77 replies to this topic

[FanSinceBoggs]

I read an interesting and provocative piece on Ellsbury by Ben Buchanan.

A few highlights:

Many observers of the Red Sox this offseason have pegged them as a team on the hunt for defense. This is not an outrageous claim—last year the Red Sox ranked 16^th in the league defensively by UZR, with notable deficiencies at 3^rd (second worst in the league), center field (worst in the league), and left field (6^th worst in the league). Between the 3 main starters at these positions, the Red Sox gave up over 40 runs more than they would have by inserting replacement-level defenders at the position.

The real question has to be the outfield. If rumors are to be believed, then Theo would be perfectly content to go into 2010 fielding the same starters as he did in 2009. Can a team that claims to be focused on defense really accept two fielders who aren't just not good, but in fact at the bottom of the barrel?

For all the world, Jacoby Ellsbury looks like he's a defensive star. His diving grabs make regular appearances on top-10 lists, he has speed enough to cover the large territory in center, and he's not prone to careless drops or tragic misplays. And yet, advanced defensive metrics will tell you that he is in fact the single worst center fielder in the majors. Observing the young speedster in action will reveal that the reason behind this—as has been discussed so many times before—is that he gets bad breaks on balls, often having to completely reverse direction after the ball is in the air, costing him valuable time and wasting much of his speed and glove.

Perhaps Theo and the front office expect Ellsbury to improve his reads next year. Maybe they think he was just off in 2009, and that he'll be back strong in 2010. But if not, what are they doing? It's puzzling why the Sox would claim to focus on defense, and then ignore their single biggest defensive problem. Especially with viable options out there. The Detroit Tigers actually crafted a seemingly perfect solution for the Sox, offering them a 1-for-1 deal: Ellsbury for Curtis Granderson, who would have been a huge defensive improvement with similar offensive production (or superior if he rebounds, as his low BABIP would predict).

LINK: http://www.overthemo...-into?ref=yahoo

I would be curious to hear opinions on Ellsbury and, more specifically, on Buchanan's analysis of Ellsbury. I am trying to gain a better understanding of Ellsbury's overall worth. I'm inclined to think that he is an overestimated player and that it would be in the Red Sox's best interest to trade him. With that said, I'm hoping to learn from others on this subject.

A few pertinent questions:

(1) Is Ellsbury's defense as bad as the defensive metrics suggest? If his defense is that bad, then why isn't Epstein looking for another centerfielder? Why the change at 3b but not CF?
(2) It has been rumored that the Tigers wanted Ellsbury for Granderson. Should the Red Sox have made that trade? Is Granderson a better player than Ellsbury?
(3) Now that the Red Sox have signed Scutaro, should Ellsbury be moved down the lineup, perhaps to the 9^th spot in the order, while Scutaro bats lead-off? Scutaro demonstrated better plate discipline in 2009, with a Swing % of 34.5 in comparison to Ellsbury's 40.9. In addition, Scutaro's O-Swing % was almost twice as good as Ellsbury's.
(4) Is Ellsbury an overestimated player? Would it be in the Red Sox's best interest to include Ellsbury in a trade for a starting pitcher or middle-of-the-order bat? If Ellsbury can't play quality defense in CF, and is not needed in the leadoff spot, then he loses a lot of his value--does he not?

Edited by FanSinceBoggs, 12 December 2009 - 03:36 PM.

[PedroKsBambino]

Well, if to really consider the issue (rather than just be controversial) one has to look at 2008 data as well and ask about a few things...just to start the list, others will no doubt add some others:

1. What changed to cause a 25 run difference between the 2008 and 2009 seasons in CF, if anything? What does the difference in data say about the underlying skill vs accuracy of the measurements?
2. How does the outrageously good defense in RF and LF in 2008 impact the assessment? Does that suggest reading the ball is the issue, and if so how fixable does this tend to be? Is there anything on career trends for guys with spectacular speed but poor defensive ratings?
3. To what degree do we have a sample size question with 1 year of data (partial answer: we do have an issue there)
4. Given the answers to those and available minor league information, how likely is it that 2010 will look like 2009 defensively?

Doesn't mean the question about Ellsbury's defense is different, but that article is not a very compelling treatment of the answer to the question, either.

[Eric Van]

The article's take on Ellsbury's defensive problems in 2009 is simplistic. Here's what I believe:

1) The break you get on balls is one of the few aspects of fielding that is subject to genuine variation from one time period to the next, in the manner that hitters blow hot and cold. I don't think anyone who watched Coco Crisp get mediocre jumps on balls before 2007 and then get tremendous jumps on balls in 2007, and then go back to being very shaky in 2008, could dount this. And when you think about this, confidence has to play an enormous factor in getting a good jump. You are reacting at the level of "muscle memory" and committing to an action, all-out.

If you lose confidence in the accuracy of that instinctive reaction (even if you are still letting it happen naturally most of the time), you will freeze a few seconds before committing. I think a lot of CFers who look great but have average defensive numbers because they get mediocre jumps have learned to simply be cautious all the time; they've trained their brains to wait a few seconds before committing to a route. (Watch Grady Sizemore and compare him to Coco in 2007; when the camera picks up Sizemore he usually still hasn't moved, whereas Coco was usually already off in the correct direction as soon as the camera had him.)

If you start to think about the jumps you're getting, so that you are no longer reacting naturally, then you're really screwed.

1b) It does take some guys a lot of reps at the MLB level to learn to read the ball well and gain confidence that they are doing so.

2) Ellsbury had a track record of fielding very well, which suggests he used to get pretty good jumps. His TotalZone by seasons: +6, +19, +20, and then you had his excellent UZR and FB numbers in '08, which you would have predicted quite well from the TZ numbers.

Last year he was fine on balls hit away from him but just awful on balls hit at him, which are the toughest balls to read. He had particular trouble going back, and played way too deep as a result, so that balls were dropping in front of him that most CFers catch even when he got a good jump on them. He also had terrible problems in the first half on balls hit near the wall, but was OK on them in the second half.

I happened to be in Baltimore (along with the whole SABR convention) the night he saved Smoltz with that astonishing catch in CF. He got a tremendous break on that ball and played the wall perfectly. The skills are there.

3) I have no doubt that they're having Ellsbury work on this over the winter and will work on it in ST.

I think there's a very good chance that Ellsbury is a +10 defender within the next two seasons. This year could be average, or he could get there just like that. And he may have upside beyond +10.

[TomRicardo]

(4) Is Ellsbury an overestimated player? Would it be in the Red Sox's best interest to include Ellsbury in a trade for a starting pitcher or middle-of-the-order bat? If Ellsbury can't play quality defense in CF, and is not needed in the leadoff spot, then he loses a lot of his value--does he not?

If you could pull off a trade for Miguel Cabrera or Halladay with Ellsbury instead of having to include Buchholz or Kelly, you do it and run. You can sign Cameron.

Ellsbury was terrible in CF last year. He has the speed to be outstanding defender but has terrible instincts. His noodle arm doesn't help.

[Eric Van]

He has the speed to be outstanding defender but has terrible instincts.

If he has terrible instincts, how did he put up seasons of +19 and +20 in the minors (according to a methodology that has been tested by comparison with UZR and is in fact pretty reliable)? How did he play a nearly full season in 2008 (144 Defensive Games) and combine the three positions for a +17 UZR and +11 FB, including +7 in CF according to both UZR and FB?

His 2009 is the outlier on a career where the numbers otherwise match the tools.

[JohntheBaptist]

If the issue is how he gets breaks on balls and his instincts, then how did he fool so many observers during his time in college and the minors into such ridiculously effusive praise? It seemed, at the time (iirc), like his defense was going to carry him.

What changed? Did that change? Were all those reports bogus? I don't buy that they're reactions to his raw speed or athleticism (as I'm sure that often colors scouting reports) relative to his peers because, if the issue is, again, instincts, then that's something separate from athleticism and speed. They become his only hope absent instinct. Maybe that's what they sort of glossed over? "Sure, he'll get better at reading the ball, and he'll always be this fast." I still think there is a weird disconnect there. I didn't think he was great this past year, but I don't know if I buy the idea that he was "terrible."

Maybe it's the yips? Made a couple bad breaks, heard about it, now overthinks it and gets to indecisive when it's hit to him?

I'd be fine with using Ellsbury in the right trade and then signing someone like Cameron, but I don't know that I'm writing off his defense going forward based on what we saw in '09. It is something to keep an eye on though, I think.

As a sidenote, reading that article- Ellsbury is one of those players that serious baseball fans sort of obnoxiously start to view through the prism of the hype generated by all the "idiots who don't know what they're talking about." I'm fully on board with the idea that he is probably overvalued by your average fan, but it shouldn't play a part in the way he's evaluated. That's sloppy. Maybe this article stops short of that, but it creeps. I think he occupies a spot comfortably in the gray.

[biollante]

I honestly don't care how Ellsbury did in the minors now that he is in the majors. He has to wake up. Maybe it is the yips or some sort of young developmental problem - or maybe this is the way he is - fast but not instinctive. Time will tell. I will not throw him under the bus yet but maybe next December I will.

[LondonSox]

I honestly don't care how Ellsbury did in the minors now that he is in the majors. He has to wake up. Maybe it is the yips or some sort of young developmental problem - or maybe this is the way he is - fast but not instinctive. Time will tell. I will not throw him under the bus yet but maybe next December I will.

You honestly don't care how Ellsbury did in the minors? You think him being good defensively in college and his entire time in the minors is irrelevent because he has to "wake up"? Seriously?
He should have the best coaching available to him, he should get better at Fenway over time. History is the key to judging whether this was a freak year or not. Given you have two opposites in his two years in the majors, how can you suggest ignoring his minor league history is the way to go.
This is the same awesome logic that had people calling for the demotion of Pedroia. His swing is too big, he can get away with it in the minors but that doesn't matter - this is the bigs.
I do however agree this time next year if he doesn't improve then this will be a bigger issue.
In the meantime of course Theo believe Ellsbury will be a solid CF this next year, because there is only recent and limited evidence to suggest otherwise, and years of data that suggests it is so.

Meanwhile, the article is insane, the biggest defensive issues were Bay in LF, Lowell at 3B and SS. SS has been addressed, 3B is seeing Lowell traded, so clearly is expected to be upgraded, and LF is still an open question, but the least important by far.
Also the course of Varitek is at most a backup not an everyday catcher.

That to my mind is an upgrade expected to C, 3B, SS, maybe LF and a resonable expectation of an improvement of the same guy playing CF.
That seems a potentially (while very uncertain until the roster is set) large defensive upgrade to the team.

[iayork]

Is Ellsbury's defense as bad as the defensive metrics suggest? If his defense is that bad, then why isn't Epstein looking for another centerfielder? Why the change at 3b but not CF?

I wondered what the predictive value of this swing in UZR would be. How many other outfielders have had >20-point swings in UZR (I actually used UZR/150) from one year to the next, and what happened to them the following year?

Using FanGraphs' UZR/150 since 2002, I found 24 outfielders (defined as >65 games at a particular position, to include Ellsbury) with that much swing, positive or negative. For the negatives, when someone drops that far, that fast, they're often not back there the next year (in 7/24 the drop was their last year at the position; for another 6, including Ellsbury, the drop was in 2009, and some of those presumably won't be back).

Of the 12 who showed a >20 point drop and continued to play, at least half bounced back, at least temporarily:
Andruw Jones (from 34.7 in 2005 to 13.1 in 2006; then 22.2 and 0.2, and out)
Kenny Lofton (19.9 in 2005; -17 in 2006; 8.3 in 2007)
Corey Patterson (-11 in 2002; then 14.8, 33.8, 11.3, 14.2, 1, 0.7)
Willy Taveras (22.6 in 2006; then -7.1, -3, and 14.1)
Jeff Francoeur (30.1 in 2005, 7.4 in 2006, 16.9 in 2007; but then -4.9, -19)
Juan Encarnacion (all over the place: from 2003 - -11.4, 13.5, -11.1, 7.1, -26.7)
Reggie Sanders (12.9 in 2002; then -7.2, 4.3, -, 9.2)

Some of the rest were just swinging from awful to horrible (Manny Ramirez, Brad Hawpe), so they're not really comps for Ellsbury. I don't see a really close match -- a guy who has a dropoff right near the beginning of his career. Taveras might be the closest and he's one of the most encouraging, UZR-wise, I think. Corey Patterson might be another but I don't have his first year numbers. Francoeur bounced back but only temporarily ... Overall, I don't see a strong pattern, but certainly I don't see any sign that this kind of a drop means that Ellsbury is hopeless as a defender.

Here's the table (thanks, AJ). I have to help one kid with his homework, give baths to both, and get them to bed, but after that I can make a start on the "Comments" column (any pointers welcome, right now most of my comments would be "Fat", "Old", "Awful", and "Beats the hell out of me")

UZR/150: Annual change of 20 points (Outfielders)

CF
First	Last	2002	2003	2004	2005	2006	2007	2008	2009	Comment
Alfredo	Amezaga					12.1	32.1	10.3
Mike	Cameron	8.5	12.5	4.3		-0.6	-10.4	15.6	10.3
Andruw	Jones	23	23.4	28	34.7	13.1	22.2	0.2
Jacoby	Ellsbury							6.9	-18.3
Coco	Crisp			13.5		1.3	22.4	-15.4
Cody	Ross							18.3	-9.5
Kenny	Lofton	-2.5	3.7		19.9	-16.9	8.3
David	DeJesus			-3.1	19.9		5.6	-23.5
Corey	Patterson	-10.7	14.8	33.8	11.3	14.2	1	0.7
Vernon	Wells	-8.7	-18.7	12.2	3.2	7.9	-1.1	-24	-18.7
Willy	Taveras				12.6	22.6	-7.1	-3	14.1
LF
First	Last	2002	2003	2004	2005	2006	2007	2008	2009	Comment
Garret	Anderson	1.7	13.7		-9.4	-7.6	-1	19.1	-16.5
Ryan	Klesko			-29.8	-9.3
Lance	Berkman	-39.7	18.5	-8.3
Reed	Johnson				23.3	39.1	4.4
Manny	Ramirez		11	-9.3	-15.7	-24.2	-28.3	4.8	-15.4
RF
First	Last	2002	2003	2004	2005	2006	2007	2008	2009	Comment
Alex	Rios			28.5	10	13.7	11.2	31.3	-5.6
Reggie	Sanders	12.9	-7.2	4.3		9.2
Jeff	Francoeur				30.1	7.4	16.9	-4.9	-18.7
Sammy	Sosa	-11.7	-8.6	11.1	-14.9
Jayson	Werth							35.6	6.1
Juan	Encarnacion		-11.4	13.5	-11.1	7.1	-26.7
Brad	Hawpe				-16.2	1.8	-27.2	-46.6	-25.9
Bobby	Abreu	0.4	15.2	-9.6	-5	-13.8	-2.5	-25.3	-12.1

Edited by iayork, 13 December 2009 - 08:00 PM.

[Adirondack jack]

I wondered what the predictive value of this swing in UZR would be. How many other outfielders have had >20-point swings in UZR (I actually used UZR/150) from one year to the next, and what happened to them the following year?

If someone will explain to me how to make one of those nifty table thingies, I'll put the numbers up. Here are all the names I found, anyway.

Phreniles link and explanation

Maybe I am off base, but I think it would also be helpful to compare large swings in offensive performance with the data you're looking at, along with some other things.

Edited by Adirondack jack, 13 December 2009 - 07:10 PM.

[SMU_Sox]

Maybe I am off base, but I think it would also be helpful to compare large swings in offensive performance with the data you're looking at, along with some other things.

I think having a "possible explanation column" would be a good idea - and if you want help on that please PM me.

[FanSinceBoggs]

If you could pull off a trade for Miguel Cabrera or Halladay with Ellsbury instead of having to include Buchholz or Kelly, you do it and run. You can sign Cameron.

I'm in agreement with this idea. Obviously, Halladay is now off the trade market, but if the Red Sox sign Cameron, I would like to see them shop Ellsbury for a bat. They could always swing a three-team deal, trading Ellsbury for prospects, and then trading those prospects to a third team for a bat. Some team will covet Ellsbury's speed.

I don't think it would make sense to play Cameron in CF and move Ellsbury to left--Ellsbury's bat is not good enough for a corner outfield spot, especially for a big market team. Ellsbury has limited power potential, and his approach at the plate leaves something to be desired. As such, if the Red Sox sign Cameron, I would consider Ellsbury expendable.

Ellsbury had a track record of fielding very well, which suggests he used to get pretty good jumps. His TotalZone by seasons: +6, +19, +20, and then you had his excellent UZR and FB numbers in '08, which you would have predicted quite well from the TZ numbers.

I think there's a very good chance that Ellsbury is a +10 defender within the next two seasons. This year could be average, or he could get there just like that. And he may have upside beyond +10.

You know much more about defensive metrics than I do. I'm wondering if this kind of decline in defensive metrics is unusual especially for a player in the prime of his career? Is it unusual for great, or above-average, defensive players to experience a significant decline in their defensive metrics or are such metrics generally stable from year to year?

Maybe Ellsbury's 2009 defensive metrics were a statistical aberration, an imperfection of sorts in measurement. Perhaps Ellsbury's 2009 defense was not much different from his 2008 defense, and that the differences in measurement were the result of the difficulties (and instabilities) in quantifying defense through statistical means.

Intuitively, it strikes me as being kind of weird, that a player could go from being a high quality defensive CF to being one of the worst in the league. Accordingly, maybe we shouldn't make too much out of Ellsbury's 2009 UZR until it turns into a clear and consistent trend.

Edited by FanSinceBoggs, 14 December 2009 - 09:01 PM.

[Eric Van]

I don't think it would make sense to play Cameron in CF and move Ellsbury to left--Ellsbury's bat is not good enough for a corner outfield spot, especially for a big market team.

This is completely untrue. The average starting LF (2007-9) has an EqA of .282; Ellsbury was .276 (adjusted for the AL East) last year and something like .291 from June onward. Considering how good his glove would be, he'd be very much an asset out there. You are forgetting his base-stealing, which adds about .009, and you are overvaluing power and undervaluing OBP.

(The average starting CF, BTW, was .267. These figures are not to be confused with the less meaningful figures for all players at each position -- there is absolutely no difference in hitting between backup CF, RF, and LF any longer, and that artificially depresses the number for all LF compared to all CF. I just found that out in the last couple of days and may have more to say about it later.)

Finally, seeing the progress Jacoby made at the plate last year, does anyone doubt that he could have put up that .291 EqA over an entire season? How many CF were better than that last year? After Beltran got hurt, just two: Torii Hunter at .297, Matt Kemp at .293.

Is it unreasonable to think that Jacoby could be a .295 EqA, +3 baserunner, +8 defender within the next two years? I think that's a perfectly reasonable hope -- not best-case, but good-case. Well, that's a 58 RARP player. That's the best CF in baseball. That's the best player on the team. That's a guy who's better than everyone but Pujols, Mauer, Hanley, and Utley, maybe Longoria (and Zobrist if he's for real).

Basically:

1) Folks are really unaware of the mediocrity of CF offense these days.
2) People are forgetting how genuinely good he was from 5/31 on. .303 / .366 / .438, in the AL East, and the best base-stealer in the game.
2) If Jacoby were an average base-stealer and got his base-stealing value from HR instead -- which is to say, 12 HR from May 31 on -- there would be no one on this board who wanted to trade him.

[LahoudOrBillyC]

As I said in the other thread, I am pretty bullish on Ellsbury, who I consider the best CF in the American League. I would rather they traded Clay than Jacoby.

[Foulkey Reese]

As I said in the other thread, I am pretty bullish on Ellsbury, who I consider the best CF in the American League. I would rather they traded Clay than Jacoby.

Torii Hunter had 103 OPS points on Ellsbury in 2009. Jacoby was -18.6 UZR vs Hunter's -1.4. I'm somewhat bullish on Ellsbury as well, but there was no way he's the best CF in the AL. I'll agree that he has the potential to get there though.

[Snodgrass'Muff]

Ellsbury also stole 52 more bases than Hunter last year, which OPS doesn't account for. Ellsbury had 41.4 VORP last year, Hunter was at 41.0 so I'd argue they're closer than you think. Though I still come down on your side and would take Hunter's 2009 over Ellsbury's. Ellsbury has the advantage of being younger and having room to continue improving, though, which Hunter very likely does not.

Long term, Ellsbury is the better option but I can see Hunter having a better year in 2010. Even still, I'd trade him for a player like Adrian Gonzalez in a heartbeat, and I'd probably do it before I'd move Clay. The rotation is going to get very expensive, very soon if they extend Beckett and they're going to need a cheap option to fill in the back end. Or they'll simply need to replace Beckett if he walks. Lackey provides a nice fall back, but we're either looking at a really expensive rotation, or a hole to fill. I'd much rather have a guy with the upside of Clay than more year to year reclamation projects. And Clay has a better chance to succeed at the major league level than any of our pitchers in the minors.

So I move Ellsbury before I move him.

[LahoudOrBillyC]

Torii Hunter had 103 OPS points on Ellsbury in 2009. Jacoby was -18.6 UZR vs Hunter's -1.4. I'm somewhat bullish on Ellsbury as well, but there was no way he's the best CF in the AL. I'll agree that he has the potential to get there though.

I think you, and others, ridiculously underrate the value of Ellsbury's basestealing.

[iayork]

I'm wondering if this kind of decline in defensive metrics is unusual especially for a player in the prime of his career? Is it unusual for great, or above-average, defensive players to experience a significant decline in their defensive metrics or are such metrics generally stable from year to year?

Having one very good year (by UZR/150) does make a player more likely to have other good or very good years (even not counting Andruw Jones' awesome run). But there is a lot of variation, and there are several cases where players went from very good, to so-so or bad, and then back again.

Here are the outfielders who have had a >20 UZR/150 since 2002 (similar, but not identical, to the list of >20 year-to-year swings in UZR/150 that I posted above). Interesting cases include Corey Paterson, Jeff Francouer, Austin Kearns, and Alex Rios.

Outfielders with UZR/150 over 20

Pos	First	Last	2002	2003	2004	2005	2006	2007	2008	2009
LF........	Gregor	Blanco							20.1
LF........	Chris	Burke				27.6
LF........	Reed	Johnson				23.3	39.1	4.4
LF........	Carl	Crawford		13.9	23.2	14.2	9.6	-1.4	25.6	17.5
LF........	Willie	Harris						9.3	30.9
LF........	David	DeJesus					15.0		22.9	15.7
LF........	Alfonso	Soriano					6.6	22.4	5.2	-15.9
LF........	Jay	Payton		2.0				2.1	29.5
....CF....	Alfredo	Amezaga					12.1	32.1	10.3
....CF....	Darin	Erstad	22.3	34.7
....CF....	Nook	Logan				20.2		-1.8
....CF....	Andruw	Jones	23.0	23.4	28.0	34.7	13.1	22.2	0.2
....CF....	Shane	Victorino					25.5		7.8	-4.2
....CF....	Felix	Pie						20.7
....CF....	Franklin	Gutierrez								27.1
....CF....	Coco	Crisp			13.5		1.3	22.4	-15.4
....CF....	Corey	Patterson	-10.7	14.8	33.8	11.3	14.2	1.0	0.7
....CF....	Jason	Michaels				37.1
....CF....	Willy	Taveras				12.6	22.6	-7.1	-3.0	14.1
........RF	Alex	Rios			28.5	10.0	13.7	11.2	31.3	-5.6
........RF	Gary	Matthews, Jr.	36.9		14.9
........RF	Franklin	Gutierrez						21.0	33.5
........RF	Jeff	Francoeur				30.1	7.4	16.9	-4.9	-18.7
........RF	Trot	Nixon	5.5	5.5		23.6	3.7	-12.9
........RF	Ryan	Sweeney							29.6	32.5
........RF	Will	Venable								24.6
........RF	Randy	Winn					21.2	9.8	17.8	17.4
........RF	Clete	Thomas								30.6
........RF	Jayson	Werth							35.6	6.1
........RF	Austin	Kearns	21.7			0.9	7.2	13.7	18.5
........RF	Gabe	Kapler			1.2					25.8
........RF	Jeff	DaVanon		35.1

[Lose Remerswaal]

Curiouser and Curiouser

Over the past three years, Red Sox center fielder Jacoby Ellsbury has emerged as a scintillating defensive asset, and he was recognized for his accomplishments by MLB.com.

The Web site named Ellsbury its Defensive Player of the Year on Thursday, when the online voting results of the eighth annual This Year in Baseball Awards were announced.

Take THAT all you Dreamboat Haters!

[BucketOBalls]

2) If Jacoby were an average base-stealer and got his base-stealing value from HR instead -- which is to say, 12 HR from May 31 on -- there would be no one on this board who wanted to trade him.

So a HR is worth about 5 stolen bases? That sounds about right, but it would be interesting to know where it comes from. Anyway, 20 HR + average base running(>10 I would think) does sound much better.

[TheBenzingerGame]

So a HR is worth about 5 stolen bases? That sounds about right, but it would be interesting to know where it comes from.

I'm guessing it comes from Linear Weights, or a rough facsimile thereof. Common coefficients for HR and SB in these kinds of formulas are ~1.4 and ~0.3, respectively.

[donutogre]

Not sure why I'm posting this...perhaps just to show how backwards most of the media is these days. Perhaps because if I had to suffer through this, I want others to do the same

http://www.nesn.com/...eaningless.html

Choice quote:

But while stats are invaluable, they are far from infallible.

There is no better example than the case of Jacoby Ellsbury. The man has the speed, quickness and instincts to steal 70 bases in a season in the majors, yet defensive numbers say he's a terrible outfielder. And that's supposed to make sense?

Despite the statistics, Ellsbury was named Defensive Player of the Year on MLB.com. Granted, the honor was voted on by fans, but it proves a point: The kid can play defense.

Stat geeks will tell you that those magnificent diving catches are the result of Ellsbury's speed making up for his bad read of the ball off the bat. That's all well and good, but chances are, if the ball ends up ensnared in Ellsbury's glove, the Red Sox are happy. Terry Francona probably sleeps a little easier knowing he's got such solid defense in center, and Red Sox pitchers likely don't mind knowing that no bloop will fall in shallow center and no blasts will hit the dirt in the triangle.

I love how we're trusting the fans to accurately judge quality defensive play now.

[Jimy Hendrix]

Stat geeks will tell you that those magnificent diving catches are the result of Ellsbury's speed making up for his bad read of the ball off the bat.

But that's not even a stat, that's a thing you watch. With scouty eyes even. I thought we're supposed to love things a grizzled baseball guy could see.

[donutogre]

But that's not even a stat, that's a thing you watch. With scouty eyes even. I thought we're supposed to love things a grizzled baseball guy could see.

He also disses scouts in the same article, when he talks about how scouts thought Mike Lowell's bat speed had slowed significantly after 2005.

[JMDurron]

That article could become pretty amusing if Cameron gets the bulk of the playing time in CF over Ellsbury. Well, more amusing, I guess you could say.

[Sprowl]

Not sure why I'm posting this...perhaps just to show how backwards most of the media is these days. Perhaps because if I had to suffer through this, I want others to do the same

http://www.nesn.com/...eaningless.html

Choice quote:

Red Sox pitchers likely don't mind knowing that no bloop will fall in shallow center and no blasts will hit the dirt in the triangle.

I love how we're trusting the fans to accurately judge quality defensive play now.

The bolded part above is a laughable statement. In 2009 Ellsbury let innumerable bloops fall in shallow center, especially shallow right-centerfield. On the other hand, Ellsbury did get a lot better at covering deep center as the season wore on.

[Dogman2]

The bolded part above is a laughable statement. In 2009 Ellsbury let innumerable bloops fall in shallow center, especially shallow right-centerfield. On the other hand, Ellsbury did get a lot better at covering deep center as the season wore on.

Defensive positioning has quite a bit to do with this, which is unfair to Ellsbury not catching bloops and running down the deeper balls.

[Sprowl]

Defensive positioning has quite a bit to do with this, which is unfair to Ellsbury not catching bloops and running down the deeper balls.

Did you notice Ellsbury being positioned deeper in centerfield as the season went on? It's plausible, but I didn't notice it myself. It seemed instead that his breaks were noticeably poorer on balls coming in, possibly because he was so concerned about not letting balls get hit over his head that he froze, or at least reacted more slowly to softly hit balls.

In neither case would it argue for calling Ellsbury a great fielder in 2009, but it might spread part of the blame for poor results onto the coaching staff.

[Dogman2]

Did you notice Ellsbury being positioned deeper in centerfield as the season went on? It's plausible, but I didn't notice it myself. It seemed instead that his breaks were noticeably poorer on balls coming in, possibly because he was so concerned about not letting balls get hit over his head that he froze, or at least reacted more slowly to softly hit balls.

In neither case would it argue for calling Ellsbury a great fielder in 2009, but it might spread part of the blame for poor results onto the coaching staff.

In no way was I posturing for Ellsbury as a good CF. I would think some of the blame has to be on the staff because playing him deeper negates some of the effect of having a weaker arm. It's just another example of his poor defensive season.

[Pandemonium67]

The bolded part above is a laughable statement. In 2009 Ellsbury let innumerable bloops fall in shallow center, especially shallow right-centerfield.

Coulda sworn I saw a bloopie piece of crap fall in front of him not that long ago. Off the bat of Vlad Guerrero, maybe...

[amarshal2]

In no way was I posturing for Ellsbury as a good CF. I would think some of the blame has to be on the staff because playing him deeper negates some of the effect of having a weaker arm. It's just another example of his poor defensive season.

I really didn't think his season was poor. It wasn't great, but I thought he was at least average and probably above. My opinion on Ellsbury is that people's opinions are swayed by the unreliable metrics. This could be a classic psychology experiment. If his UZR had come out as +10, there would be hoards of posters here recalling the great plays Jacoby made this year. Because UZR is negative, we get the opposite result. Of course, while I don't think much of UZR, I am a Red Sox fan, so there's biases here I can't control too.

[Eric Van]

So a HR is worth about 5 stolen bases? That sounds about right, but it would be interesting to know where it comes from. Anyway, 20 HR + average base running(>10 I would think) does sound much better.

In this case, I just wiped out the SB and CS numbers from my Contextual Runs spreadsheet and added HR until I restored his RAA value. The conversion between the two will vary a bit depending on the player and the team he's on, though.

[Eric Van]

I really didn't think his season was poor. It wasn't great, but I thought he was at least average and probably above. My opinion on Ellsbury is that people's opinions are swayed by the unreliable metrics. This could be a classic psychology experiment. If his UZR had come out as +10, there would be hoards of posters here recalling the great plays Jacoby made this year. Because UZR is negative, we get the opposite result. Of course, while I don't think much of UZR, I am a Red Sox fan, so there's biases here I can't control too.

I dunno. I was surprised by the huge negative rating but I would have been at least as surprised by a +10, if not more so. It started early in the season when he failed to catch a bunch of balls near the wall. And it was hard to not notice a shortage of good plays made in front of him. I think I would have guessed -4 based on watching him all year. That's because when a guy plays too deep, balls fall in front of him that should be outs and you don't even notice because he can get a good break and still come up short.

I do think his FB rating of -11 is probably closer and maybe much closer to what he really did than his UZR of -18.

[brimac]

I do think his FB rating of -11 is probably closer and maybe much closer to what he really did than his UZR of -18.

Further broken down from BillJamesOnline.net:
Shallow: minus-12
Medium: plus-3
Deep: minus-5

That makes sense: Other than the play where he went over the fence in Baltimore, most of his great catches were going side-to-side in the gaps, right?

[amarshal2]

I dunno. I was surprised by the huge negative rating but I would have been at least as surprised by a +10, if not more so. It started early in the season when he failed to catch a bunch of balls near the wall. And it was hard to not notice a shortage of good plays made in front of him. I think I would have guessed -4 based on watching him all year. That's because when a guy plays too deep, balls fall in front of him that should be outs and you don't even notice because he can get a good break and still come up short.

I do think his FB rating of -11 is probably closer and maybe much closer to what he really did than his UZR of -18.

Agreed that he was often too deep. I put that on the coaching staff. Even if it was Jacoby's fault, is it predictive? You just said he could get a good break and come up short. Seems fixable. When I evaluate him I'm more concerned with the "good break" than the "come up short." Obviously the stats don't differentiate.

[The Boomer]

Where do Ellsbury and Cameron rank in good old fashioned putouts compared to other major league centerfielders? Somebody with average speed but good reads and jumps on balls hit to the outfield might not catch as many balls as an Ellsbury with exceptional speed but below average reads and jumps. It seem to me that the number of putouts per innings played in centerfield ought to be a relatively simple way to compare defenders, particularly because the bottom line measure of how good an outfielder is ought to be the number of outs recorded. How many balls get hit to centerfield is a variable that might be hard to quantify for comparison but, all things being equal, every team should hit approximately as many balls to left, center and right in the outfield over the course of a 162 game season. I'm not sure exactly what UZR and other metrics measure so a primer about those stats would be nice. However, it will be enlightening to compare the putouts per innings played in centerfield for everyone.

[Snodgrass'Muff]

I'm not sure exactly what UZR and other metrics measure so a primer about those stats would be nice.

There's some great discussion with John Dewan about defensive metrics from the chat done here in July.

[Eric Van]

Where do Ellsbury and Cameron rank in good old fashioned putouts compared to other major league centerfielders? Somebody with average speed but good reads and jumps on balls hit to the outfield might not catch as many balls as an Ellsbury with exceptional speed but below average reads and jumps. It seem to me that the number of putouts per innings played in centerfield ought to be a relatively simple way to compare defenders, particularly because the bottom line measure of how good an outfielder is ought to be the number of outs recorded. How many balls get hit to centerfield is a variable that might be hard to quantify for comparison but, all things being equal, every team should hit approximately as many balls to left, center and right in the outfield over the course of a 162 game season. I'm not sure exactly what UZR and other metrics measure so a primer about those stats would be nice. However, it will be enlightening to compare the putouts per innings played in centerfield for everyone.

In fact, pitching staff differences in fly balls vs. ground balls allowed make the raw fielding stats very funky and hard to use. Bill James has done a lot of work trying to eliminate the noise in order to evaluate guys historically for Win Shares. If you don't adjust for GB/FB ratios, you get stuff like Richie Ashburn being the greatest defensive CF ever when he was in reality just a very good one playing with a pitching staff (led by Robin Roberts) that induced a ton of fly balls.

[Reverend]

I really didn't think his season was poor. It wasn't great, but I thought he was at least average and probably above. My opinion on Ellsbury is that people's opinions are swayed by the unreliable metrics. This could be a classic psychology experiment. If his UZR had come out as +10, there would be hoards of posters here recalling the great plays Jacoby made this year. Because UZR is negative, we get the opposite result. Of course, while I don't think much of UZR, I am a Red Sox fan, so there's biases here I can't control too.

The psychological phenomena you are thinking of is called anchoring.

In fact, I'm pretty sure the Gold Glove award voting is about 83.47% based on anchoring effects as per my calculations from my Unfounded Zeal of Respondent metric.

[Dogman2]

I really didn't think his season was poor. It wasn't great, but I thought he was at least average and probably above. My opinion on Ellsbury is that people's opinions are swayed by the unreliable metrics. This could be a classic psychology experiment. If his UZR had come out as +10, there would be hoards of posters here recalling the great plays Jacoby made this year. Because UZR is negative, we get the opposite result. Of course, while I don't think much of UZR, I am a Red Sox fan, so there's biases here I can't control too.

Let's be real here. My opinion of Ellsbury's defensive season has never been formed strictly on a UZR data basis. As Van said, his CF misreads and non-plays at the wall in the early part of the season meant a different positioning prior to any jump he may get on a ball, be it shallow or deep. That positioning is absolutely a coaching decision given his early season misreads. The data, when looked at after the season, shows something similar to what my eyes have told me. He was either poorly placed to start AB's and was moved accordingly based on those misreads or the staff tried to negate his weaker arm problems by moving him deeper. This isn't psychology, it's a confirmation of "your eyes can deceive you" based on a full season's worth of data. Let's not pretend I am swayed or biased by any unreliable metric, that's a rather tawdry accusation.

Besides, Rev's metric is completely reliable.

Edited by Dogman2, 19 December 2009 - 01:44 PM.

[Sprowl]

Further broken down from BillJamesOnline.net:
Shallow: minus-12
Medium: plus-3
Deep: minus-5

That makes sense: Other than the play where he went over the fence in Baltimore, most of his great catches were going side-to-side in the gaps, right?

That division makes sense to me too, although I also remember several very good deep catches near the walls (both left-center and straightaway) in Fenway during the summer.

If there were similar numbers for Ellsbury in 2008, I would guess that he would rank closer to plus-10-to-15 medium if all three outfield positions were combined, but that his problems deep would have been worse than in 2009.

A lot of the results could be skewed by a disproportionate number of balls hit to his strong or weak zones. Coco Crisp seemed to have an inordinate number of opportunities for balls just within reach in 2007; Ellsbury looks (to me, at any rate) to have had more bloops to right-centerfield in 2009 than 2008.

Another factor that might play into subpar performance: Ellsbury hurt his wrist on an attempted diving catch in short right-centerfield in 2008, and that set his batting back for several months. There may have been some carryover timidity on similar balls in 2009.

Or maybe I just remember all the ones that fell in...

[Eric Van]

That division makes sense to me too, although I also remember several very good deep catches near the walls (both left-center and straightaway) in Fenway during the summer.

If there were similar numbers for Ellsbury in 2008, I would guess that he would rank closer to plus-10-to-15 medium if all three outfield positions were combined, but that his problems deep would have been worse than in 2009.

In 2008 he was -1 shallow, +15 medium, +8 deep.

So in 2009 he was (according to FB) 11 worse shallow, 12 worse medium, 13 worse deep.

He was just, like, worse.

[brimac]

It's pretty clear that he's best medium, next-best deep, and easily worst shallow.

Whether that has to do with the way he reads the ball or the way he positions himself in the outfield probably takes a lot more video analysis than anyone here can do.

[Toe Nash]

In fact, pitching staff differences in fly balls vs. ground balls allowed make the raw fielding stats very funky and hard to use. Bill James has done a lot of work trying to eliminate the noise in order to evaluate guys historically for Win Shares. If you don't adjust for GB/FB ratios, you get stuff like Richie Ashburn being the greatest defensive CF ever when he was in reality just a very good one playing with a pitching staff (led by Robin Roberts) that induced a ton of fly balls.

Another important point is the nature of the flies that got by. Someone who played shallow might have more putouts, but when balls got over their head, they would go for doubles and triples. Someone who played deeper would allow more hits but they would cut off more balls in the gaps. So that's another reason why raw putouts are flawed.

Edited by Toe Nash, 19 December 2009 - 03:09 PM.

[amarshal2]

Let's be real here. My opinion of Ellsbury's defensive season has never been formed strictly on a UZR data basis. As Van said, his CF misreads and non-plays at the wall in the early part of the season meant a different positioning prior to any jump he may get on a ball, be it shallow or deep. That positioning is absolutely a coaching decision given his early season misreads. The data, when looked at after the season, shows something similar to what my eyes have told me. He was either poorly placed to start AB's and was moved accordingly based on those misreads or the staff tried to negate his weaker arm problems by moving him deeper. This isn't psychology, it's a confirmation of "your eyes can deceive you" based on a full season's worth of data. Let's not pretend I am swayed or biased by any unreliable metric, that's a rather tawdry accusation.

Besides, Rev's metric is completely reliable.

Sorry, I shouldn't have quoted your post as it certainly looks like I was singling you out when that was not my intention. I have no idea how you evaluate players or separate yourself from bias and did not mean to indicate otherwise.

That said, I stand by my general assertion that those who put a lot of stock in UZR, as a group, allow it to affect their opinion of their visual assessment of players both in real time and hindsight. This absolutely is psychology, and it's been well tested and demonstrated. Analytically, this could be a good thing. As you note, your eyes can deceive you, so when we're talking offensive stats it's better to check JD Drew's 2009 wOBA and selectively remember all the times he got on base to create runs rather than harp on the times grounded out w/RISP. However, since I don't put much more than directional emphasis on UZR, I think it's taken a bit too far around here. I realize how unreliable it is, but I still think I can better gauge a player's defensive ability by watching him play 125+ games a year than I can by checking his UZR.

I guess I kinda buried my lead there: my opinion is that people put too much stock in Ellsbury's UZR. That said, I have no desire to get into another tired discussion about the merits of UZR, so I probably should have posted nothing.

[Adirondack jack]

It's pretty clear that he's best medium, next-best deep, and easily worst shallow.

Whether that has to do with the way he reads the ball or the way he positions himself in the outfield probably takes a lot more video analysis than anyone here can do.

I think that you're basically correct in saying Ellsbury's greatest defensive deficiency is coming in and going back on balls. As others have asserted, positioning could have played a large role in the performance fluctuation (both the measured and perceived) this past year as well as a lingering injury could have altered the way he was playing out there. Probably, like offensive fluctuations, the real talent level is somewhere in between. If I were a gambling man, my guess is that Ellsbury will improve in these deficient areas next year, but I would not go as far to say that he will magically possess great baseball instincts once the calendar turns, either.

edit: While I can agree that positioning has a pretty large role in coming in and going back on balls, judging a batted ball off the bat must be pretty key in this specific skill-set also. While Ellsbury may be able to improve in these areas, I think that as he ages, those improvements will be off-set at least slightly by aging/slowing down, but perhaps I am underrating how much someone at this stage in their career can improve their instinctive reads on balls.

Edited by Adirondack jack, 19 December 2009 - 04:01 PM.

[JimMonaghan]

My head is swimming with all these numbers. Paralysis by analysis anyone?

I hate to be simplistic, but to me it comes down to this - does the ball end up in his glove?

[Eric Van]

While Ellsbury may be able to improve in these areas, I think that as he ages, those improvements will be off-set at least slightly by aging/slowing down, but perhaps I am underrating how much someone at this stage in their career can improve their instinctive reads on balls.

You saw Coco Crisp do it between 2006 and 2007, at age 27, after he'd had 281 starts in CF. He projected to be average after being +17 in LF in 2005 (which is like a +5 in CF) and -3 in CF in 2006 -- and then was +19 instead.

Jacoby is 26 and has had 225 starts there. If he improved by 22 runs, he'd be +7 -- which is what he was in 2008.

[Savin Hillbilly]

I hate to be simplistic, but to me it comes down to this - does the ball end up in his glove?

Yes. Of course the ball also ends up in the glove of every other outfielder in the major leagues. I guess that means they're all equally good?

[FanSinceBoggs]

I still think I can better gauge a player's defensive ability by watching him play 125+ games a year than I can by checking his UZR.

Perhaps there is a better alternative than this. For example, several professional scouting reports on a player's defense would be eminently more reliable than one. A single scouting report is too subjective, might be erroneous, and there is no way to reach a consensus through one report. Alternatively, if you amass maybe five professional scouting reports on a player's defense, and identify points of consensus within those reports, you would minimize error, as well as the problem of subjectivity raised by Dogman: "your eyes can deceive you." Ultimately, accumulating several professional scouting reports, and observing points of agreement among them, might be the best way to properly evaluate a player's defense and reach definitive conclusions.

By "professional scouting reports," I mean reports written up by professional scouts, those who are employed by professional organizations and have the training and experience to evaluate a player's defense. To be sure, I don't mean the self-described "scouts" from the Internet community who lack professional training and experience and are mostly amateurs, or fans of the game, and thus should be differentiated from professional scouts.

Michael Hurley wrote a piece on Ellsbury's defense that contrasts nicely with Buchanan's analysis. I think the piece is weakly argued. Nevertheless, the highlights:

While stats are invaluable, they are far from infallible.

According to FanGraphs, Ellsbury's ultimate zone rating was -18.6 in 2009 – good for dead last among center fielders. Only Toronto's Vernon Wells had nearly as low a rating (-18.2), and only two other players (Colorado's Dexter Fowler and the Cubs' Kosuke Fukudome were in the negative double-digits.

But for anyone who wants to live and die by the defensive statistics, then what is there to say about Ellsbury's 2008 statistics? In 2008, Ellsbury's UZR was a cool 16.5. Yes, on the other side of zero. So if the stats are to be believed, Ellsbury spent the 2008 offseason eating Cheetos and working on deadening his reaction skills. His jump was just too good -- nearly dangerous -- for his own good, so he must have wanted to get much, much worse. You know, for safety's sake.

The reality is that only so much can be determined through numbers and stats, while other things can be surmised through a pair of eyeballs. Ellsbury's defense leads the list.

According to The Sports PhD, "UZR attempts to measure how well a fielder can turn a batted ball into an out." If that's the case, why don't we take into account the park in which Ellsbury plays half his games. For starters, he's got a wall in right-center field that's 420 feet deep. And that wall is 17 feet high and extends deep into a unique corner, thanks to the jutting out of the Red Sox' bullpen.

Then there is the whole matter of the 37-foot wall looming over Ellsbury's right shoulder. The Green Monster complicates matters for outfielders -- notably in that they can't catch baseballs that are 10 feet up on a wall.

Fenway undoubtedly works against an outfielder at Fenway Park. Batted balls that are outs in 29 other parks bang off the Monster; bombs to center field that clear the fences of most ballparks can become some of the trickiest outs of the season.

LINK: http://www.nesn.com/...eaningless.html

Edited by FanSinceBoggs, 21 December 2009 - 11:56 AM.