions

Yep, it's from Baseball-Reference.

The full explanation is here. I’m by no means an expert on it, having just started looking into it on B-R. When possible, it works on similar principles to UZR and +/-, in that it looks at who fielded what kind of ball and where, and it then compares a player’s performance in making plays to the league average. It’s definitely not as sophisticated as UZR, because it relies mostly on simple Retrosheet data (i.e., any ground ball to left is just a ground ball to left). However, different methodologies are used depending on the available data; when hit data is unavailable, it gets a little weirder.

What makes it really neat, though, is that because it uses Retrosheet data, it’s available for historical players as well. So you actually can compare, say, Ripken and Jeter, or Brooks and Schmidt, or Blair and Jones. And the results, according to B-R (and in my own observations so far, for whatever that’s worth), correspond very well to UZR and +/-.

I’m planning on doing some historical comparisons in the future, so I need to look into the methodology in more detail, but it seems like a really cool stat to use.

Weaver's Fourth Law: Your most precious possessions on offense are your twenty-seven outs.

by Vuff on Mar 18, 2010 1:03 PM EDT up reply actions