Baseball Musings
Baseball Musings
January 31, 2005
Probabilistic Model of Range, Centerfielders

Here is the table lising 2004 centerfielders on the field for 1000 balls in play.

Probabilistic Model of Range, Centerfielders, 2004, 1000 balls in play.
PlayerInPlayActual OutsPredicted OutsDERPredicted DERDifference
Wily Mo Pena1211144 135.42 0.119 0.112 0.00708
Corey Patterson3830324 301.71 0.085 0.079 0.00582
Andruw Jones4164389 374.85 0.093 0.090 0.00340
Jay Payton3144333 322.31 0.106 0.103 0.00340
Grady Sizemore1033105 102.31 0.102 0.099 0.00261
Luis Terrero1443111 107.44 0.077 0.074 0.00247
Lew Ford1028101 99.32 0.098 0.097 0.00163
Vernon Wells3510327 321.70 0.093 0.092 0.00151
Mark Kotsay3809345 340.13 0.091 0.089 0.00128
Luis Matos2403221 218.19 0.092 0.091 0.00117
Tike Redman3643340 340.24 0.093 0.093 -0.00006
Preston Wilson1432118 118.13 0.082 0.082 -0.00009
Jim Edmonds3738314 314.49 0.084 0.084 -0.00013
Endy Chavez3304301 301.54 0.091 0.091 -0.00016
Marquis Grissom3799342 342.66 0.090 0.090 -0.00017
Mike Cameron3772354 355.96 0.094 0.094 -0.00052
Torii Hunter3346312 313.81 0.093 0.094 -0.00054
Nook P Logan1179117 119.19 0.099 0.101 -0.00185
Laynce Nix2752222 227.64 0.081 0.083 -0.00205
Milton Bradley2349230 234.97 0.098 0.100 -0.00212
Scott Podsednik4168392 400.93 0.094 0.096 -0.00214
Coco Crisp2472206 211.47 0.083 0.086 -0.00221
Rocco Baldelli3278342 349.51 0.104 0.107 -0.00229
Juan Pierre4257365 378.59 0.086 0.089 -0.00319
Kenny Lofton1657162 168.29 0.098 0.102 -0.00379
Marlon Byrd2268196 205.04 0.086 0.090 -0.00398
Craig Biggio1636134 140.68 0.082 0.086 -0.00408
Carlos Beltran4235397 415.38 0.094 0.098 -0.00434
Steve Finley4148359 377.62 0.087 0.091 -0.00449
Johnny Damon3792349 366.12 0.092 0.097 -0.00452
Aaron Rowand3117291 306.32 0.093 0.098 -0.00492
Jason Michaels100096 102.67 0.096 0.103 -0.00667
Jeromy Burnitz1622114 126.52 0.070 0.078 -0.00772
David DeJesus2361231 252.60 0.098 0.107 -0.00915
Randy Winn3304341 372.91 0.103 0.113 -0.00966
Alex Sanchez2082178 200.73 0.085 0.096 -0.01092
Ken Griffey Jr.2077173 199.64 0.083 0.096 -0.01283
Chone Figgins103592 105.30 0.089 0.102 -0.01285
Garret Anderson2393211 243.09 0.088 0.102 -0.01341
Bernie Williams2616214 255.18 0.082 0.098 -0.01574

The first thing I notice is that Andruw Jones is very good and Bernie Williams is very bad. So in this case the system appears to be getting the end points right. The thing that really surprises me is the equality of Biggio and Beltran. Here's a table comparing them just with the Astros:

PlayerInPlayActual OutsPredicted OutsDERPredicted DERDifference
Carlos Beltran2242200 203.15 0.089 0.091 -0.00140
Craig Biggio1636134 140.68 0.082 0.086 -0.00408
Jason Lane1258 8.54 0.064 0.068 -0.00436

Beltran was much better with the Astros than Biggio; his poor fielding was a result of his time with Kansas City last year.

PlayerInPlayActual OutsPredicted OutsDERPredicted DERDifference
Ruben Mateo13312 9.62 0.090 0.072 0.01789
Carlos Beltran1993197 212.23 0.099 0.106 -0.00764
David DeJesus2361231 252.60 0.098 0.107 -0.00915

So the question for the Mets is, which Beltran will show up in centerfield next season? Neither is better than Cameron, and one is a lot worse.


Posted by David Pinto at 01:40 PM | Defense | TrackBack (0)
Comments

I took David's fascinating work, applied some adjustments, and reformatted.

The adjustment is to lower the predicted outs by .00256 per team BIP for each CF.

Numbers below shows the outs above average, per 162 games. The "games" is team BIP divided by 4500.

Listed are all players with at least 70 "games".

Outs Games Player
38 138 Corey Patterson
27 150 Andruw Jones
27 113 Jay Payton
18 126 Vernon Wells
17 137 Mark Kotsay
17 87 Luis Matos
11 131 Tike Redman
11 135 Jim Edmonds
11 119 Endy Chavez
11 137 Marquis Grissom
9 136 Mike Cameron
9 120 Torii Hunter
2 99 Laynce Nix
2 85 Milton Bradley
2 150 Scott Podsednik
2 89 Coco Crisp
1 118 Rocco Baldelli
-3 153 Juan Pierre
-6 82 Marlon Byrd
-8 152 Carlos Beltran
-9 149 Steve Finley
-9 137 Johnny Damon
-11 112 Aaron Rowand
-30 85 David DeJesus
-32 119 Randy Winn
-38 75 Alex Sanchez
-46 75 Ken Griffey Jr.
-49 86 Garret Anderson
-59 94 Bernie Williams

IIRC, Aaron Rowand was great according to UZR, which matches fans' observations. But UZR loved Winn too, while fans hate him.

Nothing can save Junior and Bernie.

It's also good to remember that David is looking at slices and hardness of ball hit, and does not look at distance. I don't think David is getting a good approximation of "hang time" based on slice and hardness of ball hit.

There's no better stat that is not measured than "hang time". Distance or grid location, to me, is a necessity, esp for OF.

Posted by: tangotiger at January 31, 2005 03:40 PM

David, how can Cameron, Hunter, and Beltran all come up as "below average" CF's? At least the above comment makes two of them above average, which makes sense to me.

And who is Nook P Logan?

Posted by: Al at January 31, 2005 04:08 PM

I still don't see how most major leaguers can be below average fielders. Tangotiger's comment above helps a lot, but, like others have said previously, -0.1283 doesn't have much meaning to me.

David, can you do something with that? I'm sure it's only another simple mathematical adjustment so we laypeople can have an easier time putting our brains around all this? Thanks.

Posted by: sabernar at January 31, 2005 04:30 PM

Nook is/maybe the Tigers' CF. He was a mid/late season call up, IIRC. One reason why the Mud Hens finished the season 6-28. It was fun to watch him play though. Not a lot of pop in the bat, but very speedy. Nook itself is a nickname.

Posted by: Robert at January 31, 2005 06:56 PM

I hope Nook sticks. One of my favorite Dr. Seuss rhymes involves a Nook.


BUT
A Nook can't read
So a Nook can't cook.
So what good to a Nook
Is a hook cook book?

Posted by: David Pinto at January 31, 2005 07:02 PM

I have started doing some work on converting these figures to runs. The URL will take you to my blog where I'm doing that, the explanation is a few entries down ... the method (of converting these outs to runs) works far better for MIF than for OF, btw ...

Posted by: Los Angeles Waterloo of Black Hawk at January 31, 2005 07:43 PM

sabernar, the way that I turn these ratings into something meaningful is to look at the "Actual Outs" and "Predicted Outs" columns. That tells you how much better than "predicted" the player was, in real outs. I find that easier to digest than the rate.

I agree that it would be nice to at least have the 2004 average listed in the table, and preferably an outs above average or something. Sounds like David's already working on the latter.

Posted by: Vinay Kumar at February 1, 2005 12:28 AM

I want to believe these CF numbers are accurate. Finley's nowhere near a gold glover, but at least he'll be a huge improvement on having Garret Anderson in center.

Posted by: rallymonkey at February 1, 2005 10:41 AM

While I've posted my doubts about how the infielders compare in UZR and PMR, the outfield ratings may be different and still consistent.

In Finley's case, it could be he's only a little below average in recording outs, but he plays shallow so that the ones he misses results in extra bases more often, costing his team a bunch of runs.

Posted by: rallymonkey at February 1, 2005 10:48 AM

Having seen Beltran play in KC, I have serious doubts that he is a below average fielder. He gets to balls that he shouldn't regularly. He may not be as good as Cameron, but I have a lot of trouble believing he is below average. Is it possible that more balls are judged to be in his zone due to the limited range of KC's RF and LF's? Could it be possible that balls in play are scored to CF when they fall in because he's the closest fielder, even if the ball was in left? I have seen this happen several times and I'm curious how the BIP's are scored.

Posted by: Matthew at February 1, 2005 03:50 PM

They say Johnny Damon has no arm ,did Bernie Williams have one? Mickey Rivers?Could you please name the past Centerfielders since 1955

Posted by: GARY at December 27, 2005 08:38 AM
Post a comment









Remember personal info?