The Manny Question… Again
Welcome to the new fall season of “Trading Manny,” (Manny Swap? Super-Manny?) the hit reality show that’s now entering its fourth season!
Once again, when it comes to considering trading Manny, it’s hard to judge any deal unless you can pin down Manny’s defensive value. To that end…
As some of you have probably already noticed, there is some interesting stuff going on over at Dialed In, where Chris has linked to a Rally Monkey’s list of park factors for ZR. There, Rallymonkey finds that Fenway LF deflates a player’s defense by 14 runs/400 chances. He says that Manny came out at -25 per 400 chances, so his new rating should be more like -11 runs/400 chances. I’m quite shakey on whether his choice to not regress was a good one, and where he came up with the 400 chances # or the -25 #. Manny was -34 by Rally’s method in 05, so I’m assuming this adjustment would make it ~-20. He actually says that Manny came out at -31 in 248 chances. Using the computation below, we get -22 total and -35/400 chances.
Things get tricky if you try to apply this adjustment to Dial’s #s. If you look at Dial’s methodology, you’ll see he concludes that leftfielders average 348 chances per 162 games. Therefore, if we take Manny’s 1031.3 innings played this year, you’d expect 246 chances. This would tell us how to scale down that /400 chances number above (divide it by 1.63), which makes it approximately -9 runs over that span. That would make Manny’s adjusted Rspt in 06 -23, and his RS/150 -32. In 05, Manny was -34 RS/150, so the adjusted number would be ~-23 in 05.
One more set of numbers to consider. If you recall, when I tried to figure out Manny’s defensive value after 05, MGL provided me with these UZR numbers:
2003, -20, 1073; 2004, -20; 2005, -50. Those numbers are /150 games, and are park adjusted. However, a few months later MGL wrote this article, in which he said UZR showed Manny’s “total defensive contribution over the last 3 years was -52 runs, or -20 per year (150 games). I don’t know why those two are so different (it’s not just arm, which is only 6 runs of it). Let’s just assume that MGL’s article has the real UZR number, whatever that means these days.
So, of the park-adjusted PBP defensive models that we know about, here’s what I have:
UZR: -20/150g avg from 2003-05
Rally: -20/400 chances in 05, -35/400 chances in 06
Dial ZR: -23 /150g in 05, -32/150 in 06
(Other #s: DSG had him at -35/150 in 05 but that’s not PBP data, and fielding bible had him at -20/150 games, but it’s not clear if that’s park adjusted. Joe Arthur’s figuring of 2006 Fenway home/away splits comes out to something around -9.)
I could be totally mangling these numbers (I’m not very bright, I’ve heard). But to me, that’s a pretty strong consensus that Manny’s a very poor OF. Something like -15 to -20 should be expected going forward (and I’d say that’s conservative). How does that affect whether you’d be willing to trade him or what you’d accept back for him?
Darren
Posted: October 31, 2006 at 09:19 PM |
67 comment(s)
Related News:
Boston
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.
You have 25 six-sided dice, 4 ten-sided dice, and one 20-sided die. The average roll is 3.5 for the first 25, 5.5 for the next four, and 10.5 for the last one. Do you regress the average for the 20-sided die to 5.5 or 3.5? No. You KNOW it's not like the others.
The Monster is an outlier, but it's a legitimate outlier. Regression eliminates the noise, but it also eliminates the legitimate variation. That wouldn't be an issue if the former outweighed the latter, but it doesn't.
If it's to be regressed at all, perhaps it should be regressed to an N-year average of itself. I guess that depends on how you feel about using multi-year data to develop single-year park factors. I'm not that comfy with this notion, either. (For projection, fine; for retrospective work, no.)
I just got an answer from Rally. Those 06 numbers were just an example, not Manny's actual numbers. I'll fix it when I get the actual #s.
How does that affect whether you’d be willing to trade him or what you’d accept back for him?
I work the other way. If I'm the GM I want to know what everyone would be willing to pay, then decide whether it's worth trading him. If an unsavvy front office makes an offer that clearly ignores the -15 to -20 on defense, it might be worth making the deal. If a savvy front office makes such an offer - or, better yet, if many teams do - I'd start to wonder about the legitimacy of that -15 to -20 range. Given the uncertainty that still remains about defensive worth and the effects of the Monster on those estimates, I don't think I could find an offer I'd feel comfortable with.
And given that, plus Manny's 10/5 no-trade status and the many other holes the team must fill, if I'm the GM I'm not going to spend much time on this issue. I've got bigger fish to fry.
I have estimated a zone rating park factor before based on UZR park factors, and got something like 18 runs per 150 games. Other estimates have all been close to Rally's as well, with very different methods.
Look, these park factors make perfect sense. The average Red Sox left fielder will end up just around average -- either you believe that the Sox have had great left fielders over the past 20 years, or you have to buy this park factor.
Plus, Manny looks bad too. I don't see why anyone would argue that he isn't a terrible left fielder. -20 runs a year sounds about right.
I think so. I used Ralley's park factors and got 17 plays per 400 chances counted against a Boston LFer that shouldn't be (14 / 0.831 = 17). Now Manny had 248 chances in LF in 2006, but when you adjust for the Monster it goes down to 237 which brings his ZR up to .724. This puts him at an actual -30 plays and -1 assists for -26 runs and -37 runs per 150 (the average AL LFer saw 340 chances in 2006).
Yes, he did doubt his numbers, but he did so before Rally's PF were published. He seems to think those handle the Fenway issue.
If a savvy front office makes such an offer - or, better yet, if many teams do - I'd start to wonder about the legitimacy of that -15 to -20 range.
Do you work for Grayson Moorehead?
"Joe, I was looking at your Fenway park splits for 2006.
By your data, the monster shows a run park factor of -36 per 400 plays. If we had multiple years of splits, that's what we should use.
Right now I'm leaning towards averaging the two, your stuff is more accurate for 1 year, while I have 20 years of data. That would yield a PF of -25 per 400 chances.
I have Manny, unadjusted, at -31 runs with 248 chances.
So using my adjustment by this article, he's -22
Using the AVG, Manny's -16
And using just your data, he's -9.
I think the truth is somewhere in that range."
For defensive park factors, we're looking at about a fraction of the data - I can't imagine the error range of 1-year defensive park factors being small enough to have any real value. I would personally use all defensive park factor data going back to the most recent architectural changes at the relevant position and ignore 1-year factors.
So that's why I suggest averaging my 20 year stuff with Joe's 1 year.
That being said, another comment posted yet another analysis of fenway defense, this time including home/road splits of all LF who ever played in fenway (this past year only).
http://mb9.scout.com/fbaseballfrm8.showMessageRange?topicID=1182.topic&start=14&stop=40
The -0.220 factor is much larger than Rally Monkey's -0.042. Obviously we can't be sure until they crunch all the numbers, but my gut feeling is that the park factor is extremely underrated in fenway as well as other high-walled stadiums (minny, houston, arizona, etc).
also, i have to wonder since Dial's numbers relate expected chances (and not actual chances) to zone rating (already flawed), if there is some value "lost in translation." granted, Dial writes off the possible difference between expected to actual chances by sheer volume of data analyzed, but all of manny's LF chances are always shrowded in outlying situations, which despite sheer volume of data, could still veer from statistical norms.
all that being said, i am in no way arguing that manny is a good defender. i'm just saying maybe he's not as bad as everyone seems to think he is. maybe in some strange/bizzaro way, manny's defense is like jeter's...seemingly bad, but not as bad when truly analyzed (vs. jeter whose is seemingly good but not that good when truly analyzed).
and maybe that just adds even more value to guys like Randy Winn who posted a +2 & +4 RSpt & RS/150 score in PacBell's RF (or that Willingham was not that bad in Joe Robbie's LF).
That's Joe Arthur's work that I referenced above. And the .220 difference only applies to home games - you have to divide by two to apply to the whole state line.
And in post #8 you can see Manny's D with my factor, Joe's, and an average of the two.
Now, if there are types of outfielders who are inherently well-suited or inherently ill-suited for the Monster, we should try to identify what those types are and which type Manny is. That would give us some insight into what his real value is to us, and what his real value is to someone else.
***
It's not too bad, actually. The margin of error is +/- .047 (+/- .023 if we divide the park factor by two to apply to the player's overall zone rating -- which is what we'll do). Thus, the one-year park factor is somewhere between .087 and .133 points. Now there are still some issues that the margin of error does not cover, as conditions change from year-to-year, but the lower range of Joe Arthur's park factor seems reasonable. Rally's adjustment is .042, my estimate was .064, which turns out to be about halfway between the two.
And because of the Monster, most LFs have already peeled off to play the caron even when the ball hits at less than six feet.
-- MWE
Using the AVG, Manny's -16
And using just your data, he's -9.
To be clear, this is for the 248 chances, right? If we were to prorate to 150 games, the numbers would come out to about -31, -22, and -12, right?
Yes, thats for the 248 chances.
If I was trading for Manny, I'd be looking at him as a DH, so I'm not too concerned about his defense. I don't know why you guys would want to trade him at this point though.
1. You can't replace him this year. Soriano and Lee are not in his class as hitters.
2. Contract is really not a problem anymore - only 2 years left. I'd rather have Manny 2/40 than Lee at 5/70 or whatever.
3. He changes his mind all the time. If you ever do get a deal worked out Manny will probably just decide to veto it. Even if he wanted to be traded somewhere, I would be surprised if he didn't threaten to veto unless the new team picks up his options - and that's very likely a deal killer.
No, I want to trade him mainly because the Red Sox have something like 36 DHs in their starting lineup. If his defense is indeed -20 runs then I think it's possible to replace his production more cheaply. This assumes that other teams are going to either acquire him to be a DH or do not think he's that bad of an OF. For example, if the Tigers were willing to trade Bonderman for him, I'd have to really consider that. Or if the Dodgers decided that they needed power and were willing to give up Drew+. Or if the Angels wanted to give up, oh I don't know... maybe... Santana! You'd trade Santana for him, wouldn't you? Of course you would!
In the same vein, I'd say his contract is pretty bad if his defense is indeed -20ish. As for point three, I certainly hope that before they lift a finger to trade Manny without very clearly establishing the terms with Manny in advance.
As in would you score it any differently if someone positioned themselves perfectly to play a one hop carrom as opposed to someone batting a ball down and retrieving it...
If there's one thing Manny's good at defensively (and there probably is only one thing), its playing balls off the wall and keeping the hitters to a single. I don't believe any defensive statistic measures this, and its a flaw in the system. That said, I don't believe it makes him a good fielder, or average, just not as bad as the systems suggest.
Regardless of his fielding, I think the Sox will be hurt in the short term if they trade him. I don't think they'd get near equal value for him in a trade, and I don't think they would sign anyone reasonable with the money that would be freed up (assuming they don't have to eat part of the rest of the contract). 40/130/.320 is not easy to replace, and I do think Manny provides protection for Ortiz, particularly when the game is on the line.
1) as measured by ZR, in 2006 Manny had 63% of his opportunities at Fenway, so whatever you like for a park factor, in evaluating his 2006 performance, it shouldn't be applied as if he had a 50/50 split in opportunities
2) 2006 could be the start of something new in terms of a park factor. The ".406 club" [1988-2005], a glass-enclosed giant luxury box behind home plate, was dismantled and replaced by the open air "State Street Pavilion" before the 2006 season. Immediately in 1988, the .406 club [originally known as the 600 club] was blamed for changing air currents in the park and suppressing the distance fly balls travelled. It was speculated in preseason this year that fly balls should carry better with the .406 club gone. Better carry should mean more balls off the wall ( and over it). As we accumulate more data the next couple of years, we'll be able to tell better how well the ZR differential I measured for 2006 holds up. If it more or less does, it needn't contradict Rally's park factors here, because his almost entirely coincide with the presence of the .406 club.
3) Manny had a horrific level of performance in ZR after the all star break. Realistically this has to be attributed to his knee injury, and it is unclear how healthy his knee will be going forward. This has to be taken into account as a risk for his future value and future defensive projection.
I agree with villageidiom in #2; multiyear park factor data is good for projecting, but it is problematic to apply it to evaluate a single year where performance may be altered by effects which need not cancel out within the year, such as unusual weather or an unusual skew in the quality/handedness of the pitchers who happened to pitch in the home games.
I collected data for 2006 at the granularity of each home series. I personally was struck by the consistency of the differential in LF between Fenway and other parks; the 200+ point differential emerged early in the season and seemed to stay stable. To me that stability suggests that it wasn't luck or weather making 2006 an outlier. If 2006 is an outlier compared to prior years, I'm betting on the demise of the .406 club as the structural cause. If anyone feels they can perform some sophisticated analysis to estimate likely error by looking at subsamples of my data, let me know and I can send you the raw data.
I don't know how many hit off the wall in 2006, but Fenway severely depressed home runs last year.
Only SF had a worse 1 yr HR factor.
The issue arises in getting fair (or better) value for Manny in a trade. If you make a trade based on the notion that Manny is -20 (or -47) on defense, and he's actually -8, you won't get fair value in return.
The key to it is exactly as you suggest. They need to understand the Fenway Effect.
I don't know how many hit off the wall in 2006, but Fenway severely depressed home runs last year.
Poor David Ortiz.
the basic question is, is there a team out there that values Manny more than we do? If so, we should make a deal, preferably for players that we value more than they do.
To approach that question, wee need to figure out how much we value Manny. The Monster factor has no bearing on this at all, unless we can separate out whether Manny is particularly well suited, or particularly poorly suited, to the Monster. We don't know that.
Personally, I suspect that Manny is particularly well suited to the Monster. The short porch and the virtual absence of any foul territory make for a very small area to cover, and that is excellent because Manny has nearly zero range. He has experience playing the caroms and he has, if not a strong arm, an accurate arm with a quick release for making the pegs to second, third, and home.
And of course he has made friends with the Wall scoreboard operators. ;-)
I think Manny would be overexposed playing RF in Fenway or playing any outfield spot anywhere else. As an outfielder, he is more valuable to us than to any other team. As a DH, on the other hand, he is probably more valuable to other teams than to us, because we have a DH.
Bottom line: if we trade with an AL team who plans to DH him, we could probably work out a sensible win-win deal. The Angels might be the best fit. Otherwise, the only way we trade him is if we can exploit someone else's stupidity.
playing balls off the wall and keeping the hitters to a single
Almost all defensive metrics concentrate on outfielders making outs as opposed to them holding singles to singles and doubles to doubles, &c;. And certainly the ability to catch the ball in the air seems like a reasonable proxy for being able to track it down on the bounce. The thing is that, bad an outfielder as Manny may be, he is still at least physically capable of running down the more routine hits out there and making sure they stay routine. Most other players who hit like he does -- e.g. Thome, Ortiz, Hafner, Giambi -- would be major disasters in left field. Manny is at least only a minor disaster. His routine ability has a certain amount of value that tends not to get measured at all.
Indeed. 22 HR at home, 32 on the road. His OPS was nearly the same, Fenway helps his average, but it probably cost him the AL home run record.
playing balls off the wall and keeping the hitters to a single
There was a study on this, all the way back in the 1984 Abstract. Jim Rice was pretty good at it.
The Angels might be the best fit.
I still don't see a deal. I think Juan Rivera would have to be part of it. Upgrading to Manny's bat is a big plus. Downgrading from Rivera's glove to Anderson's hurts (though Scioscia is determined to play Anderson and DH Rivera pretty often anyway - so thats a minor point)
The salary difference between ManRam and JRiv is enough to buy us any free agent we want though. I'm thinking for any deal to happen the Red Sox would have to take some $ back. Any interest in fat, injured, former Cy young winners?
The salary difference between ManRam and JRiv is enough to buy us any free agent we want though.
Not sure I follow. I thought your earlier point was that Manny could not be replaced in this market. If he can't be, then you can't buy a similar free agent with your leftover dough.
I'm thinking for any deal to happen the Red Sox would have to take some $ back.
But you just said his contract was no longer bad. Why would they have to take money back if they're trading a good contract?
Any interest in fat, injured, former Cy young winners?
Can you really get us Clemens?
Seriously, is Colon expected back at full strength in 07? I'd definitely consider him if he is.
Darren, I know you know how good his peripherals were last year. Considering he's 24 and is two years away from FA, why would you consider it? Manny's an amazing hitter but I think the Red Sox would be very happy to make that deal.
as a line drive goes screaming over his head, etc.? If he was a good defender back then is his defensive
decline normal or a bit beyond normal?
Its not bad - but he's no bargain either. I'd take Manny and his contract. But what would I give up for it? Not a top prospect, not a cheap young pitcher.
Rivera is a bargain. Actually, Colon might not be a bad deal for the Red Sox. Assuming they were happy with his rehab progress. Colon for one year at 14? million is probably a lot better than some of the other options out there, like Suppan for 3/30 like I've been hearing.
With Schilling, Beckett, Wakefield, and Papelbon seems like the Red Sox will need to trade for or sign one starter.
Not sure I follow. I thought your earlier point was that Manny could not be replaced in this market. If he can't be, then you can't buy a similar free agent with your leftover dough.
I'd rather have A-Ram than ManRam right now. Don't think the Red Sox are in the same market for a 3B, with Lowell and Youkilis. If A-Ram resigns with the Cubs I'd be more interested in ManRam.
1. Is he willing to be the primary DH or is he going to ##### about not playing in the field?
2. Is he going to demand one or both of his options be picked up as a condition of waiving his 10/5 rights? If so, there can be no deal.
I'm OK with paying 20 mil a year for ages 35-36 to a guy who's as good a hitter as ever. I'm not OK with doing the same for ages 37-38 as well.
The Red Sox do need at least one starter, but I don't think they accomplish anything by dealing Manny for Colon. Or did you mean Colon and Rivera? Colon seems like an awful risk and I don't trust Boston's medical staff much. What about Aybar, are you guys ready to throw him into a deal, what with all your young superstar SS/2B types?
One other option would be to include OCab, who's probably similarly overpaid to Manny (if you consider Manny a DH). I'm not sure what kind of a deal would be built around that though.
1994 +2
1995 -4
1996 +4
1997 +4
1998 +4
1999 +1
2000 -6
Using the park factor based on an average of my stuff and Joe's:
2001 +7
2002 -2
2003 -5
2004 -10
2005 -17
2006 -16
Colon + Rivera + a shortstop could work if the Angels got something useful back.
You guys happy with Mike Lowell? Might be just what our defense needs.
My preferred SS for 2007 is Maicer, and beyond that Brandon Wood.
Lowell didn't hit much in the second half, and I doubt he does all that well in our park, but our defense killed us last year - and with Manny I wouldn't worry about not getting above avg offense out of third.
Colon, Rivera, Aybar for ManRam and Lowell. You can have OC too, or leave him out. Either way.
1) Yun Taragoashi (yes that was me)
2) Paul Kilgus (remember how funny he was?)
3) Sid the dickfuck (that guy was great)
4) EV-9D9 (remember how he was way funnier than Ackbar?)
5) I'm sure i could think of some other ones if you gave me a chance to think about it. I was well known around here a few years ago. I gave Idaho a bad name. I don't live there anymore, but I imagine if you tried to log on at BSU or U of I you would have a tough time thanks to me.
I know more about baseball than anyone i've ever met. i just can't help but act like a middle schooler online.
he gets the ball to the infield fast
he makes the routine plays
sometimes he makes great plays
Russian is on to something,he accelerates slowly--2 years ago Millar commented about this when Aramis Ramirez blew out a hammy coming out of the box--it was in defense of Manny coming out of the box methodically and taking a beating for it
the Monster IS one outlier,but his positioning has to have the equivalent value vis-a-vis zone studies
I don't remember seeing this article at the time, but it made the opposite suggestion about wind currents at Fenway in 2006. In case the link disappears, in this Boston Globe article, the architects of Fenway Park's reconstruction had comissioned a study of wind effects with their new construction. The consultant's conclusion: 1) the .406 club had suppressed carry on fly balls 2) replacement by the State Street Pavilion basically would have no effect 3) however a new tall (12-17 story) building complex built 800 feet away was expected to interfere with the prevailing spring/summer winds, and suppress fly ball carry to a degree the architects were unwilling to attempt to quantify.
So the poor HR park factor for 2006 which Rally notes may turn out to be a permanent factor.
The effect on zone ratings is still unclear without further study. In LF there would be a tradeoff between fly balls which would have been home runs turning into uncatchable balls off the wall, while uncatchable balls off the wall are turning into catchable balls in front of the wall. In center and right, uncatchable homeruns might turn into catchable (but more difficult than average) deep flies, etc., etc.
Anyway, don't know how much of that construction was already in place in 2005, but this article does imply that at a minimum we should be putting less weight on older years in constructing multi-year park factors, at least for any urban park like Fenway which isn't surrounded by parking lots. Who suspected that it wasn't enough for the field dimensions and fence height and playing surface to be unchanged to compare the park across years, but that changes in the neighborhood itself need to be taken into account?
If Jessica Alba was willing to pay me a million dollars for a night of passion, I'd have to strongly consider it.
Manny's contract is still utterly terrible. He's about four wins a year at twice market value.
Not much. I think most of the work in 2005 dealt with site remediation and the first 1-2 stories. When I came back in the spring it was huge.
wouldn't have improved itself for the future either (beyond payroll flexibility -- which, though important, is not enough, given the trade).
Put it this way: Lowell for Colon is not a good trade for the Red Sox. Lowell costs $6 million less and, given Colon's last three years, age, and problems with conditioning, has to be worth more in terms of projected onfield contribution. The Sox have enough injured pitchers without needing to take on another.
The team should not be thinking about trading Ramirez in order to waste time on the treadmill of mediocrity in 2007. They're roughly where the Cubs were when they talked about trading Sosa for Nixon. At the time, a Sosa for Nixon trade, on the field, was insane for the Cubs, but given salary and value (both at the time and as events worked out), it would have been an overall win. If the Sox are looking to trade Manny Ramirez, it should be for Nixon-like players. Kendry Morales or Maicer Izturis or (dare to dream) Brandon Wood or Howie Kendrick and so forth. But not Aybar, who again, did not hit in the PC and is no better than the Angels' third or fourth-best shortstop. All of these guys are capable of producing right now, but are unlikely to be worth as much in 2007 as Ramirez. If the Angels want Ramirez, one of these guys has to be part of the package, especially if the Sox are filling the Angels' holes at two positions by giving them Lowell and Ramirez and taking on
someone ridiculous like Colon.
I take it this means they're not moving Manny to shortstop.
The only two ways a trade of Manny - or anyone, really - works out in Boston's favor is if another team values him far more than Boston does, or if we're trading from surplus to fill a hole. Regarding the former, I think Manny's negatives scare other teams more than Boston. On the latter, I don't think we have enough of a surplus of slugging OF/DH types, and the commodity we'd want (quality pitching) won't be offered.
Time will tell, of course. I'd expect a midseason deal rather than offseason, if it happens at all; but it depends on who else we pick up in the meantime.
You mean the garbage that he's never denied and his agent has actually confirmed in the past? Yeah, why do they keep making all that stuff up?
I certainly prefer having Manny in the mix, especially with the injuries all those guys have suffered in the recent past... but if the Sox _are_ shopping Manny as aggressively as some reporters claim (no given, I realize), perhaps they're crunching defensive numbers with that arrangement in mind?
I reckon that if the Red Sox can get Loney, Broxton and another decent prospect (Kemp?) for Manny, they should FedEx Manny to LA at the crack of dawn...
It's weird because the tools they bering to the table imtuitively opfposite of what they seem to be good at but there you have it.
if Drew signs, he's going to be the centerfielder. He'll be way better than the other two.
As for the Dodgers and Manny: some have speculated that this latest business about the Giants was leaked by Boston to tweak Colletti and McCourt. Just speculation, of course, but seems possible.
I know. That's the cunundrum that's going to be created if Manny is dealt and both Crisp and Pean become full-timers. The least objectionable scenario is that Crisp plays left and Pena right. Then only one outfield spot would be relatively weak.
If you have to have a weakness in the outfield, center is that last place you want ot put it.
You must be Registered and Logged In to post comments.
<< Back to main