2010 in Review: Shortstop
Posts in this series:
| Team Offense | Team Pitching |
Team Defense |
| Catcher | First Base | Second Base |
| Shortstop | Third Base | Designated Hitter |
| Left Field | Center Field | Right Field |
| Starting Pitching | Relief Pitching |
Part I: Review
2010 Overview
Asdrubal Cabrera broke his arm on May 17th and missed two months. The Indians had all kinds of trouble finding a decent defensive replacement, as Jason Donald was playing out of position and Luis Valbuena was barely viable on both offense and defense when he played there. When Cabrera came back, the defense didn't get much better.
Defense
Here's Fangraphs' defensive numbers for the position:
| Name | Pos | Inn | DPR | RngR | ErrR | UZR | UZR/150 |
| Asdrubal Cabrera | SS | 825 | 1.3 | -8.6 | -0.4 | -7.7 | -13.4 |
| Jason Donald | SS | 388 | -1.2 | -6.1 | -1.3 | -8.7 | -22.7 |
| Anderson Hernandez | SS | 127 | -0.4 | -3 | 1.1 | -2.2 | -22.4 |
| Drew Sutton | SS | 50 | -0.5 | 1.3 | 0.4 | 1.2 | 46.7 |
| Luis Valbuena | SS | 42 | 0.3 | -1.2 | -2.5 | -3.4 | -90.4 |
The three players who logged the most innings were all poor with the glove, according to UZR. I thought Cabrera was pretty good in 2009 but slipped this season; UZR from 2009 for Cabrera was -4.8 (bizarrely, it had Peralta as having much better range). Cabrera wasn't in very good shape coming into this season, and probably the main reason why he didn't receive a long-term contract offer. His bat still looks good, but he's going to have to improve on the defensive end, or his viability at the position is going to come into question.
Jason Donald was the main substitute for Cabrera while he was out, but he just didn't look comfortable playing the position. Drew Sutton, who they picked up off waivers in August, was the only player of this bunch who looked good at the position. The Indians essentially had a bunch of second basemen around the infield, and it showed.
Offense
The Indians' shortstops ranked in the middle of the pack on offense:
| Rk | PA | AB | 2B | 3B | HR | BB | SO | BA | OBP | SLG | OPS | BAbip | tOPS+ | sOPS+ 6 | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | TOR | 690 | 635 | 34 | 1 | 25 | 40 | 104 | .265 | .312 | .439 | .751 | .281 | 97 | 115 |
| 2 | BOS | 749 | 675 | 40 | 0 | 16 | 61 | 88 | .273 | .334 | .403 | .737 | .292 | 88 | 112 |
| 3 | CHW | 659 | 615 | 29 | 3 | 18 | 28 | 84 | .278 | .308 | .423 | .731 | .294 | 94 | 110 |
| 4 | NYY | 768 | 689 | 28 | 3 | 10 | 65 | 111 | .269 | .338 | .361 | .699 | .306 | 79 | 102 |
| 5 | MIN | 629 | 568 | 30 | 4 | 7 | 47 | 96 | .268 | .321 | .371 | .692 | .309 | 82 | 100 |
| 6 | KCR | 640 | 605 | 30 | 2 | 17 | 26 | 68 | .261 | .291 | .402 | .693 | .269 | 89 | 99 |
| 7 | OAK | 629 | 558 | 29 | 8 | 7 | 52 | 103 | .251 | .317 | .369 | .686 | .295 | 96 | 98 |
| 8 | CLE | 691 | 620 | 31 | 3 | 6 | 47 | 114 | .266 | .323 | .355 | .678 | .316 | 94 | 96 |
| 9 | TBR | 672 | 599 | 30 | 4 | 6 | 53 | 117 | .252 | .318 | .346 | .663 | .303 | 81 | 92 |
| 10 | TEX | 745 | 652 | 17 | 3 | 0 | 69 | 108 | .264 | .339 | .299 | .638 | .316 | 72 | 86 |
| 11 | DET | 638 | 573 | 23 | 1 | 9 | 43 | 97 | .246 | .301 | .337 | .638 | .279 | 71 | 84 |
| 12 | LAA | 695 | 631 | 21 | 4 | 6 | 39 | 108 | .239 | .291 | .314 | .605 | .279 | 74 | 75 |
| 13 | SEA | 592 | 547 | 23 | 3 | 2 | 23 | 110 | .230 | .276 | .294 | .570 | .282 | 79 | 65 |
| 14 | BAL | 598 | 556 | 15 | 1 | 1 | 27 | 69 | .236 | .277 | .272 | .549 | .266 | 58 | 60 |
| TOT | 9395 | 8523 | 380 | 40 | 130 | 620 | 1377 | .258 | .312 | .357 | .669 | .292 | 83 | 93 |
Only three clubs had their shortstops top a .700 OPS, but a .700 OPS was good enough to rank as above-average among all hitters. Two years ago, a .700 OPS would equate to an OPS+ in the mid-80s. So adjust your standards accordingly.
Here's how the Tribe shortstops did individually:
| Rk | G | PA ▾ | AB | 2B | 3B | HR | RBI | BB | SO | BA | OBP | SLG | OPS | BAbip | tOPS+ | sOPS+ | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Asdrubal Cabrera | 94 | 423 | 379 | 16 | 1 | 3 | 29 | 25 | 60 | .274 | .325 | .346 | .671 | .317 | 100 | 94 |
| 2 | Jason Donald | 47 | 169 | 153 | 10 | 2 | 2 | 16 | 11 | 33 | .255 | .313 | .386 | .699 | .314 | 102 | 101 |
| 3 | Anderson Hernandez | 15 | 55 | 53 | 3 | 0 | 0 | 2 | 2 | 7 | .264 | .291 | .321 | .612 | .304 | 122 | 77 |
| 4 | Drew Sutton | 7 | 24 | 22 | 1 | 0 | 1 | 3 | 2 | 7 | .273 | .333 | .455 | .788 | .357 | 112 | 126 |
| 5 | Luis Valbuena | 5 | 20 | 13 | 1 | 0 | 0 | 2 | 7 | 7 | .154 | .450 | .231 | .681 | .333 | 154 | 103 |
| Team Total | 162 | 691 | 620 | 31 | 3 | 6 | 52 | 47 | 114 | .266 | .323 | .355 | .678 | .316 | 94 | 96 |
Asdrubal Cabrera took a major step backward in his offensive game. Jason Donald filled in nicely during Cabrera's absence, so the team didn't see any drop off from what they had, but the Indians were expecting the offense to be a positional advantage, much like center field. Cabrera's slugging especially dropped off, from .438 in 2009 to .346 this season.
Part II: Looking Towards 2011
Not Coming Back
Anderson Hernandez (Claimed off Waivers, 7-21-2010)
The Indians essentially swapped Hernandez out for Drew Sutton in late July. Anderson finished the season with the Houston, and is a free agent after refusing outright assignment by the Astros.
Returning
Asdrubal Cabrera (1st Year Arbitration) - 2 Option Years Remaining
The Indians don't really have a Plan B at shortstop, so they need Cabrera to come to Arizona next spring in shape to play shortstop and be one of the team's main run producers. Cabrera picked a really bad time to have a bad year, as he's heading into arbitration with the Indians presumably in no mood to pursue a multi-year contract.
Jason Donald (Pre-Arbitration) - 1 Option Year Remaining
Luis Valbuena (Pre-Arbitration) - 1 Option Year Remaining
Neither of these guys should be playing much shortstop if the Indians can help it. Donald should be in the mix for the starting second base job next season, while Valbuena seems ticketed for Columbus.
Drew Sutton (Pre-Arbitration) - No Option Years Remaining
I think the Indians might keep Sutton on the roster through the winter. He looked good enough defensively at shortstop to get a look as a utility infielder.
110 comments
|
0 recs |
Do you like this story?
Comments
Cabrera’s defensive regression is extremely disappointing. Unless I am mistaken, I remember people raving about his defense just a few years ago. How could he have fallen so far so fast? Is it really just his conditioning, or was he never quite as good as advertised? I used to think the Indians had a plus defender and decent bat locked down at shortstop for a long time; now it looks like they may have another Peralta on their hands.
Well, let’s see how he responds. By all accounts, he had a poor year defensively and offensively , and interrupted by injury. Hopefully not all of our players have their career year at 23.
by dgcambridge on Oct 21, 2010 11:30 AM EDT up reply actions
The Indians have a serious absence of middle infielders viewed as plus defenders. The situation isn’t just a problem at Cleveland, it extends throughout the minor league system. As Antonetti takes charge and reassesses what is happening throughout the organization, I hope he spends some time figuring out how to improve the defensive quality of players already in the organization.
Was this part of a plan to get offensive production from the right end of the defensive spectrum? Jhonny was a perfect example of such thinking: tolerable (and reliable) defensively, but an offensive force. You sacrifice defense (especially at the right end) but you gain offense from typically offense-challenged positions. A gloveman at catcher or shortstop would seem to run contrary to the thinking of Shapiro.
Not necessarily a great strategy when your pitching staff is heavily dependent on ball-in-play outcomes
That’s where the plan fell apart, of course, but that was the hand dealt to them. If Seattle overestimated the value of defense, the Indians may have underestimated the necessity of gloves.
I think you have it right in these two comments.
Intuitively, I don’t think the Indians are wrong to valuable a selective eye at the plate over stellar defense. I just wish I didn’t have to watch the results.
I think part of it is a previous system trope in trying to find (or force) plus positional value in a moneyball kinda way – selecting for potential offensive value at traditionally defensive positions, because you could “produce” value there.
At its extreme, such a philosophy shuttles the best bats to the most weighted defensive positions, regardless of defensive ability.
In practice, you end up with slightly fatter shortstops and catchers with great bats that get hurt more often because they play catcher.
Good points. You force your best hitters into the most critical defensive positions, and your first basemen and left fielders hit like shortstops.
The 1b and LF hitting like shortstops is not really a necessary outcome of that scenario, even if it happened.
In the new Geico commercial, Marte sings "Let me be myself" on Wedge's front lawn (with the cavemen).
by V-Mart Shopper on Oct 22, 2010 3:57 PM EDT up reply actions
Yes, you’re right. But it is the outcome of thinking that it is easy to find good-hitting first basemen and left fielders.
That’s true, but the presumption is that one can find competent first baseman hanging around Home Depot.
I’m presuming the presumption is Shapiro’s. Something that’s discussed late at night in the darkened lobbies of the Marriott with other enlightened general managers. Otherwise, why convert Santana to catcher?
Traditionally, such a presumption would be appropriate. There was a time that first basemen and leftfielders were abundant. The present time is not a good one for third basemen, first basemen or outfielders. It is a good time for second basemen and short stops. This sort of thing comes and goes.
Otherwise, why convert Santana to catcher?
It’s not like he took John Kruk and put him back there. Santana is (or at least has a chance to be) really, really good as a catcher.
Come on, four billion!
He also has a chance to be better than anyone else on the Tribe at first base. With less chance of injury. The question then is whether Marston is better than LaPorta as a hitter.
personally, I think to even assume Marson>Laporta is ridiculous.
Yes, laporta had a bad season, but Marson was just flat out awful when he was here.
I teach good life choices. That’s why I almost didn’t graduate High School.
Intensive Purposes? I could care less...
your whole argument is a fallacy!
LaPorta managed a .306 OBP in his age 25 season. I’ll bet Marson has a higher on-base percentage next year, in his age 25 season.
so
a) its all about OBP
b) one year makes marson THAT close to LaPorta?
You really want to talk stats?
Laporta’s Slg. is about 80 points higher than Marson’s.
Marson had a .560 OPs and a 59 OPS+ Both of these marks were just flat out terrible. Laporta wasn’t great, but .668 and 88 are at least somewhat passable marks for a major leaguer.
Marson showed nothing to convince me he is a major leaguer or that he will post a .306 OBP next year. Hell, a .306, while it may not be good, is a large (~30 pt) increase compared to marson last year.
You can say that marson may be able to reach that mark, but what evidence is there to support it?
The only think you can be at all positive about marson’s offensive performance is the fact that he got 8 steals; a number and rate he never achieved anywhere in the minors.
I teach good life choices. That’s why I almost didn’t graduate High School.
Intensive Purposes? I could care less...
your whole argument is a fallacy!
Is .668 and 88 at least somewhat passable marks for a left fielder/first baseman? I’m not so sure.
What evidence is there to support my assertion that Marson will surpass a .306 OBP next season? He had a career .366 OBP in 535 minor league games. He even had a .433 OBP in 2008 at Reading.
One would expect LaPorta to be the better hitter, so it makes sense to leave Santana behind the plate, but my point is LaPorta appears to have failed. Maybe it’s injury, but I’m not as sure of that as I was in July.
I wouldn’t judge Marson’s major league abilities on the basis of 370 plate appearances.
I am sure. Those numbers are not at all passable for a defensive zero.
Answering bross’s question, maybe it is all about OBP in this comparison. If LaPorta can’t beat Marson in OBP, it’s unlikely that his power advantage will be enough to outweigh the difference in defensive value.
thats true.
I didn’t realize we were actually talking defense. at this point considering the defense of both players, I would call it an overall wash. I thought this conversation started about offense but I am comfortable in saying that from what I have seen from him overall this year is not significantly better than what we saw from marson.
I teach good life choices. That’s why I almost didn’t graduate High School.
Intensive Purposes? I could care less...
your whole argument is a fallacy!
Maybe its not amazing for a LF/1B. but even for a position like Catcher, were hitting isn’t a premium, .560 and 59 are just terrible. If Marson was the only catcher on our team, we would have finished dead last among backstops in the AL in OPS+. So even for a Catcher, that is flat out bad.
He had a career .366 OBP in 535 minor league games.
and LaPorta had a career .390 OBP in the minors…So I would say there is more evidence to support Laporta improving than Marson.
but my point is LaPorta appears to have failed
I wouldn’t judge Marson’s major league abilities on the basis of 370 plate
370 is nowhere near enough plate appearances to judge a guy’s ability at the ML level but after 623 PAs, a guy has “failed”? I really don’t understand this logic.
So, if after 250 PAs next year and marson shows nothing has he “failed”? I would say its more likely, especially considering at least laporta had somewhat of an excuse for a while in that he was playing hurt.
Maybe you made high expectations for Laporta and he didn’t meet them. However, you are tauting all these stats of how marson will surpass LaPorta while Laporta’s stats were better.
I teach good life choices. That’s why I almost didn’t graduate High School.
Intensive Purposes? I could care less...
your whole argument is a fallacy!
I can see that you are making some adjustment for the difference value between a fine catcher and a mediocre 1B/LF. Your adjustments are just not big enough, however.
Per WAR, just for example, LaPorta was very slightly under replacement level for 2010. Marson, on the other hand, translates as a 1.5-win player over a full season — and that’s if he doesn’t improve at all! (That’s using the B-R numbers. Using FanGraphs WAR, both are a little lower, but Marson is still 1.0 win ahead.)
Even just looking at the offensive side, adjusted for position, Marson edges LaPorta out, 0.3 to 0.2.
LaPorta is also older. Granted, it’s only 18 months. I would argue, however, that the difference between a 25.5-year-old slugger and a 24.0-year-old defensive catcher is actually pretty significant. The slugger at 25.5 is supposed to be improving but not totally lost. The catcher at 24.0 may still be getting it together at the plate, having focused largely on being an effective game-caller and defender up to this point.
The standard is not the same. In terms of where each guy is supposed to be, LaPorta is clearly further behind as he nears his 26th birthday.
I can agree with your last paragraph.
are those WARs adjusted for positional value i am assuming? Yeah, then marson probably has a higher WAR because of what is expected from the position.
I definitely agree with odarek’s notion that Laporta was a disappointment and is not coming along as fast as he should nor as fast (considering positional value and age) as Marson. I guess I just thought it was ridiculous that he jumped to the extreme that after one injury plagued season where he disappointed, he has “failed” an “busted”.
I am anxious to see what Laporta can do next year and if he can turn it around then. but if he disappoints again in this fashion, I don’t mind saying he “busted” because by then, there’s a good chance he does very little in the majors.
I teach good life choices. That’s why I almost didn’t graduate High School.
Intensive Purposes? I could care less...
your whole argument is a fallacy!
I don’t think anyone is saying we should be rash and dump the guy, but understandably expectations are way down. He eventually may not be a bust. Arguing over whether that means he’s a bust so far but may become un-busted, or whether it’s more like he’s not a bust yet because he hasn’t sucked for long enough … this is really academic.
Yes, positional adjustments. Yes, the point is exactly that Marson’s bad looking numbers are a little better than LaPorta’s bad looking numbers. Does this difference justify having a polar opposite view of the two players? Perhaps not. But the age thing really does magnify the position thing considerably.
LaPorta came with higher expectations, and we’ve been waiting twice as long for him to do something. It’s understandable that some of us are more ready to throw him under a bus.
I don’t think anyone is saying we should be rash and dump the guy
usually, when someone says they believe a player has failed, that implies that they have busted. that is what odarek said.
I do think its understandable to throw him under the bus because of the expectations, but I’m just not ready. He has pretty much just one season of Major League work to show us (over the course of 2 seasons). Like I said, I think next year is key, and if he still slumps, I wouldn’t mind throwing him under the bus.
I teach good life choices. That’s why I almost didn’t graduate High School.
Intensive Purposes? I could care less...
your whole argument is a fallacy!
Otherwise, why convert Santana to catcher?
That’s a question for the Dodgers.
Though I look right at home, I still feel like an exile
by Manhattan Tribe Fan on Oct 23, 2010 5:52 PM EDT up reply actions
Why are Replacement Players so hard to find? By definition they are the types of players you can find hanging around in the parking lot, looking for work.
I think your hyperbole is a bit confusing here. Replacement-level guys are not working at home depot, but they are lighting up an independent league or thriving in Triple-A at an age too old to be a prospect.
If they’re AAAA guys or Northern League sluggers—Ernie Youngs—how come the Tribe has to play sub-replacement level players?
We don’t have to. We can play all the Jordan Browns and maybe go 72-90… or we can put guys with more potential to grow out there, win a couple fewer games, but be in better shape down the line.
Adam Miller is healthier than Stephen Strasburg.
Sure, I agree with that line of thinking. But my point is that it isn’t easy to find even replacement-level players. The whole WARP concept endorses the belief that replacement players are fungible. It creates a standard that doesn’t exist.
It may well be that replacement level, in that sense, has been defined at too high of a level of productivity (which also means, incidentally, that we’ve been underrating the value of average players in particular).
It may also be a question of risk/reward, however. That is, in the hopes of developing players who are far better than replacement level, you put guys in there who may be more likely to produce at a sub-replacement level. The alternative is to play guys who are safe bets to produce at replacement level — Ernie Youngs — but astronomical long-shots to produce much better than that for more than a few weeks.
No, the strategy would never force the best bats to defensive positions. I think that is a misunderstanding of the logic or at least a misapplication of it.
I agree. You could wind up with a Garko in the outfield.
by mcrose on Oct 21, 2010 11:48 PM EDT up reply actions 1 recs
Cabrera’s UZR has been crappy for a while now. The question is if you buy into that statistic.
LGT's resident moderate Yankee hating fan.
by Joe. on Oct 21, 2010 12:57 PM EDT via mobile reply actions
I don’t buy into any defensive statistics, in that I think the collective subjective impressions of a player’s defensive value over time or more “objective” than any metric out there. In other words, just watching them play over the course of the year(s).
UZR is one of the better ones, but how can you rely on it when it consistently has head-scratching inconsistencies? How can you quote UZR as a meaningful index for player X and then parenthetically note that, “bizarrely, UZR had the famously slow of foot player Y as having much better range”, when you know for a fact that it isn’t true?
Droobs at his best is a gold glove caliber second baseman and merely a good SS. That’s still the best SS we have, there is nobody else within shouting distance. This year he was not at his best, and the broken arm (wait, how did he break that arm again?) definitely had something to do with that.
Depth is a big issue, but a healthy and fit Cabrera at SS should be at least league average in the field and at the plate. He’s a good player.
UZR is one of the better ones, but how can you rely on it when it consistently has head-scratching inconsistencies? How can you quote UZR as a meaningful index for player X and then parenthetically note that, "bizarrely, UZR had the famously slow of foot player Y as having much better range", when you know for a fact that it isn’t true?
That’s why I mentioned the inconsistency. I don’t think UZR is infallible, heck I mention that just about every time I use it in a post. It’s the best free “objective” defensive measure out there, and I’ll say whether or not my subjective eye test agrees with it or not.
UZR had Peralta as a below-average fielder from 2004-2008. Only in 2009 did the rating go positive, and that might have been because of a small amount of playing time at the position. Conversely, I though Cabrera’s range was pretty good in 2009, and pretty bad this year. I don’t think the injury had much to do with his range, it was more conditioning than anything. I think he still has the ability to play the position.
Yes, and I use it when I have little or no idea of a player’s defensive abilities as a rough (and sometimes wrong) indicator of defensive worth.
I’m just pointing out that when talking about players we are very used to watching day in and day out, the metrics are less reliable than simple cumulative impression.
For example, while I can sorta agree that Droobs’ range didn’t suffer a whole lot after his return from injury, I would definitely say that his overall defense suffered. Whether it was because of lack of conditioning or making up for a leftover physical deficit, Droobs was a less fundamentally sound SS in the second half – he took more chances in the field, rushed throws, made stabs at balls rather than getting behind them, etc.
I can see a given player having consistent defensive metrics over the course of a year, with his actual defensive value and effectiveness actually varying a great deal (due to injury, focus, whatever).
Sorry Jay – I don’t agree with elsandito’s general comment below that defensive stats are merely used to corroborate a subjective notion. I don’t use defensive metrics to corroborate anything – only use them as a starting point when I have nothing to corroborate.
And the logic that gets Jeter gets gold gloves has nothing to do with exactness or metrics – its just that he’s Yankee captain Derek Jeter. Logical and exact in its own right, but nothing to do with defensive value.
Disagree. Gold Gloves are based almost entirely on subjective observation with the occasional glance at fielding percentage. The voters “know” that Jeter looks awesome on that jump-throw play and doesn’t know that he doesn’t get to many balls.
Subjective analysis is certainly not infallible and as we can see, sometimes even the “objective” stats are not infallible and 100% accurate at finding the best player. The best thing to do would be some kind of balance.
Subjective analysis definitely has its merits, but shouldn’t be solely relied upon.
I teach good life choices. That’s why I almost didn’t graduate High School.
Intensive Purposes? I could care less...
your whole argument is a fallacy!
We need to begin, however, by recognizing each type of information for what it really is.
UZR is not inaccurate. It is a report of what actually happened — just like batting average. Whether that report is complete or reflective of a player’s “true ability” — and whether it’s “projectable” into future results — is another question.
That “true ability” is a phantom which may or may not really exist.
Subjective observation is extremely limited. If you want to try to find the best five subjective evaluators, make them watch every play that every defender makes all season and then make their reports, I’m all for that. But we’re not ever going to have that data.
Short of that, the stats are inherently limited in their usefulness, but at least they are real and unbiased information.
That "true ability" is a phantom which may or may not really exist.
I think this sums it up. The quest, again, for an essentialist number that tells us the truth is folly. A fielder of significant “true ability” can still boot ten grounders. The number of variables makes such knowledge impossible to attain.
I am amused to read pitch speeds that move to the right of the decimal point (83.4 mph). Such false accuracy makes people think they have discovered the truth about baseball—and that it can be measured and we needn’t look up from our laptops to observe the game itself. Despite the assertion that the game can be truly measured, there is more than mere numbers at work.
by odradek on Oct 22, 2010 12:53 PM EDT up reply actions 1 recs
we needn’t look up from our laptops to observe the game itself
You know, laptops and actually observing the game aren’t mutually exclusive. I think most people who like exact stats – even some who do something as admittedly ridiculous as adding a decimal to pitch speeds – like them to supplement the observation of the game, not to replace it.
Come on, four billion!
by Joel D on Oct 23, 2010 2:29 PM EDT up reply actions 2 recs
Just quoting mcrose (below). The “problem” with subjective observation is that it isn’t easily transferable. The numbers endure in the form of box scores, but the observational assessments are variable and not easily repeated. Exact stats will thus have more thoroughfare than observation.
All the sabermetricians are waiting for that time, when you don’t have to look up from your laptop to tell how good a guy is with the glove.
Yes, Subjective observation is extremely limited. However, so are statistics measuring the play of a fielder. they do not always tell 100% of the story, because multiple people here do point out the consistencies in a stat like UZR.
Now I believe that stats generally carry more weight than subjective analysis, however neither one is 100% and some harmony needs to coexist between them to get the most accurate evaluation.
I teach good life choices. That’s why I almost didn’t graduate High School.
Intensive Purposes? I could care less...
your whole argument is a fallacy!
How do you “know” it isn’t true?
If a guy looked great to you at the plate but had a 670 OPS, would you say that OPS was a bad stat?
If UZR “consistently has head-scratching inconsistencies,” it may be in part because defensive performance (or play outcomes which are not entirely indicative of raw performance) is itself very unstable, and perhaps UZR merely reflects that fact.
Every saber admits that defense doesn’t fall into the same stat analysis category that offense does. The variables are a lot more difficult to eliminate, which is the essence of stat analysis. At some point, perhaps soon, we’ll be able to track exactly the velocity, angle and trajectory of a batted ball, and rank defensive players ability to get to those same batted balls and turn them into outs depending on their initial position on the field and which way they are leaning to keep a guy close to a base. All the sabermetricians are waiting for that time, when you don’t have to look up from your laptop to tell how good a guy is with the glove.
But we’re not there yet.
Until then, I have to say that Droobs is a better SS than Jhonny.
I can’t speak for sabers, I think the whole idea of that term is stupid.
A lot of the “sabers” are completely clueless as to the limited application of many stats.
Case in point, when we have all that detailed data you say we’re going to have, the sample size will make nearly every scrap of the data insignificant. It may actually be harder, not easier, to find anything meaningful in that type of data.
It may be stupid, but its a definite split to date – stats vs scouting, as if one can’t supplement the other. THAT is stupid, and it will only be less so when a given GM weighs them appropriately, and the actual individuals getting paid for their analysis take full advantage of the strengths of each. There are assuredly less GM’s than the guys that do take advantage of that and can help their bosses.
As far as sample size, I don’t get what you mean – defensive stats are sort of the final frontier right now. Pure statistical approaches, UZR etc, are getting more refined as measurement allows, and cumulative subjective approaches like Tango Tiger has been trying to do (more power to him) are zeroing in on value as well.
Seems to me more data rather than less is the trend, and getting better – its just not there yet for real rankings.
I do have some reservations as to whether or not pure statistical metrics will ever truly measure defensive value, since defense is much more of a team effort and field coordination than batter vs pitcher.
I do really like Tango’s approach, which does suffer from sample size, unfortunately. I’ve always been struck by how accurate an umpire the 10 to 40 thousand fans in the stands make while calling balls and strikes. The cumulative viewpoint, and resulting murmurs, boos and huzzahs from the stands, tends to be amazingly accurate. Unless they’re cultishly rooting for the home team of course.
I’m having a little trouble understanding what you’re trying to say here.
The stats vs. scouting debate is over. It was stupid, yes, and it also was exaggerated. Here’s a great roundtable from almost six years ago that showed that even if you get the most rabid experts from both sides together, it evetually devolves into more or less a baseball lovefest.
An interesting side element of the non-debate is the fact that the stats people are, simply, better at dealing with math-related problems. So it’s not just a matter of evaluating players — who’s going to be good? — but also valuing different types of players and making decisions about them — what should we be willing to pay for and/or trade for a certain player? — and that arguably is where they’ve had more influence. That influence hasn’t always yielded good results, however.
You say you think the defensive data is getting better, but do you really understand the data?
We already have a huge stability issue with defensive data. You probably need at least 3-4 years of data to figure out how good a guy really is. You then need to make some kind of a correction for health issues, maybe also adjustments for coaching and “adjacent” defenders. And of course, by the time you have 3-4 years of data, the player has changed — he’s possibly improved through experience and probably slowed down with age.
Understand that the instability is correlated directly with the sample size, and that we already have a type of data where 1-2 years of data is not enough to tell us anything with any confidence. And now you want to slice the data up further:
the velocity, angle and trajectory of a batted ball, and rank defensive players ability to get to those same batted balls and turn them into outs depending on their initial position on the field and which way they are leaning to keep a guy close to a base
It sounds awesome. It also sounds like for any one particular type of batted ball, you’re going to end up with a sample of maybe 5-20 plays for each defender. So it’s cool how specific the data is, but the sample size will make all this data manifestly meaningless. It’s going to be like evaluating a hitter based on maybe ten games.
I see what you’re saying about small sample size now. And you’re right – if you divide all possible batted balls into some kind of reasonable categories, you end up with not that many plays within each to compare.
Remember my initial point was its tougher to measure defensive ability! I trust my eyes telling me that Droobs is a better defender than Jhonny over UZR telling me the reverse.
As far as future stats, as pure measurement of bodies in motion during a game gets better, or at least internally more robust and consistent, I can’t help but think there will be some statistical method that will yield a “better” defensive metric than UZR and its ilk currently do.
Defensive metrics should be used to corroborate what we observe. The notion that we think a player fields a position well but the metrics disprove it, or we think a player is not fielding a position well, but metrics say he is, will only lead to uneasiness.
I believe Cabrera can play an efficient SS and, eventually, the metrics will agree.
No one should put most of their faith in any metric nor should they put most of their faith in naked observation. Both of these paths lead to mistaken impression.
So the metrics are right when they agree with you? I’m not sure that is what we are going for. There are problems with how we quantify defense, but I don’t think the goal of quantifiable defensive metrics is to get them to agree with what we already think.
The goal of most metrics is to get them to agree with what we already think, although this is rarely admitted.
True in so many ways. Except for the exceptions. In baseball, the only metric that really penetrated recently to change the already thought was OBP. Which led to OPS, and a sudden appreciation of anecdotally recognized truths like R/L splits and, courtesy of Bill James, the significance of age when evaluating prospects.
I disagree almost completely. Metrics attempt to do a number of things. Many attempt to simply record what happens in some form. But on a larger scale, metrics attempt to make things comparable by quantifying them, which isn’t really the same thing as saying “agree with what we already think.”
Codification does add objectivity to the way we appreciate performance. Nobody denies that. But, it also contains limitations to the way the metrics were designed. There are no perfect translations between reality and whatever models we build. That is why I say it is foolish to doubt one’s ability to qualify performance using their senses.
Why are you so ready to accept the flaws in metrics but so willing to believe that people can “qualify performance using their senses”? There are just as many flaws inherent in sensory observation as there are in “metrics.” I think it has been demonstrated (though I can’t pull data to back this up right now) that eye witness testimony is one of the least reliable kinds of evidence in a court of law.
Come on, four billion!
by Joel D on Oct 22, 2010 5:43 PM EDT up reply actions 1 recs
however, eyewitness testimony is still used. it is not very reliable but is used because it does have its own merit. It is much more useful as quantitative evidence rather than qualitative evidence. The same can be said for “scouting reports” and other subjective analysis in any sport. On its own, it holds little water. However, combined with statistics and data that support it, it can be powerful.
I teach good life choices. That’s why I almost didn’t graduate High School.
Intensive Purposes? I could care less...
your whole argument is a fallacy!
Metrics actually don’t have goals or attempt things. People do. And while there are exceptions, both in stats that are simply too hard too ignore, and with people who use numbers as tools for investigation, most of the time I think odradek has it right – they are used selectively to back up an already held notion.
That’s in the broadest human sense – baseball suffers probably less from it than say, politics.
As I typed that, I was reminded that the cutting edge of statistical political analysis is represented now by a Baseball Prospectus alum:
Metrics actually don’t have goals or attempt things
Of course, but metrics are in fact designed to do something in particular. “Hits” record how many times a batter successfully gets a hit. It doesn’t tell you anything about what kind of hit it was, where it went, or what pitch it was off of, but it turns part of something we see (in this case a partial coding of the outcome) into something we can quantify and compare (Look! Trevor Crowe had 6 more hits than Asdrubal Cabrera this season). But wait…Crowe played 25 more games than Cabrera, so we probably want some metric that will simply look at how they did while they were playing. We could look at batting average (H/AB), to do that, and see that Cabrera’s BA was actually 25 points better than Crowe’s. Again, we’ve taken something we have casually observed (plate appearances of Cabrera and Crowe with partial assessment of outcome – safe or not safe) and turned it into something we have compared. Of course, the sabermetricians have spent a lot of time trying to get metrics that better record performance, often with the goal of making better comparisons. Essentially, “who is better” or “who is more valuable”. So it turns out OBP is better than BA, and maybe something like wOBA is better than OBP.
But getting back to defense. What I see is some of Cabrera’s performances throughout the season. I see games in which Cabrera is playing at a certain place in the field, surrounded by a certain set of fielders, in a certain game situation, with a certain pitcher on the mound throwing a particular pitch to a specific batter. I also see the outcome of these at bats and some of what happens (Cabrera moves to his left, picks up the ball, and throws the runner out….Cabrera moves to his right, the ball gets by, the batter is safe). That is a lot of information, but it is very difficult information to synthesize into something reliably comparative. Humans simply aren’t good at remembering and interpreting that kind of information, and certainly not over 162 games or even a sample of 40 that we watch very closely. Nevertheless, we develop impressions about what we see and use those to form subjective judgements about the defensive quality of players.
No defensive metric that exists attempts to do the above. Most fancy defensive metrics are based on the nature of outcomes of balls in play hit to a certain part of the field relative to everyone else in the league. So Asdrubal recorded fewer outs on balls hit in his general direction than the league as a whole. That is a very different assessment of defense than what our eye sees, or even what happens in a baseball game. It can be effective, particularly if we have a lot of data and we are confident that data is recorded well…but there is a huge number of factors that metric does not attempt to control for. And there are a huge number of differences between when the metric has been designed to do, and what we do when we visually assess a player.
by APV on Oct 22, 2010 11:14 AM EDT up reply actions 7 recs
That’s a great description. Both what it “attempts” and why it falls short.
I’m not sure I completely buy into the “humans aren’t good at it” thing. Some are, and, as in a crowd, an averaging of a lot of perspectives can often be quite accurate.
What humans aren’t good at is agreeing amongst themselves what they saw. You can close your eyes in a stadium and listen to the crowd respond to pitches and get a good idea of what was a ball and what was a strike, while hearing individual arguments over same after every pitch.
I think that’s really the allure of stats. They are used to silence argument over subjective perceptions. You may continue to argue over what it means to be a great player, but when confronted by a spreadsheet listing ISO’s, you won’t argue over who is the better slugger anymore.
In any case, I take it that despite UZR, you think Droobs is a better SS than Jhnonny?
you think Droobs is a better SS than Jhnonny
I think he should be better. If we’re defining “better” as preventing runs by creating more outs, I’m not sure he has been the past two seasons compared to Jhonny in 2007-2008.
Why was this statement ridiculous the million times I said it this last year? Cabreta’s defense is overrated.
I swear this is just like what happened on this site regarding Josh Barfield a few years back.
LGT's resident moderate Yankee hating fan.
by Joe. on Oct 22, 2010 9:56 PM EDT via mobile up reply actions
I agree, Joe, but it’s easier to perceive good news than bad. The tendency is to prefer not to recognize negative indications. Kind of like a prize fight: in order to take the belt away from the champ (or, in this case,an old paradigm), you really have to make it obvious.
Crowd-sourcing is a fad. There are places where it seems to work great, and there are places where it doesn’t work at all. Eventually, I suspect, we’ll find that there are very few areas where it works reliably over the long haul. It is a fun topic to watch, but that doesn’t mean it should be taken seriously.
Key phrase from APV: “it is very difficult information to synthesize into something reliably comparative”
That is the part humans aren’t good at. There are simply too many subjective elements to sort out, and the goal isn’t just to form an opinion about a player, but to be able to COMPARE him to other players effectively. Without a RELIABLE comparative ability, the observation is a whole lot less useful.
In fact, I can still argue who is the better slugger. The models we use are very rough models of what is reality. If I take 2 sluggers with the same numbers and parse the data so that I can see how each slugger did against 1& 2 starters vs 3&4 starters, I might find that slugger A beats up on weak pitching, but slugger B hits strong and weak pitching equally well. I might find other factors that we don’t now use to rate the quality of the hitter as a team player. I can concede that these hitters are similar, but there will always be ways to refine our models so that they better mirror reality. And this would allow us to make even better decisions.
Bravo. This is exactly what I was thinking as I read a couple comments in this thread. And a perfect example of the kind of thing I could have written but am not about to.
Adam Miller is healthier than Stephen Strasburg.
Asdrubal Cabrera and his -13.4 UZR/150 follows a long line of Shapiro players
who are always chosen with offense over defense
Indians had GB pitchers like Carmona, Westbrook, Talbot and Masterson, yet their INF had a league worst UZR/150 at every position
Bravo Mark Shapiro and Kudos Larry Dolan for allowing it for the past 10 years
follow me on twitter @sternfan10
Ummm…no
Here is Jay’s summary of the scouting report on Cabrera at the time we acquired him (my emphasis added):
Signed out of Venezuela in 2002, Cabrera is considered a defensive wizard, and Baseball America named him the Mariners’ sixth-best prospect to start the season. He’s spent 2.5 seasons in the minors at shortstop and one at second base, dazzling scouts at both positions and always at or near the top in fielding percentage. The rushed development and position flips have been the result of a crowded picture in Seattle’s middle infield; Cabrera has been progressing through the minors alongside fellow shortstop prospect Adam Jones, just four months older, and the Mariners signed Cuban defector Yuniesky Betancourt as a major-league shortstop in the offseason.
At age 17, Cabrera was an All-Star shortstop in the Venezuelan Summer League, and he made another All-Star team in the low-A Northwest League despite being the league’s youngest starter at 18. While Cabrera’s recent numbers suggest a weak bat, on any given day, only a handful of 20-year-olds are playing above Single-A, anywhere in the country. This season, Cabrera has been striking out once every four at bats, but at lower (more age-typical) levels, his peripherals were solid. Notably, he has maintained a .10 walk rate since the start of 2005, even as he was promoted four levels in one year. Cabrera becomes the youngest player on the Buffalo roster by more than two years, and one of only three Bisons under 25.
The Indians have been painfully bereft of solid middle-infield prospects since Jhonny Peralta and Brandon Phillips ascended to the majors. Acquiring Cabrera addresses this problem but also seems rather pointedly aimed at the Indians’ infield defense issues. That said, it seems unlikely the Indians would deem Cabrera a solid contender for a starting job in 2007; he seems unfinished. That said, the Indians face several middle infield problems for 2007 — including Ron Belliard’s likely departure and Peralta’s newfound defensive yips — and Cabrera might reasonably solve any number of them.
Go back and look up scouting reports of Cabrera in the minors and you’ll see the above summary is correct.
I'm sure Indians pitchers feel so much better now
knowing that Asdrubal as an all star in the VSL, as he allows another ball in the hole or behind 2B to get through for a base hit
follow me on twitter @sternfan10
follows a long line of Shapiro players who are always chosen with offense over defense
This is what you said. And it is wrong. Cabrera was chosen in spite of his offensive performance to date (though not without some optimism in his forecast), and because of his ability to play the position well defensively.
I am not a Cabrera fan, but his offensive season of '09 is why he was handed the job last year
Even before his injury he had regressed his BB% and wasn’t able to correct it all year
If i’m the Indians i make a deal for a SS, put Cabrera at 2B where his limited range won’t hurt
follow me on twitter @sternfan10
And? What’s your point?
Of course he was handed a job based on being one of teh better shortstops in baseball last year. Do you think some other front office wouldn’t have handed him the job?
Asdrubal’s walk rate over the first six weeks was cause for concern, but he’s had great walk rates for years. It is alarmist and frankly a little crazy to assume this is a long-term problem.
Seriously? We can’t even discuss defense and defensive stats without taking a shot at Larry Dolan?
by Brad D on Oct 22, 2010 2:16 PM EDT up reply actions 3 recs

by 
















