Brewers' Team Defense?

by keephopealive on Jul 8, 2011 12:47 AM CDT

I just read this article on the importance of team defense and how the Brewers are attempting to buck the trend in pursuing a pennant with good pitching and hitting but abysmal defense:

The author talks about defensive efficiency, which is the ratio of outs to balls in play, as a good measure of team defense. This is pretty close to 1-BABIP. I am not a saber-head, but I suspected that defensive efficiency had as much to do with the type of balls that are hit, as well as plain old luck as it did with defense.

My basic check confirmed this. Here is what I did. I looked at the distribution of teams' hitting 1-BABIP, pretty close to the ratio of outs made to balls in play, call that "offensive efficiency". I then compared it to the distribution of defensive efficiency in the league. The idea was that if defensive efficiency was really driven by defense and not luck or the type of hits that pitchers allow, then we should see more dispersion in it than we do in "offensive inefficiency". That is, "offensive inefficiency" would have nothing to do with hitters, only the the defenses faced, but teams play fairly balanced schedules, and so this would average out a great deal, leaving less dispersion in my "offensive inefficiency" numbers.

In fact, when I look at the full year of 2010 data, Iit's not consistent with this story. The standard deviations across teams are almost identical, but actually slightly higher for offensive inefficiency: 0.0105 vs. 0.0095. This isn't likely due to the the slight difference between defensive efficiency and 1-BABIP, since the means are very similar (0.703 for offensive, 0.707 for defensive). Also interesting is that the correlation is almost zero (0.04), indicating that ballparks probably don't seem to drive much of the dispersion either. If Petco, for example, lead to higher BABIP because of its dimensions, then we should see the Padres having both low defensive efficiency but also low offensive inefficiency (higher offensive BABIP). We don't see that.

Anyway, as I said, I'm not a stat guy, but I was wondering what people thought. Is the author simply using a weak metric for team defense? Is there some error in my reasoning?

Tweet Comment 16 comments | 2 recs |

Do you like this story?

Comments

I think you are right

To me, Defensive Effiency/DER seems like a fairly outdated stat since, as you point out, it doesn’t look at the specific properties of the balls in play like more advanced stats like UZR or TZ do (probably because batted ball data wasn’t available when DER was developed a couple decades ago).

Basically every other metric has the Brewers as a team roughly average defensively this season (another indicator that DER might be a little outdated):
UZR/150: -1.8 (5th in NL)
TZ: +2 (6th in the NL)
DRS: -18 (9th in the NL)

I think that article is tailored to fit a preconception (which many Brewers fans have as well) that all of the Brewers are horrendous defenders, which, outside of Yuniesky Betancourt, isn’t really fair. They’re by no means good, and Weeks/Gomez/Morgan are the only starters that I would argue are plus defenders, but there’s a difference between below-average and “historically bad, omg!!!” like some people suggest.

With ‘08-’11 UZR/150s of -4.7 (Prince Fielder), -6.4 (Ryan Braun), -5.6 (Corey Hart), the Brewers who are generally editorialized as “horrible defenders” are poor defenders, sure, but they’re not Yuniesky Betancourt bad (-20.1 UZR/150 this season) or Lance Berkman bad (worse than a -20 UZR/150 at both 1B and RF this year). People act like, outside of centerfield, the Brewers have eight Adam Dunns out there, which is ridiculous.

"PLUSH ALERT: THERE WAS AN UNTUCKING AT FENWAY!"

by SRB on Jul 8, 2011 2:52 AM CDT reply actions 2 recs

thanks

Yeah, that is what I kind of suspected. Sometimes I think journalists have the pressure to produce, and they have the idea for the article, facts be damned.

by keephopealive on Jul 8, 2011 9:40 AM CDT up reply actions

It's results vs. true talent

If the claim is simply a team has shown defensive skills, then DER could be used to describe the outcome of a team’s defense so far. That’s a bit like using OBP to demonstrate that a player has been successful or unsuccessful without taking into account a BABIP aberration.

If one is trying to make a claim that a team is deficient in defensive talent, then DER seems kind of worthless. I haven’t seen a Team Defense UZR stat anywhere (I’m sure you’re well aware UZR is additive only within position), but it would be nice to have at some point.

Founder of the BCBCU - Est. 2011

by ecocd on Jul 8, 2011 10:20 AM CDT up reply actions

Fangraphs has team totals for UZR

Why doesn’t that work?

Get a ife broseph

by Supertramp on Jul 8, 2011 10:39 AM CDT up reply actions

They just sum the UZR of each individual player

But UZR is scaled by position, which means it’s only additive within position. For example, a ball in the outfield that is in zone for LF, but is instead caught out of zone by a CF causes a greater increase in UZR for the CF than if the LF had made the routine catch. Some measure of Team UZR would be different than the sum of the UZR of each player.

I base that on KL’s theory put forth in the offseason that OF defense might be better defined as a collective effort rather than the individual components. The poor defense in the corner outfielders for the Brewers makes Gomez’ defensive value even more valuable than if the Brewers were to have average or good corner outfielders. fWAR based on defense for Gomez may be understated when considering he plays with Braun and Hart (or worse, Kotsay). A team UZR would take that into account.

Founder of the BCBCU - Est. 2011

by ecocd on Jul 8, 2011 11:52 AM CDT up reply actions

Would it matter for IF though?

Very, very rarely does a ball get by an IF and another IF is able to make an out on it. Has there been actual evidence that a corner OF UZR is dependent on the CF is, or are you talking just theoretically?

I’d think if the UZR rating of a player is heavily influenced by the rest of his team, they’d have very little application on an individual basis – yet WAR and UZR are almost universally accepted among writers and fans.

Get a ife broseph

by Supertramp on Jul 8, 2011 12:28 PM CDT up reply actions

Thinking further

Adding up individual UZR makes more sense than looking at them individually. If Gomez takes a ball away from Braun, it will add value to Gomez and subtract value from Braun. If you look at them individually, it might be deceiving — but if you look at the team as a whole it all evens out.

Or I’m completely off-base. Hadn’t really considered it until I read your post.

Get a ife broseph

by Supertramp on Jul 8, 2011 12:32 PM CDT up reply actions

UZR probably gets more respect than it deserves

While it’s probably the best measure available with data available to the general public, It has some notable faults as a precise measure of defensive talent.

It does basically nothing (can’t do anything, really) to take into account defensive positioning. It throws out plays with a shift and I haven’t been able to find anything regarding moving to ‘double play depth’ or a 1st or 3rd baseman charging on a bunt. There are plenty of plays that involve slight changes in positioning that can make a large difference over the course of a year. The Brewers IF defensive stats are likely skewed pretty heavily by Roenicke’s defensive positioning schemes.

Another point is that it doesn’t necessarily measure individual tendencies. TH has noted that Yuni moves much better to his left than his right and it wouldn’t surprise me if the Brewers are positioning him a half-step more to his right than Wilson or Counsell. A step in one direction or the other will affect a player’s Zone Rating over the course of the season, but the total distance he covers to get to a ball is the same.

UZR passes the sniff test in the sense of Gomez being picked out as a fantastic defender and Jeter’s defense decreasing over time, but it’s not without its faults.
-——————
With regard to your point about Gomez and Braun, since an out was made on the play, I don’t believe it explicitly subtracts from Braun’s UZR. It counts as a non-play for him. If Gomez calls for it and ducks out of the way letting the ball drop in Braun’s zone, then it would count negatively for Braun (but not Gomez).

That said, a catch for Braun in zone wouldn’t result in as much of a gain in UZR as a catch for Gomez out of zone. Since UZR is scaled to the position, any given play has different value for each player. If Zone Rating were to get as much flash as something like batting average, we might see ball-hogging accusations similar to the whisperings that Ichiro is hurting his team by focusing on singles rather than hitting for power and making a few more outs.

If a person is trying to get a feel for a player’s true defensive talent level, they can’t technically even compare UZR across seasons, because they’re dependent upon all other players at the position. Tejada and Yuni put together might collectively raise every other SS’s UZR 0.1 points simply due to their gross ineptitude.

Don’t get me started on UZR/150…

Founder of the BCBCU - Est. 2011

by ecocd on Jul 8, 2011 1:23 PM CDT up reply actions

Nice summary

And I get most of that, but if you’re going to try and measure defense, you have to use something, right? Otherwise we’re stuck at fielding percentage and errors.

If I look at the team UZR and see the Brewers are 19th, I would think that I could likely say that the Brewers are somewhere around the 15th-25th best defense. As you mention, it seems to pass the smell test.

Despite the flaws you mention, I don’t have a major problem using it for rough guidelines how how good a team is on defense.

Get a ife broseph

by Supertramp on Jul 8, 2011 1:34 PM CDT up reply actions

"Best we have" is certailny better than nothing

To be fair, we have Plus-Minus and Zone Rating so we wouldn’t be completely reliant on fielding percentage. I agree that UZR is the best available, but I don’t see it as robust as something like wOBA; it seems like it’s starting to get as much respect as the mature offensive stats.

Founder of the BCBCU - Est. 2011

by ecocd on Jul 8, 2011 1:47 PM CDT up reply actions

Yup, I'd agree with that

Always find it somewhat interesting when people refer to WAR so exactly, like a guy with a 1.5 WAR is clearly better than a guy with 1.0 WAR – they could easily be the same player with a few random breaks on defense throwing off the UZR rating

Get a ife broseph

by Supertramp on Jul 8, 2011 2:04 PM CDT up reply actions

They all have flaws, and the biggest drawback to UZR might be that it fluctuates so much (so what is it measuring?)

I still think it’s the best defensive stat people have been able to come up with though. Certainly more advanced than DER.

"PLUSH ALERT: THERE WAS AN UNTUCKING AT FENWAY!"

by SRB on Jul 8, 2011 2:12 PM CDT up reply actions

Nicely done

for a guy who is “not a stats guy”, you sure used a lot of math/stats to investigate your theory and prove your point. Great job. You just may be the next Zoros McCracken.

"That's not a weird stat. Rickie is a run-scorer," Yost said. "It doesn't matter. It doesn't matter," Yost told reporters. "See, you guys have no concept. He's a run-scorer. So there's nothing weird about it. That's what he does."
BCB Fantasy Football League 1 Champ

by Hyatt on Jul 8, 2011 10:45 AM CDT via mobile reply actions

Comments For This Post Are Closed

FanPosts

Community blog posts and discussion.

Recent FanPosts

+ New FanPost All FanPosts >

NL Central Standings

	W	L	PCT	GB	STRK
Cincinnati Red Reporter	26	20	.565	0	Won 1
St. Louis Viva El Birdos	25	22	.531	1.5	Lost 3
Houston The Crawfish Boxes	22	24	.478	4	Lost 1
Pittsburgh Bucs Dugout	22	24	.478	4	Won 2
Milwaukee Brew Crew Ball	19	27	.413	7	Lost 1
Chicago Bleed Cubbie Blue	15	31	.326	11	Lost 11