Baseball for the Thinking Fan

Login | Register | Feedback

btf_logo
You are here > Home > Sox Therapy > Discussion
Sox Therapy
— Where Thinking Red Sox Fans Obsess about the Sox

Reader Comments and Retorts

Go to end of page

Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.

   1. Halofan Posted: November 02, 2005 at 12:16 AM (#1715160)
Manny Ramirez for Bill Stoneman, straight up, no Steve Finley salary balancer necessary.
   2. Dr. Vaux Posted: November 02, 2005 at 12:18 AM (#1715165)
How about hiring DePodesta to be the personel and numbers guy, and somebody else to be the "people person?" Lucchino can handle being the media contact guy. Really, it seems like DePo in Boston could work, because the GM doesn't have to be the face of the franchise. Gillick looks like he's off the list; Ken Rosenthal says he's going to sign with the Phillies (which must mean no one in his right mind would work from Frank McCourt; I can't imagine that a highly-touted 68-year-old would choose Philadelphia over Los Angeles, given the markets, payrolls, climates, etc. otherwise).
   3. Dr. Vaux Posted: November 02, 2005 at 12:22 AM (#1715170)
The question, really, is why he would take the Phillies job either, now that the Red Sox are available. In either case you've got a stupid, meddling boss, but with the Red Sox you've got a better team to start with, a better farm system, a bigger payroll, and one smart boss. You've also got a blood-thirsty media, though. Gillick's only worked in places with an easy media. Philadelphia will easilly be the toughest media he's worked in front of, with Baltimore surprisingly easy, and Seattle and Toronto both being total pushovers.
   4. 1k5v3L Posted: November 02, 2005 at 12:24 AM (#1715172)
Joe Garagiola, Jr!

heh heh
   5. 1k5v3L Posted: November 02, 2005 at 12:28 AM (#1715181)
How about Peter Gammons? He's well versed at navigating among the d!kheads in boston's front office and he's well trained at toeing the party line. And he'll be cheap too; half a dozen beers a day and a spit bucket.
   6. Maury Brown Posted: November 02, 2005 at 12:38 AM (#1715197)
Towers.

He worked with Lucchino prior, so he's not going to blind-sided. Only reason I see him back out of the idea, if it came calling, is if he wants to be able to have more control than what he'll get in Boston.

As far as DePodesta... Not exactly been impressed with his work with the Dodgers.
   7. Buster Olney the Lonely Posted: November 02, 2005 at 01:13 AM (#1715239)
Slightly off-topic: When are the GM meetings?
   8. Darren Posted: November 02, 2005 at 01:28 AM (#1715255)
Ya know, some people were suggesting Gammons over at SOSH. Not kidding.

Eric M Van for GM. He's now an expert on the inner workings of the Red Sox, and he could use studies of players 3-week splits to determine who's "made the leap!" He could also impose curfews on the players who need it!

Or how about Eric M. Van and Mike Gimbel as co-GMS?
   9. Mike Emeigh Posted: November 02, 2005 at 02:31 AM (#1715314)
Tell me what the role of the GM is going to be in a Lucchino-run organziation before you tell me who should be the GM - because that will define the skill set that's needed for the job.

-- MWE
   10. Harold can be a fun sponge Posted: November 02, 2005 at 02:36 AM (#1715320)
Towers.

He worked with Lucchino prior, so he's not going to blind-sided. Only reason I see him back out of the idea, if it came calling, is if he wants to be able to have more control than what he'll get in Boston.


And that's a potential problem for him if he stays in San Diego, as he's getting squeezed on both ends here between Alderson and Fuson.

When he interviewed with Arizona, all the rumors here were that Fuson would get the GM job if Towers left; I imagine it must be a little frustrating to be on the verge of getting the big job, only to be shuffled back.
   11. Шĥy Posted: November 02, 2005 at 02:36 AM (#1715322)
He could also impose curfews on the players who need it!

My favorite of the few SOSH threads that I have read was the one where there were about 50 posts discussing whether Arroyo should be forced to wear a long sleeved shirt when he pitches.
   12. RP Posted: November 02, 2005 at 02:47 AM (#1715334)
Here's what should happen: Red Sox hire Duquette away from the Orioles. Red Sox give the Orioles a prospect -- say, Lester -- as compensation. Orioles hire Depodesta to replace Duquette as Flanagan's righthand man. Everybody wins!
   13. Baseball Crank Posted: November 02, 2005 at 03:09 AM (#1715351)
We Mets fans need Duquette to go to Tampa so he can give us Kazmir back.
   14. Шĥy Posted: November 02, 2005 at 03:17 AM (#1715359)
We Mets fans need Duquette to go to Tampa so he can give us Kazmir back.

I'll settle with him going to Boston and giving the Mets Petunia for Zambrano.
   15. Darren Posted: November 02, 2005 at 03:17 AM (#1715360)
Why does RP want to make me cry?
   16. Darren Posted: November 02, 2005 at 03:19 AM (#1715361)
Good thing the Mets got rid of Duquette before he did something stupid like trade for Mienkeweiwioavntz.
   17. karlmagnus Posted: November 02, 2005 at 03:37 AM (#1715373)
Dan Duquette. Just as good as Theo if not better, won't take #### from Lucchino and will wreak his revenge on Gammons and CHB. No More Mister Nice Guy!
   18. Rafael Bellylard: Built like a Panda. Posted: November 02, 2005 at 03:41 AM (#1715375)
ESPN announced a rumor that Epstein and the Dodgers may be talking.
   19. The George Sherrill Selection Posted: November 02, 2005 at 03:46 AM (#1715379)
ESPN announced a rumor that Epstein and the Dodgers may be talking.

Will they let Theo bring his computer along?
   20. Champions Table Posted: November 02, 2005 at 04:08 AM (#1715391)
Why would the Dodgers want Theo but not DePo? Aren't they both guys that Tommy Lasorda would make meatballs out of?

The GM meetings are Nov. 7-11.
   21. Joel W Posted: November 02, 2005 at 04:22 AM (#1715401)
frankly, Gammons would actually be perfect. Everybody loves him and knows him, including LL. He's actually become more SABR oriented of late, even if he'd err on the personality side. I know I'd prefer DePo, but I think we could do worse than Gammo. My college roommate has been wanting gammo to get a job somewhere since i've known him.
   22. The George Sherrill Selection Posted: November 02, 2005 at 04:56 AM (#1715429)
If Gammons was a GM, would he kiss his own ass?
   23. Matt Clement of Alexandria Posted: November 02, 2005 at 05:07 AM (#1715436)
This is going to sound crazy, but let me just toss a name out there...

Jeter!!

Sure, he's employed by the Yankees... during the season. But in the offseason, I'm sure he could take evening conference calls, impart calm-eyed leadership and do all the little things that the best GMs do, the things that don't show up in the WARP3 charts.
   24. IronChef Chris Wok Posted: November 02, 2005 at 05:09 AM (#1715437)
Curt Schilling
   25. Dr. Vaux Posted: November 02, 2005 at 05:50 AM (#1715463)
Yeah, Gammons probably would be a decent GM. As good as half the guys with jobs, at the very least. Which, when you think about it, makes him a great baseball writer, since most of them would be the worst GM ever.

Epstein to the Dodgers: that could happen, because he's won a WS. That makes him legitimate in the eyes of folks like Lasorda and Placshke. Why would he put up with McCourt and Lasorda, though, if he won't put up with Lucchino?
   26. OlePerfesser Posted: November 02, 2005 at 04:28 PM (#1715815)
Yeah, Gammons probably would be a decent GM.

That Insider piece Gammo wrote sure seemed like he's campaigning for the job, Vaux.

And when you think about it, maybe he that's not so far-fetched. He's already "in the Hall," being on TV must be losing its thrill... why not prove your genius in another sphere?

Damn, now I've talked myself into being really worried...
   27. Norcan Posted: November 02, 2005 at 04:32 PM (#1715825)
I would be thrilled with Grady Fuson. He's done a good job scouting talent wherever he's been. But I highly doubt the Red Sox will consider him. I just can't see Lucchino being comfortable trying to have the same boss-mentor GM (young 28 theo, young 35 Towers) dynamic with a man his same age.
   28. Joel W Posted: November 02, 2005 at 04:40 PM (#1715844)
I know this is weird on a site like this (maybe not), and I think I've been influenced by Philly on SoSH in this regard: but as teams become more stat savy, is it possible we'd rather have a great scouting-type GM. Eventually, the stat stuff is going to be easy, and I hope Henry is smart enough to hire a guy who can use the stat stuff that his lower-level employees provide. The hard thing to do is to look at 18 year old and 21 year old guys and see which ones are actually going to be good, to see 25-year-old bullpen help and pick them up after they've had 6 years in the minors. Isn't it this type of thing that GMs will increasingly need to get good at?
   29. Danny Posted: November 02, 2005 at 04:56 PM (#1715872)
I can't imagine that a highly-touted 68-year-old would choose Philadelphia over Los Angeles, given the markets, payrolls, climates, etc. otherwise).


The Phillies have a higher payroll than the Dodgers and are in one of the biggest markets in the country.
   30. Jim Furtado Posted: November 02, 2005 at 04:58 PM (#1715875)
JoelW,

I don't think you want a GM with a specialty in any particular area. You want a GM who's comfortable and knowledgeable in both areas. He doesn't have to be the most proficient person in your organization in both areas. He just needs to have enough understanding of all the things that factor into his job performance to properly choose his advisors and properly weight their input.

That's why the Theo situation is so frustrating. He was the perfect guy for the Sox. Not only did he blend the diverse viewpoints, but he also had the people skills to integrate the more extreme viewpoints in a respectful way.
   31. Joel W Posted: November 02, 2005 at 05:03 PM (#1715885)
Jim 1, Joel 0
   32. villageidiom Posted: November 02, 2005 at 05:36 PM (#1715950)
Towers.

He worked with Lucchino prior, so he's not going to blind-sided.


Given what we perceive about Lucchino as being unbearable, why would someone who IS familiar with him be a more likely candidate than someone who ISN'T? I'd think it would be the opposite.

I'm guessing Lucchino fills the position on an "interim" basis, that interim being at least one full year.
   33. Maury Brown Posted: November 02, 2005 at 05:42 PM (#1715970)
Slightly off-topic: When are the GM meetings?

Next week.
   34. tfbg9 Posted: November 02, 2005 at 05:43 PM (#1715974)
WEEI is saying there are talks involving the Sox, LAAoA, and ARZ involving a 3 way deal that would net the Sox prospects, Glaus, and Erstad for Manny. I hate it.
   35. Maury Brown Posted: November 02, 2005 at 05:44 PM (#1715977)
We Mets fans need Duquette to go to Tampa so he can give us Kazmir back.

Hunsciker looks to be going to the DRays
   36. Biff, highly-regarded young guy Posted: November 02, 2005 at 05:45 PM (#1715981)
Why would the Dodgers want Theo but not DePo?

COUNT DA RINGGGGGGGGG
   37. Matt Clement of Alexandria Posted: November 02, 2005 at 05:45 PM (#1715982)
I'm a bit confused here. Who's in charge of the Red Sox and Diamondbacks to make these trades? Moorad and Lucchino? They're going to make huge trades before the new guys take over?
   38. bunyon Posted: November 02, 2005 at 05:49 PM (#1715992)
LL will be the new Sox GM.
   39. Old Matt Posted: November 02, 2005 at 05:49 PM (#1715994)
I think it would be wildly hilarious if the Sox brought Gammo in for an interview.

Just for kicks.
   40. chris p Posted: November 02, 2005 at 05:56 PM (#1716002)
if erstad plays center and glaus plays first i don't mind it.
   41. JB H Posted: November 02, 2005 at 06:03 PM (#1716016)
- Manny for Erstad and Glaus makes no sense, so I'm sure there's nothing there (or at least, a lot missing). There's no point in trading Manny if you're just going to take on worse contracts.

- Theo didn't blend stats/scouting at the major league level. Almost every move he ever made was a slam dunk statgeekwise. The two big moves that weren't were the Varitek and Renteria deals - one is a disaster, and the other has a pretty good chance of becoming a disaster. Please just give me another statgeek.
   42. chris p Posted: November 02, 2005 at 06:12 PM (#1716034)
<i>- Theo didn't blend stats/scouting at the major league level. Almost every move he ever made was a slam dunk statgeekwise.<?i>

The Nomar trade wasn't really a "slam dunk statgeekwise." It was based on the (correct) perception that Nomar couldn't play a passable shortstop. There is no way there was enough data at the time to statistically evaluate Nomar's D.
   43. JB H Posted: November 02, 2005 at 06:23 PM (#1716051)
IIRC, Nomar's UZR was fine in Chicago after being terrible in Boston

The Nomar deal made plenty of sense statistically if they thought he was going to miss a lot of time the rest of the season.
   44. Joel W Posted: November 02, 2005 at 06:33 PM (#1716070)
Manny for Erstad and Glaus makes no sense, so I'm sure there's nothing there (or at least, a lot missing). There's no point in trading Manny if you're just going to take on worse contracts.


Who would the prospects be? Could we get Giles? How good is Erstad in center now? What do we plan on doing in center? If we could get giles, you'd basically be paying the same for Giles + Glaus + Erstad as you would be for Manny + Damon.

I'll wait on proclaiming that to be a bad deal.
   45. JB H Posted: November 02, 2005 at 07:13 PM (#1716194)
It looks like Glaus and Erstad are owed $44 million. Erstad might be worth the $8 million he's owed next year as a CF, but Glaus is an albatross. He doesn't look like he'll be all that much more than an average 1B over his contract, maybe a 120 OPS+, and he's making $12 million a year.

Trading 3 season of a star LF for one pretty good CF season, 3 pretty good 1B seasons and ~$7 million is a bad deal.

I doubt the prospects would be worth much. The Sox don't look to be getting anything from Arizona. I doubt Anaheim would send us Wood or Santana or anything, and I don't think B/B- prospects are all that valuable.
   46. 1k5v3L Posted: November 02, 2005 at 07:45 PM (#1716284)
How is Glaus an albatross? Do you even know the definition of albatross?

The guy has 40 home run, .370 OBP, potential, and you call him an albatross?

And if $12m/year is too much, please help yourself to Paul Konerko.

And, the way I read it, the prospects will be going to AZ actually.

At least the good ones.
   47. tfbg9 Posted: November 02, 2005 at 07:52 PM (#1716298)
Nice press conference. Real revealing.

[end sarcasm]
   48. Jim Furtado Posted: November 02, 2005 at 08:34 PM (#1716371)
Manny for Glaus, Erstad, and a few prospects? Unless the prospects are Ervin Santana and Howie Kendrick, the Sox should just say no.
   49. 1k5v3L Posted: November 02, 2005 at 08:39 PM (#1716384)
Good luck getting that, Jim.
   50. Joel W Posted: November 02, 2005 at 08:53 PM (#1716413)
Sing it with me: Manny is a -25 runs defensive player. He's an outstanding hitter. Since every legitimate metric thinks that's the case, his WARP the past 2 seasons isn't 6.7 but 5.5.

Somebody explain to me how Glaus 3b/1b + Erstad CF + x (maybe Giles) LF + Youks 3B/1B is worse than Manny LF + X CF + Youks 3b/1b + X.

As I see it, we have no centerfielder, no first baseman, a left fielder who can rake, is atrocious in the field, is 34, and doesn't want to play in the city every so often.

At the Red Sox current level of hitting and pitching, 13 runs of offense is worth 10 runs of defense right? So Ertsad, LF X, and Glaus at 1B is getting a bit of a bump on Manny.

Look, I love Manny, he's funny, i like his swagger, I love his bat. I understand the aversion to Erstad (can't hit much) and Glaus (was hurt in 2003-2004, then again, hasn't been hurt outside of those seasons and is 28). But I really don't see how it's necessarily a bad deal minus prospects plus the possibility of getting Giles in LF is really appealing. Frankly, what Giles has done in San Diego is really impressive, and he'd kill the ball in Fenway. And you know, maybe not suck in the outfield.
   51. chris p Posted: November 02, 2005 at 09:04 PM (#1716441)
glaus was hurt pretty much all of this season and still put up 37 hrs and a .363 obp (125 ops+). to me, erstad is a more appealing option than johnny damon for center field--his contract is NOT bad if he plays center field with anywhere near the ability he did a couple years ago. giles? all he did was put up a 148 ops+ (manny was at 156), in other words, a case could be made that giles is a better baseball player than manny.

manny is a dh, not a left fielder. he is playing out of position as a left fielder. -25 runs is really really bad.
   52. karlmagnus Posted: November 02, 2005 at 11:19 PM (#1716730)
If Manny's a DH, let's dump Ortiz, who could get us really good stuff in exchange. The money difference between the two contracts is fungible, and Manny, not Ortiz, is going to the HOF by 2020. Ortiz is a really nice player, but his value is at its ABSOLUTE peak. Plus he's underpaid and will whine about it, so should whine to somebody else.
   53. chris p Posted: November 02, 2005 at 11:36 PM (#1716778)
karl, i pretty much agree, but they cannot afford the pr backlash of trading ortiz. also, i believe manny prefers to play the field.
   54. villageidiom Posted: November 02, 2005 at 11:51 PM (#1716824)
Seriously, is Manny currently costing the Red Sox one run per week on defense, relative to replacement?

Not that BPro is the authoritative source on defensive metrics, but they have Manny on FRAR2 at +3, 0, and +1 for 2003-4-5. Subjectively, Manny has been getting a lot better: his throws more accurate in 2003, his routes more sensible in 2004, and a ridiculously short time from catch to throw in 2005 that didn't lose accuracy. He's still no Gold Glover, far from it.

Matsui drops more catchable fly balls per year than Manny does, takes disastrous routes, and makes horrible throws, yet Matsui's at -9.

One run per week? Other than "yeah, that's what the formula says," does the magnitude make intuitive sense to anyone?
   55. chris p Posted: November 03, 2005 at 12:13 AM (#1716877)
One run per week? Other than "yeah, that's what the formula says," does the magnitude make intuitive sense to anyone?

on the road it's pretty clear he can't get to balls that a decent left fielder makes an easy catch on. he's less of a liability at home.

Other than "yeah, that's what the formula says,"

bottom line is that he makes less plays than other left fielders, park factors considered.
   56. Flynn Posted: November 03, 2005 at 12:25 AM (#1716902)
Manny's 2005 UZR was -47, not -25.
   57. fret Posted: November 03, 2005 at 12:33 AM (#1716920)
Manny's 2005 UZR was -47, not -25.

Where did you read that?
   58. villageidiom Posted: November 03, 2005 at 12:36 AM (#1716924)
bottom line is that he makes less plays than other left fielders, park factors considered.

And certainly he does. But that doesn't defend the magnitude, only the relative ranking. I have no problem that he's ranked last.

The numbers for 2005 suggest that Manny + 2 average OF is about as bad as Matsui x 3. And on several levels that doesn't make sense. Replacement level for LF is pretty bad to begin with, and Manny is worse than that by a run per week?
   59. villageidiom Posted: November 03, 2005 at 12:41 AM (#1716933)
Manny's 2005 UZR was -47, not -25.

Two runs a week worse than replacement in 2005? One run every 3-4 games?
   60. Josh Posted: November 03, 2005 at 12:49 AM (#1716944)
Negative 47? This is too confusing.

How many BIP are hit to the LF? 250-270 a year? This what I figure Manny had given his RF and his F%.

How many non-out BIP to LF would 47 runs equal? Maybe 100?

Manny's F% is about avg, and lets assume his arm is average (b/c it makes my life easier). So, if I'm doing this right, according to UZR Manny didn't get to 100 BIP that an average fielder should have gotten to. That means an average LF would have gotten to about 350 BIP?

I must be missing something. That just seems really really wrong. I'm sure I'm missing something.
   61. Joel W Posted: November 03, 2005 at 01:21 AM (#1716986)
Why is it so difficult to believe that Manny is that bad at fielding? He's not fast, he misjudges balls, and the only thing he has going for him is that he plays the ball well off the wall.

The UZR thing: remember that is compared to average, not replacement. I think 25 runs below replacement is a relatively solid estimate of his defensive ability.

Why would we trust our intuition with defensive valuations?

.85 runs per play was the value just used in primate studies. So I don't know why the idea that manny doesn't get to 35 balls that a replacement level left-fielder gets to is so outrageous.
   62. Josh Posted: November 03, 2005 at 01:29 AM (#1717000)
85 runs per play was the value just used in primate studies. So I don't know why the idea that manny doesn't get to 35 balls that a replacement level left-fielder gets to is so outrageous.

35 BIP I'd agree with.

55-60 (as a -47 UZR indicates) seems a bit high, but a lot less hysterical than the 100 I posted above (using numbers from OOMA).
   63. Joel W Posted: November 03, 2005 at 01:50 AM (#1717020)
-47 is one season of UZR which we'd have to regress, and really that would be about -35 above replacement, regress that and look at some older seasons, realize that Manny is 33/34 and we get: a really bad fielder who isn't getting better...could Manny play 1st? If not, why?
   64. villageidiom Posted: November 03, 2005 at 02:05 AM (#1717038)
That's what I mean, Josh. The more I roll around these numbers in my head and try to translate them into something more tangible, the less I end up with something that makes sense.

ESPN.com says that Manny had 267 total chances last year in LF, so, yeah, your math is right.

If you take Total Chances (PO+A+E) and divide by ZR, that should give a sense of the number of Total Opportunities in Zone (TOiZ). Subtract non-hits (PO+E) and that should roughly be the number of Hits in Zone (HiZ). This is what I get:

<table width=50% cellpadding="3" cellspacing="1"><tr align="right"><td align="left">Name</td><td>TOiZ</td><td>HiZ</td><td>OBAiZ</td></tr><tr align="right"><td align="left">Johnson</td><td>152</td><td>17</td><td>.112</td></tr><tr align="right"><td align="left">Catalanotto</td><td>195</td><td>32</td><td>.165</td></tr><tr align="right"><td align="left">Podsednik</td><td>296</td><td>33</td><td>.112</td></tr><tr align="right"><td align="left">Crisp</td><td>332</td><td>34</td><td>.103</td></tr><tr align="right"><td align="left">Holiday</td><td>279</td><td>36</td><td>.128</td></tr><tr align="right"><td align="left">Mench</td><td>269</td><td>37</td><td>.139</td></tr><tr align="right"><td align="left">Klesko</td><td>249</td><td>41</td><td>.164</td></tr><tr align="right"><td align="left">Crawford</td><td>384</td><td>41</td><td>.106</td></tr><tr align="right"><td align="left">Gonzalez</td><td>315</td><td>42</td><td>.132</td></tr><tr align="right"><td align="left">Burrell</td><td>287</td><td>44</td><td>.153</td></tr><tr align="right"><td align="left">Bay</td><td>311</td><td>45</td><td>.146</td></tr><tr align="right"><td align="left">Dunn</td><td>299</td><td>48</td><td>.161</td></tr><tr align="right"><td align="left">Floyd</td><td>335</td><td>50</td><td>.150</td></tr><tr align="right"><td align="left">Cabrera</td><td>248</td><td>55</td><td>.222</td></tr><tr align="right"><td align="left">Matsui</td><td>279</td><td>57</td><td>.204</td></tr><tr align="right"><td align="left">Stewart</td><td>311</td><td>58</td><td>.187</td></tr><tr align="right"><td align="left">Lee</td><td>376</td><td>63</td><td>.168</td></tr><tr align="right"><td align="left">Ramirez</td><td>366</td><td>116</td><td>.317</td></tr></table>
This suggests that Manny allowed at least twice as many hits in his zone than any other LF (save Lee). Said another way, for every hit that falls in LF with a replacement-level player, Manny allows at least two.

The last column is HiZ/TOiZ, which is roughly the opponents' batting average when hitting into the LF zone. Manny is at .317; said another way, the typical batter performs like Vlad Guerrero when hitting toward Manny.

If we are to believe ZR, the above must be true. But intuitively it's nowhere near being true.
   65. karlmagnus Posted: November 03, 2005 at 02:06 AM (#1717041)
Why would trading Ortiz annoy the fans more than trading Manny? If Manny's defense is such a problem, he should DH. If Ortiz' defense is so cosmically bad that Manny can't DH, then the -47 runs that the idiot formula says Manny's costing should be billed to Ortiz's account not Manny's.

Either way, Ortiz is the player that will light up the switchboards with attractive deals if we trade him, and Manny is the player who is likely to provide more consistent value to the Sox (even though he's older) as well as being a fan favorite and a near-certain HOF/retire the shirt number candidate. The difference in salary is something the Sox can absorb more easily than all but 1 of its competitors.
   66. karlmagnus Posted: November 03, 2005 at 02:08 AM (#1717044)
For one thing, there are only 3-4 teams that can afford Manny (one being the MFY) whereas every team in baseball, loaded with juicy, succulent young propsects, can afford Ortiz.
   67. villageidiom Posted: November 03, 2005 at 02:15 AM (#1717050)
Well, that looked a lot better in livepreview. Let's see if this is better:


Name       TOiZ    HiZ     OBAiZ
Johnson     152     17      .112
Catalanotto 195     32      .165
Podsednik   296     33      .112
Crisp       332     34      .103
Holiday     279     36      .128
Mench       269     37      .139
Klesko      249     41      .164
Crawford    384     41      .106
Gonzalez    315     42      .132
Burrell     287     44      .153
Bay         311     45      .146
Dunn        299     48      .161
Floyd       335     50      .150
Cabrera     248     55      .222
Matsui      279     57      .204
Stewart     311     58      .187
Lee         376     63      .168
Ramirez     366    116      .317
   68. chris p Posted: November 03, 2005 at 02:19 AM (#1717056)
If we are to believe ZR, the above must be true. But intuitively it's nowhere near being true.

those numbers are worthless without context. how does manny do on road? how do opposing left fielderes do at fenway? how many of htose hits in zone are impoosible to catch (off the wall) ...
   69. villageidiom Posted: November 03, 2005 at 02:19 AM (#1717057)
Yeah, that worked better. Again, with the conclusions (assuming ZR is correct):

1. For every hit that falls in LF with a replacement-level player, Manny allows at least two.

2. The typical batter performs like Vlad Guerrero when hitting toward Manny.

Just from casual observation, that seems very unlikely.
   70. villageidiom Posted: November 03, 2005 at 02:26 AM (#1717071)
those numbers are worthless without context. how does manny do on road? how do opposing left fielderes do at fenway? how many of htose hits in zone are impoosible to catch (off the wall) ...

In terms of home vs. road, I don't know. But if {opposing hitters = Vlad} overall, and if your suggestion is true that that Manny is worse on the road, it would mean that {opposing hitters = Ted Williams} on the road. And that's even more ludicrous.

Manny is observably bad, but not this bad.
   71. Matt Clement of Alexandria Posted: November 03, 2005 at 02:30 AM (#1717078)
km -

I don't get it. You've made a big deal about keeping players that the fans love, that maintain a bond with the team.

I'm not going to argue that Papi or Manny is more or less the face of the team. The point is, they're both huge. Letting either of them go would seem to violate the rule you set in place - certainly, it'd be far worse to trade Papi than to let Damon go, which you've already argued for as a move to retain the same core.
   72. chris p Posted: November 03, 2005 at 02:30 AM (#1717079)
vi, my point is that just plain zr cannot be used to evaluate a fenway park left fielder. are the zones defined differently in different parks? if they aren't, then the green monster is right in the middle of the left fielder's zone and manny positoins himself so shallow he's practically out of the left field zone. or am i completely wrong in how zr works?
   73. Spivey Posted: November 03, 2005 at 02:32 AM (#1717082)
ESPN.com says that Manny had 267 total chances last year in LF, so, yeah, your math is right.

Well, not really. Those are the balls he can get to. The question is how many he can't get to. You could easily be adding 30-50 to that number, pushing it over 300.

Oh, and as if I needed more of a reason to completely disregard BPro's fielding metrics, Manny being listed as average or anything near it is enough. ZR, DSG, and Anaheim Rallymonkey all have Manny as ~-25 in the field. If you consider him to be a -25 fielder (which I think is definitely fair), we're talking about a player who is worse than Carl Crawford, without even considering contract.
   74. Spivey Posted: November 03, 2005 at 02:35 AM (#1717085)
UZR and DSG's defensive metric I believe both include park factors for Fenway's wall, FWIW.
   75. chris p Posted: November 03, 2005 at 02:35 AM (#1717086)
the manny vs. ortiz discussion is just an exercise b/c we know there is a 0% chance they trade ortiz. however, i will say that i disagree with km's analysis here. as i've already said, the media backlash would be huge.another point, not every player *likes* to dh. we KNOW that ortiz doesn't mind and his production is at an elite level when he is the dh. manny on the other hand is unhappy to begin with, and he doesn't like to dh. i don't think moving an unhappy player to a position where he would be less happy is a recipe for success.
   76. Darren Posted: November 03, 2005 at 02:56 AM (#1717111)
If Manny's a DH, let's dump Ortiz, who could get us really good stuff in exchange. The money difference between the two contracts is fungible...

Let's see: Ortiz is 4 years younger than Manny, makes ~$10 mil less per year and is on a shorter contract. If $10 mil is fungible to you, please share the wealth a bit.

JoelW--

You are a voice of reason here. People love to say things like Manny is a star LF, or his bat is irreplaceable and ignore the fact that the best defensive metrics we have say his defense drags him down to merely good. If you mention this, you're often met with answers like "he's not that bad" or "he can't be 1 run worse than average per week." This is quite similar to Yankees' fans defense of Jeter. When UZR hated Jeter, it was a bogus stat. When it said he was average in 04, it proved he was improving. And now that it's back in the crapper, it's back to being an inferior stat.

With Manny, you don't even have the excuse that he looks good in the field. He lopes around slowly misjudging balls. If I had to guess what kind of fielder he was, it would be very close to the ratings that UZR gives him. But fortunately, I don't have to guess. I have a very good stat called UZR, which has been shown to correlate well to run prevention and to itself on a Y-T-Y basis.

Manny is far from untouchable. Trading him is a great idea if you can get someone else to see him as a star. Glaus and Erstad are not what I'd want for him, but who knows what those "prospects" are.
   77. IronChef Chris Wok Posted: November 03, 2005 at 04:53 AM (#1717188)
Can't we just stick ORtiz at 1B and end the debate?
   78. chris p Posted: November 03, 2005 at 05:07 AM (#1717191)
Can't we just stick ORtiz at 1B and end the debate?

uh no.
   79. villageidiom Posted: November 03, 2005 at 05:50 AM (#1717223)
vi, my point is that just plain zr cannot be used to evaluate a fenway park left fielder.

I guess that's my point, too. But Rallymonkey's oft-quoted -25 calculation starts with ZR, and the improvements on ZR he makes do not overcome the bad starting point.

I don't have to guess. I have a very good stat called UZR, which has been shown to correlate well to run prevention and to itself on a Y-T-Y basis.

Based on the underlying assumptions, are there specific cases where UZR shouldn't work well, even though it will work well in general?

IIRC, UZR goes like this:

- Start with outs & hits in subzones, and convert to runs based on average runs per hit for the subzone and the player's out rate & hit rate relative to league average for that subzone.
- Apply park adjustment that is calculated based on the LF zone (not the subzone) data for 10 years, looking only at FB out % and LD out % (treating them the same) over that span.
- Apply adjustments for batted ball speed and LHB/RHB, based on typical league out rates for the subzones.
- Apply adjustment for pitcher GB/FB tendencies, based on typical league out rates.

In there, you've got assumptions that:

- batted balls are assigned to subzones properly
- subzones are irrelevant to park factors
- park effects are irrelevant to batted-ball speed, LHB/RHB, and GB/FB adjustments

ZR and UZR both fall victim to the first assumption. We have no idea how big an issue this is in general, especially concerning the "shared" zones. But exactly how small the subzones are at Fenway, or how much of the LF zone is "deep", is even more problematic.

I suspect that, with any zone assignment, if a player fields a ball in an adjoining subzone that "belongs" to the neighboring fielder, the ball gets misassigned. You certainly see it with Carl Crawford, who from LF can get to balls in the CF zone, or at least a highly disproportionate number of balls in the LCF "shared" zone. If he catches them, he gets credit; if nobody does, he AND the CF share the blame. But the end result is that the balls officially marked in "his" zone are disproportionate to the typical distribution.

I think Manny has a similar problem, believe it or not. Not that he catches a lot of LCF balls, but rather that balls hit off the wall in LCF are being "coded" as being hit to LF. Generally, Damon defers to Manny on these wall-ball hits, as Damon's arm is notoriously weak. As Manny fields it, I think some of these are slotted to Manny's "territory" instead of the shared subzone. And since they're all hits, Manny gets the opposite problem of Crawford: a disproportionate share of blame for hits.

(Crawford and Manny are 1-2 in 2005 in terms of total opportunities per inning in the LF zone. Matsui - who with Bernie Williams faces a similar problem to Manny/Damon described above - is 3rd.)

The second assumption is generally irrelevant, but is certainly relevant in Fenway. Intuitively the out rate for deep balls in Fenway is very low in the LF domain, while it rises considerably for the shared LF/CF domain (which appears to be assigned entirely to LF for the park adjustment, even though it's shared otherwise). And of course this will vary differently for FB% and LD%, which the park adjustments ignore. (Park factors adjust for the combined effect, which is certainly a step in the right direction.) If I'm looking at this correctly, the sizable park adjustment for Fenway LF is actually understating the effect, and thus underestimating the number of hits.

As an aside, LD% appears to be ignored for outfielders in UZR, other than being thrown in with the park adjustment. If your pitching staff has a high LD% - the best sign of a bad staff - you should expect a higher BABIP, right? And from that you'll get fewer outs as % of hits, in pretty much every zone; and from that you'll get bad UZR; and from that you'll mistakenly conclude that some of the "true" pitching effect is defense. I'm not sure if this is relevant to Fenway LF, but I thought I'd bring it up while it was on my mind.

And I think the case can easily be made that UZR falls short on the third assumption, as it pertains to Fenway. Does an easy fly ball to deep LF mean the same thing at Fenway as in other parks? Ask Mike Torrez. OK, small sample, but you get the idea: something that would be considered a relatively sure out anywhere else is, in Fenway LF, anything but. Since the adjustments don't reflect park, they will overestimate the number of outs a Fenway LF should produce.

UZR, as good as it is, relies on a set of assumptions that will work against the Fenway LF, regardless of who it is. This will screw up the absolute number of fielding runs, and the rate, produced through UZR for Manny. I know everyone would like for convenience to assume that the effect is minimal, but that is not wise.

I'm pretty sure that Manny will only see a Gold Glove if they start selling them at F.A.O. Schwarz, so don't take my critique of UZR as suggesting that Manny's not bad defensively. He is bad, and noticeably so; and the relative ranking of Manny as the worst defensive LF is subjectively not far from the truth. But when trying to determine Manny's value (in RARP, WARP, whatever) keep in mind that the quantitative defensive measures overstate his problems by an unknown amount.
   80. Matt Clement of Alexandria Posted: November 03, 2005 at 12:46 PM (#1717367)
For one, UZR starts by accounting for the type of BIP. So the LD% thing does not really apply there.

For another, it's park adjusted. I don't know the specifics of the adjustment, but it seems like you're assuming there is none, and that's not correct.

Finally, non-Manny players have had perfectly fine UZR ratings in Fenway's LF. In 2000, Troy O'Leary was +11 in 119 games. In 2001, he was +4 in 43 games, and Dante Bichette was +3 in 30 games. In 2002, Rickey Henderson was at 0 in 38 games. Manny's low numbers are not typical of Fenway LF.

It is true that he plays very shallow in Fenway. The distribution of BIP might be affecting Manny's UZR. But if that's the case, the issue is positioning, not park adjustments. That's where I'd look if I wanted to say Manny's not as bad as his UZR/ZR.
   81. OlePerfesser Posted: November 03, 2005 at 04:42 PM (#1717535)
Good post, village--thanks.

Stuff like that is important; it's a kind of "referee report" of the type that's routine in academic research. I think we need more of that kind of thing in stathead circles. Too often, new statistical measures (esp. "proprietary" ones, the internal workings of which are veiled in secrecy) make it out into general usage without much peer review/criticism, and with few robustness tests.

I'm not singling out MGL's UZR here. I'd say the best example of this phenomenon is Bill James's Win Shares, usage of which has become quite common, but methodological comprehension of which is extremely scarce.
   82. villageidiom Posted: November 03, 2005 at 06:03 PM (#1717655)
For another, it's park adjusted. I don't know the specifics of the adjustment, but it seems like you're assuming there is none, and that's not correct.

Not often I say this - and even less often is it directed toward you by anyone - but... RTFP.

Too often, new statistical measures (esp. "proprietary" ones, the internal workings of which are veiled in secrecy) make it out into general usage without much peer review/criticism, and with few robustness tests.

When statistical measures that generally work well break down at the margins, the general reaction in the community (such as it is) is, "Well, the method works in so many cases that it can't possibly be wrong for this one." And that's ignorance by choice. If hammers work well for building a new house in general, we can use it to install the light bulbs, right?

We all make simplifying assumptions. Each person makes simplifying assumptions that serve their interests. Nothing wrong with that. But the more your assumptions rely on someone else's assumptions, the more likely you'll be wrong. When they (Stanley? Or Stanley?) built the hammer, they knew it would work for most fastening jobs in building a house, because most of that work involved wood and nails. If you assume it's good for all home-building work, you'll find yourself complaining about the quality of light bulbs.

I've seen it with defensive metrics, I've seen it with DIPS, I've seen it with Pythagorean Wins. I haven't seen it with Win Shares, but I'm assuming that's because I simply haven't looked at it closely. But for pretty much every statistical measure, someone will assume it applies to a case where it doesn't work very well. They're taking the pills without reading the warnings.

Again, Manny is bad by nearly any measure, and his ranking from any defensive metric agrees. The absolute levels, though, appear to be nutty. If you're trying to figure out precisely what total value he brings, you can't get there with the existing defensive metrics with any level of certainty because they overstate his ineptitude.

To offer something constructive... If this is a problem with defensive metrics at the margins, then you could do pretty well by subtracting 2 from the 10th percentile UZR for LF, and using that as an estimate of Manny's defensive value. If you think the 10th percentile isn't high enough to guard against goofy numbers at the fringes, start higher, and subtract more. In terms of whether it's worth it to trade him or not, or evaluating the talent you get in return, this should give as much of a quality answer as the metrics can allow.
   83. Joel W Posted: November 03, 2005 at 09:40 PM (#1718044)
VI,

I note that you didn't talk about MCA's post where he showed that O'leary and even Bichette (Bichette!) had positive UZRs in Fenway. So, his repsonse to your previous post still has the merit in that it objects to your characterization of Fenway as inherently a problem for left-fielders that UZR deals with incorrectly.

Your general claim is true of this community. However, Mike Cameron and Ichiro both managed to have good numbers in Seattle together (Crawford/Baldelli aren't in the old UZR ratings, so I found a different metric).

I think it's possible that there's another reason Manny has more opportunities than other LF: People try to hit balls that way at Fenway. Whether it's managers putting in hitters who tend to hit balls to left field, or hitters trying to go that direction at Fenway, there's certainly reason to think that LF in Fenway sees more balls because it's LF in Fenway.

You may have a valid point on the Manny/Damon thing, but I'd still like to hear from MGL on how he deals with such issues. Also, Manny had 330 BIZ last year, Damon 450. For reference, Beltran had 424 and Floyd had 315. Brady Clark 446 and Carlos Lee 355. So the range isn't that off for Manny. Given that MGL does use park factors for hit rates, it would need to be the balls in zone that was the problem.

Regardless, I don't see why it's so nutty to assign Manny such awful numbers. The difference between the best 3b at hitting last year was 100 runs. In Centerfield it was about 70 (for players w/ over 300 PAs). At 1st about 100, at 2B about 70. So let's just say about 85 runs is the difference in a given season between the best and the worst players at a position over the course of a season hitting-wise.

In the UZR numbers on Tango's site, the biggest difference is -73. That's the difference between Erstad and Cruz Jr. in Center. At shorstop, it's 50. Left-field it's 64. At 1b it's 40. So, I am wondering why I am supposed to believe that UZR is flawed on the margins, even though the range of values is perfectly reasonable. (The average number of plays in zone is like 300, so we'd expect the ranges to be smaller, which they are)

Moreover, we have good reason to think that Manny is in the outfield despite god-awful D: he's an amazing hitter, it's worth it. And it is for the most part, but I'm not going to reject generally sound methods it seems bad in certain spots. Talk to Barry Bonds and Pedro Feliz about how the difference just seems off in their abilities.
   84. veer bender Posted: November 04, 2005 at 07:00 PM (#1719127)
It's very possible I'm missing something here, but I think the fact that the range of numbers for defense and hitting are of similar magnitude supports the position that there's a problem here, rather than refuting it. I mean, hitting is the greater component of run scoring, while defense the lesser component of run prevention. Given the relatively small impact of defense compared to hitting, doesn't a -50 UZR suggest a true baseball skill level of horribility much greater in magnitude than an offensive stat (say VORP) of +50 is indicative of true talent at hitting the ball?
   85. Joel W Posted: November 04, 2005 at 07:34 PM (#1719189)
Well, actually the point was that the range is about 65-70% of what it is for hitters. So the range is much smaller. And frankly, I think we have decent reason to think that range in fielding talent ought to be higher than the range for hitting talent. This whole debate is proof of this point.

Measuring a hitter's quality is relatively easy. OPS, 1.8 OPS, RC, VORP etc. etc. They pretty much tell us the same thing. Even bad front offices pretty much know who is good and who isn't. But with defense, there is so much disparity in evaluations of a player's quality. Eric Van once showed that there is no corellation with defense as measured by UZR and salary. (I know that this is self-referential, but I think we established that within the bounds, UZR is solid).

Regardless, teams not able to measure defense are going to focus on what they can measure, and that's hitting ability. The range between Adam Kennedy and Luis Rivas is a great example of this: everybody thought that Rivas was a good fielder, UZR disagrees, but that perception that he was a good fielder means that bad fielders get more playing time than they should. Jeter and Jose Valentin are another good example. But the Jeter point would remain even if he was recognized as a bad fielder. Since people don't measure defense as well as they measure offense, hitting would rationally be a more important piece of the measurement.

We know Manny can rake, but people argue about how bad his defense really is, whether we aren't capturing something correctly, but we can defintiely say, "yeah, but he can f'ing rake, keep him." Well, that's the reason that I think we can have these outliers. Put it this way, if every GM became really stat savvy, who would go first? The players with good defensive reputations and bad defense. You don't have a reputation on offense divorced from your numbers (for the most part, the Podsednik's of the world are around, but it happens more on D).
   86. Josh Posted: November 04, 2005 at 07:39 PM (#1719199)
vi -- great stuff.

In all events, to get back to the thread's topic, what do people think of the recent cadidates mentioned? I'm particularly interested in Kim Ng -- a smart person, with a background at a top notch economics/analytics school, almost a half dozen years older than Theo when he first stepped up, and having worked in three different big markets. Anyone know anything more than her BPro interview?
   87. veer bender Posted: November 04, 2005 at 07:53 PM (#1719223)
I agree with a lot of what you say, and I think my point is not at all clear, at least not yet. The key to my point is the idea that in general, hitting is a bigger part of scoring runs than defense is in preventing them. Let's assume that (1) scoring and preventing runs are equally importnant; (2) offensive stats that use runs as the unit describe either a player's hitting or their total offensive game, so they encompass 75-100% of run scoring; (3) UZR, measured in runs, describes 10-50% of run prevention (I don't remember the best guess, and no one knows for sure).

So the extreme UZR outliers seem too large in magnitude, because of the missing, larger component of run prevention -- pitching.

Here's another totally different thought. You're right that most are better at judging offensive ability than defense, and that may lead to more variability in true defensive ability at the MLB level. But. . . .remember they have different defensive positions in baseball, with different demands, and varying difficulty and importance. This should flatten out some of that effect. Obviously lineup order affects a hitters contibution slightly, and there are different batting styles, but I think this is relatively less important. If a guy can't hit at all, he just plain sucks. If he can't play shortstop at all, he could still be an excellent player.
   88. villageidiom Posted: November 04, 2005 at 11:54 PM (#1719523)
I note that you didn't talk about MCA's post where he showed that O'leary and even Bichette (Bichette!) had positive UZRs in Fenway. So, his repsonse to your previous post still has the merit in that it objects to your characterization of Fenway as inherently a problem for left-fielders that UZR deals with incorrectly.

I don't think I ever said that you can't have positive UZR in Fenway LF. I said that with marginal situations UZR breaks down. Manny's D is marginal. Fenway LF, with respect to the assumptions underlying the defensive metrics, is marginal. Put the two together and you get something you can't use.

Since his point didn't address mine, I didn't address his.

Given that MGL does use park factors for hit rates, it would need to be the balls in zone that was the problem.

MGL uses park factors based on the zones, not the subzones. And they attribute the "shared" LCF entirely to LF. When you have such ####ed up (read: inconsistent) subzones as Fenway does, there could be an effect.

So, I am wondering why I am supposed to believe that UZR is flawed on the margins, even though the range of values is perfectly reasonable.

All you've done is to suggest - not even prove - that UZR is no more flawed at the margins than other defensive metrics. I agree.
   89. Joel W Posted: November 05, 2005 at 12:09 AM (#1719536)
I think I was suggesting that defensive metrics are no more flawed on the margins than offensive statistics. There is certainly more variability, and they tell us less, but I was trying to say that given those statements, I don't think the variability has anything to do with flaws at the margins.
   90. Darren Posted: November 05, 2005 at 03:01 AM (#1719644)
MGL uses park factors based on the zones, not the subzones. And they attribute the "shared" LCF entirely to LF. When you have such ####ed up (read: inconsistent) subzones as Fenway does, there could be an effect.

There could be, but that hasn't appeared to have hindered any other LF. Why assume that it is hurting Manny?

When statistical measures that generally work well break down at the margins, the general reaction in the community (such as it is) is, "Well, the method works in so many cases that it can't possibly be wrong for this one."

I think you've totally mischaracterized what the general reaction is. I think it's more along the lines of "This method has been tested and shown to work. If you're going to say that player J is an exception, you need do more than say 'he's really good at pop-ups,' or 'maybe the subzones are screwing him in ways that they screw no other player.'


Too often, new statistical measures (esp. "proprietary" ones, the internal workings of which are veiled in secrecy) make it out into general usage without much peer review/criticism, and with few robustness tests.

I know you're not singling out UZR here, but you must certainly think it falls into this category if you're making this comment. Do you really think that that stat has not gone through considerable peer review and testing? To me, peer review and testing are the reasons that I trust it over things like BPro's Defensive Magic.
   91. Darren Posted: November 05, 2005 at 03:16 AM (#1719652)
To sum up, vi, your entire argument seems to be that UZR may not fully reflect Manny's defense. I don't think I can argue with that. UZR is far from perfect and certainly not as reliable as hitting stats.

But it seems to me that in evaluating players, you need decide their value (duh). In Manny's case, there's little argument about his offense (it's very good). There also seems to be very little argument about his D. UZR says he is terrible. DSG's number says he's terrible. BPro says he's terrible. Scouts say he's terrible. Tango's Fan's scouting report says he's terrible. Why are you trying so hard to find a flaw in UZR, when it appears that UZR pretty much agrees with everything else about Manny?
   92. Russ Posted: November 05, 2005 at 09:37 PM (#1720081)
Why are you trying so hard to find a flaw in UZR, when it appears that UZR pretty much agrees with everything else about Manny?


Everyone agrees he's a bad fielder. However, the degree to which he is bad is very important in Manny's case. Because of that double exponential upper tail of baseball players, there's a big increase in quantity between the truly elite (e.g. ARod) and someone who is good, but not in the truly elite (Ortiz). Because there are more guys like Ortiz than there are like ARod, Ortiz should (and does) cost a lot less per year than ARod. Manny's possibly enormous defensive deficiences may make him more in line with Brian Giles than Alex Rodriguez, even though the line between ARod and Giles doesn't look that great in an absolute value sense, it is in a relative scarcity sense.
   93. Darren Posted: November 05, 2005 at 10:29 PM (#1720137)
Manny's possibly enormous defensive deficiences may make him more in line with Brian Giles than Alex Rodriguez,

As merely a bad fielder, Manny is in line with Giles.


Everyone agrees he's a bad fielder. However, the degree to which he is bad is very important in Manny's case.

Okay, but what are we talking about here? DSG has him at -15 (with PF), BPro it's something like -13. I think that UZR puts him in the -20s somewhere (unless the -47 is accurate). So it's maybe 10 runs difference? Is that really worth disecting UZR over?

Let me address a couple specifics of VI's posts:

I think Manny has a similar problem, believe it or not. Not that he catches a lot of LCF balls, but rather that balls hit off the wall in LCF are being "coded" as being hit to LF. Generally, Damon defers to Manny on these wall-ball hits, as Damon's arm is notoriously weak.

Correct me if I'm wrong, but the universe of balls that hit off the wall, for and against the Red Sox, would all be hits. So Manny would not be penalized, because the expected result of such balls would be a 1.000 batting average, which Manny would match. (Correct me if I'm misunderstanding UZR here.)

Damon's deferrel would also be irrelevant except that it would give Manny extra chances to throw out runners. (FWIW, I had not noticed Damon deferring to Manny any more than any other CF does to any other LF.)
   94. villageidiom Posted: November 05, 2005 at 11:10 PM (#1720168)
UZR says he is terrible. DSG's number says he's terrible. BPro says he's terrible. Scouts say he's terrible. Tango's Fan's scouting report says he's terrible. Why are you trying so hard to find a flaw in UZR, when it appears that UZR pretty much agrees with everything else about Manny?

I also say he's terrible. And on that, I agree with the defensive metrics.

The reasons I'm trying to point out that the existing defensive metrics are flawed at the margins - not "find a flaw in UZR" as you mischaracterize it - are because they were being used with some level of precision ("...his WARP the past 2 seasons isn't 6.7 but 5.5...") to demonstrate Manny's overall value for the purposes of making a trade. At the margins they don't work with that level of precision. If Manny is a -11, or -15, instead of a -25, he's still the worst LF around; but he might be a 6.1, or 5.9, instead of a 5.5. And the quality of the deal you make depends heavily on that answer.

If all you want to do is have a sense of who's good or bad, or get a relative ranking of the players, each of the defensive metrics out there can handle it. At the margins precision breaks down.
   95. villageidiom Posted: November 05, 2005 at 11:22 PM (#1720176)
Correct me if I'm wrong, but the universe of balls that hit off the wall, for and against the Red Sox, would all be hits. So Manny would not be penalized, because the expected result of such balls would be a 1.000 batting average, which Manny would match. (Correct me if I'm misunderstanding UZR here.)

UZR treats anything in the field of play as catchable, but it penalizes only if a fielder catches fewer than is typically caught in a particular subzone. If there were no park adjustment that would be a huge penalty for Fenway LF, because a much higher proportion of deep-zone balls would not be caught. Fortunately there is a park adjustment (a substantial one), but it treats the entire LF and LCF area as one big zone for that purpose.

If all the wall-balls in LCF are allocated to Manny he's penalized, unless the park factor adjust for that. And it won't. It'll adjust for the fact that more BIP hit the wall than in other parks; but it won't adjust for the fact that one particular OF gets all the shared-subzone hits allocated to him.
   96. Darren Posted: November 06, 2005 at 05:19 AM (#1720438)
It'll adjust for the fact that more BIP hit the wall than in other parks; but it won't adjust for the fact that one particular OF gets all the shared-subzone hits allocated to him.

How can you say that without a precise map of the zones? Isn't it possible that one of the zones encompasses all the wallballs? That would mean that, though Manny might be responsible for all of them, he wouldn't be penalized because he'd be doing just as well as other LF.

(I still don't see any evidence that Damon defers more than any other CF.) In the abscence of proof on these, I tend to believe that Manny's getting a fair shake.
   97. IronChef Chris Wok Posted: November 06, 2005 at 12:59 PM (#1720561)
Damon's deferrel would also be irrelevant except that it would give Manny extra chances to throw out runners. (FWIW, I had not noticed Damon deferring to Manny any more than any other CF does to any other LF.)

I've only seen Manny cut off Damon's throws
   98. Шĥy Posted: November 08, 2005 at 04:03 AM (#1722811)
To bring this thread back to its topic...

Red Sox GM Candidates

The Red Sox have announced their first two interviews in the GM search. Jim Beattie, the former Expo GM who was part of the two-headed disaster in Baltimore, and Jim Bowden, the Weston native who thrilled Washington by keeping the Nationals in the NL East race for much of the season, will interview for the job.

The Sox have contacted two other prospective candidates but their names have yet to be announced.

You must be Registered and Logged In to post comments.

 

 

<< Back to main

BBTF Partner

Support BBTF

donate

Thanks to
robneyer
for his generous support.

Bookmarks

You must be logged in to view your Bookmarks.

Syndicate

Page rendered in 0.6097 seconds
41 querie(s) executed