AL Central Defensive Rankings After 90 Games

by royalsreview on Jul 22, 2009 4:00 AM EDT in News Comment 22 comments

The last time we checked on the AL Central's defensive ranks, it was June 1st, and the Royals were still sorta in contention. Since then, most teams have played another forty plus games, and the UZR data -- while far from perfect -- is drifting towards being meaningful.

Listed below is each AL Central team's AL rank in UZR at each position. The division leader is listed in bold.

Do you wanna guess who is dead last overall?

	KC	CLE	CWS	DET	MIN
1B	9th	4th	1st	3rd	8th
2B	13th	12th	9th	1st	14th
SS	14th	7th	4th	6th	11th
3B	8th	10th	14th	4th	1st
RF	14th	7th	11th	9th	13th
CF	7th	11th	5th	8th	9th
LF	1st	12th	7th	4th	10th
OF Overall	9th	11th	7th	5th	12th
Overall	14th	11th	8th	3rd	10th
Overall on June 1	13th	11th	10th	2nd	9th

Takeaways:

The AL Central is not a good defensive division. Detroit has been much improved, and the Twins have a good reputation for defense that isn't matched by the data this season, and that's it.
The Royals are bad.
Butler's defense has been better than expected, but the Royals are still below average defensively at first base, which is, weirdly, one of the stronger defensive positions in the Central. Paul Konerko? First?
All that TPJ time we endured still couldn't get the Royals out of the SS cellar.
Teahen deserves some props for being passable at third after returning to the position on short notice. Imagine the Royal infield if Teahen had slipped and Butler had been what we expected?
Congrats to Jose Guillen for catching and passing Jermaine Dye for worst RF in baseball.
David DeJesus!
It's curious that the mid-July numbers and the June 1 numbers are so close. Given the way people talk about single season defensive stats, you'd think we'd see tons of crazy changes in the data. Hasn't quite happened, yet.

0 recs | Comment 22 comments | Share

Story-email Email Printer Print

Comments

When you watch baseball games every single day,

its very apparent who can play defensively and who can’t.

by 2X2L on Jul 22, 2009 5:32 AM EDT reply actions 2 recs

Moore is an expert at playing defensively

He aggressively gets defensive with every brain-dead roster decision he makes. (zing!)

by sumajestad on Jul 22, 2009 6:07 AM EDT reply actions 0 recs

a sumajestad sighting!

you’ve been awol

by marbotty on Jul 22, 2009 9:24 AM EDT reply actions 0 recs

Francoeur is hitting .363 (12-for-33) with seven RBIs in eight games since being traded.

That's the worst lookin' hat I've ever seen...

by royalblue69 on Jul 22, 2009 9:25 AM EDT reply actions 0 recs

he just needed a change of scenery...

by royalsreview on Jul 22, 2009 2:18 PM EDT up reply actions 0 recs

He needed

Tony Bernazard ripping his shirt off and challenging him to fight.

http://sportsillustrated.cnn.com/2009/baseball/mlb/07/22/Tony.Bernazard/index.html

Baseball's that swingy stick game, right?

by royalsroyalsroyals on Jul 22, 2009 4:51 PM EDT up reply actions 0 recs

Think Dayton reads SI?

http://sportsillustrated.cnn.com/2009/writers/cliff_corcoran/07/21/gold.glovers/index.html

Especially like the lead-in.

Baseball's that swingy stick game, right?

by royalsroyalsroyals on Jul 22, 2009 9:35 AM EDT reply actions 1 recs

how old is Dayton?

he’s right on the cusp

by royalsreview on Jul 22, 2009 2:18 PM EDT up reply actions 0 recs

I don't understand how these defensive statistics are evaluated

I really don’t.

Relive Royals History at royalsretro.blogspot.com

by RoyalsRetro on Jul 22, 2009 10:23 AM EDT reply actions 1 recs

You beat me to it

I’m warming u to the stastical side of the game. I’m currently reading Baseball Between the Numbers, and through 3 chapters, I find myslef with little argument against their analysis, and I think that is true with the majority of the Sabermetric findings…

but……. Paul Konerko? First?

This is where I understand EXACTLY what Dayton Moore was saying by not understanding how Defensive Stats are accrued. There is NO WAY IN HELL that Paul Konerko is the best defensive 1B in the AL. Hell, Morneau, Carlos Pena, Youkilis, Morales and surely 3-4 more would have better range and softer hands than Konerko at this point in his career.

If I lined up 1000 professional scouts to watch all the AL 1B, how many would come to the Konerko is best conclusion? 100? 200?

When that eye in the sky things that was linked a couple weeks ago is up & running in every park, that is able to measure speed of the abll, distance to left or right a defender moves, etc, THEN I (and hopefully GMDM if he is stilla round) will believe defensive metrics. When they continue to pump out stats saying that Paul Konerko is the best defensive 1B in the AL, I’ll remain skeptical…

BOOM! ROASTED!

by GoBabies!! on Jul 22, 2009 10:43 AM EDT up reply actions 0 recs

A stats guy would say: let’s see how it looks as the season progresses. His UZR on the year so far is certainly out of line with his last seven years, and if it stays that way there will be some ‘splainin’ to do.

by 2X2L on Jul 22, 2009 11:14 AM EDT up reply actions 0 recs

well, for performance analysis, yes

for talent analysis, they’d say for defensive stats, because of sample size issues, for talent evaluations, you need 2-3 years to get an idea of where a guy is at. Konerko is at +11.4 UZR/150 so far. But the prior three years he was 0.8, -1.5, -0.7.

Fielders can go on “hot” and “cold” strea"(i.e., lucky/unlucky deviations from “true talent”) just like hitters. Konerko is doing it right now.

I'm not a sabermetrician, but I do play one at Driveline Mechanics.

Can't get enough of me? Check out my Twitter feed.

by devil_fingers on Jul 22, 2009 11:16 AM EDT up reply actions 0 recs

Yes, I was going to say something about how streakiness on offensive doesn’t raise eyebrows, but I didn’t make the comment because I have no idea how much of an offensive sample size is roughly comparable to half a season on defense. A month? Six weeks? Anyway, everybody knows not to get too excited when Joe Shlabotnik is leading the league in some hitting category or other on May 10.

The other important question about UZR is this: what is the actual range of the ratings among regular first baseman, best to worst? Roughly how many defensive plays separate them? Is it one play per week? One a month? If the difference isn’t really all that much, it’s easy to see how a Konerko might streak his way to superior performance over a limited span of time.

by 2X2L on Jul 22, 2009 11:34 AM EDT up reply actions 0 recs

I'm not totally sure

It depends. I’ve been reading about it a bit more lately, I’ll post more later, perhaps. One thing that is often repeated is that three years of defensive stats is the sample equivalent of one year of offensive stats.

Another thing I’ve heard is that 100 games (~2/3 season) of defensive stat is like 200 PA of offense (~1/3 season).

If you use FanGraphs three-year leaderboards, that’s pretty good, although that will only giveyou guys who have played in the lats three years. I exported it and (probably not doing it right) found the standard deviation for the three years (caveat emptor: unweighted and unadjusted for age!) of UZR/150 to be ~4.6.

Here’s the list.

Albert Pujols 8.4
Kevin Youkilis 8.2
Casey Kotchman 6.4
Todd Helton 5.3
Derrek Lee 3.6
Lyle Overbay 2.8
Ryan Howard 2.7
Mark Teixeira 1.8
Paul Konerko 1.8
Lance Berkman 1.5
Carlos Pena 1.2
Justin Morneau -0.3
Adam LaRoche -0.6
Carlos Delgado -1.3
Adrian Gonzalez -2.4
Ryan Garko -4.9
James Loney -5.2
Prince Fielder -8.5

I’m a newbie with this sort of stuff, and it’s probably a bit of a leap to assume that this is subject to “Gaussian” rules. But wtf— Only 3 guys at the extremes where even close to wo standard deviations away from average, and 7 guys around 1, with 11 of the 18 (61%) within one standard deviation which I guess, given the sample, is close to ``what we should expect (68%).

I'm not a sabermetrician, but I do play one at Driveline Mechanics.

Can't get enough of me? Check out my Twitter feed.

by devil_fingers on Jul 22, 2009 11:53 AM EDT up reply actions 0 recs

maybe someone has done this before

but what about looking at the standard deviation decrease in samples that include 1, 2, 3, 4 5 years of defensive stats and see if there is a plateau after 3 years?

"The life of a (Royals) fan must be lived forward but can only be understood backward" -- Kierkegaard (more or less)

by benfunke on Jul 22, 2009 12:21 PM EDT up reply actions 0 recs

Thanks for the numbers.

If it wasn’t clear from my comment above, I think that to gain broader acceptance, these ratings have to be have connected in some way with recognizable units of individual performance, e.g. a difference in UZR/150 of 1.0 roughly means that the guy is responsible for turning an additional 3 (or whatever the number is) defensive plays per month into outs. I guess it’s fine if the numbers differ from position to position; I have no idea of how UZR is normalized for different positions so I wouldn’t hazard a guess as to whether they would.

Now I know that it’s handy to translate them into “wins above replacement” for agglomeration with offensive stats, but the connection with observable individual performance is also highly desirable.

by 2X2L on Jul 22, 2009 12:40 PM EDT up reply actions 0 recs

conversion to runs

is either complciated or simple. The simple way is just to do something like plays made above/below average times 0.78. But that’s just a generic average. UZR does the translations individuals for certain positions and/or zones/types of plays (double plays, extra bases prevented/allowed, putouts (range), etc.

But runs are above/below average for the position during the season.

I'm not a sabermetrician, but I do play one at Driveline Mechanics.

Can't get enough of me? Check out my Twitter feed.

by devil_fingers on Jul 22, 2009 2:04 PM EDT up reply actions 0 recs

if you never allow yourself to be surprised

you’ll never learn anything

by royalsreview on Jul 22, 2009 2:19 PM EDT up reply actions 0 recs

except for what it feels like to never be surprised

"The life of a (Royals) fan must be lived forward but can only be understood backward" -- Kierkegaard (more or less)

by benfunke on Jul 22, 2009 5:59 PM EDT up reply actions 0 recs

Konerko

Konerko only leads the majors so far this year because he has done incredibly well catching the ball this year — only one error in 711 innings — which boosts his runs saved by almost two runs (1.7). And keep in mind UZR uses average relative to the position — Konerko is not a good defender when compared to a SS or even a 3B (Konerko came up as a 3B), but against the immobile brutes who have generally occupied 1B over the years (from which the data is derived), he holds his own.

by Gopherballs on Jul 22, 2009 2:20 PM EDT up reply actions 0 recs

How much do errors, a.k.a. “quirky categorizations of defensive failures by highly idiosyncratic official scorers”, factor in to UZR? I know that in Dewan’s system a ball in play not turned into an out is counted the same way regardless of the scorer’s decision. Is the same true of UZR?

by 2X2L on Jul 22, 2009 2:25 PM EDT up reply actions 0 recs

UZR on Fangraphs includes up to four components

Range Runs (RngR), which is the baseline, plus the added components of Error Runs (ErrR), Double Play Runs (DPR), and Outfield Arm Runs (ARM). Separate numbers are listed for each (if applicable, e.g., infielders will not have an ARM number). Error Runs is defined as:

ErrR (error runs): The number of runs above or below average a fielder is, determined by the number of errors he makes as compared to an average fielder at that position given the same distribution of balls in play.

The Fangraphs glossary has links to more information about UZR (both under Defense and Win Values). If you do not want to consider errors, just look at RngR.

by Gopherballs on Jul 22, 2009 3:26 PM EDT up reply actions 0 recs