ogc thoughts
After a very poor defensive year last season - only 4 DRS, not even an extra win (10 runs), after years of sterling defense - I was wondering how this year would fare, given that Sandoval being fit that one year, in 2011, resulted in him leading the league by a mile in DRS at 3B. I'm happy to report, we are already at 6 DRS just 16 games into the season, meaning we are already ahead of what we compiled in total in all of last season.
Here were the biggest offenders last season, per DRS (oh, that's Defensive Runs Saved), that sunk us:
- Abreu: -5 (that's another reason why I was pretty sure Adrianza had the last spot vs. him)
- Lincecum: -8 (and, of course, that's just in his 32 starts)
- Pagan: -9 (that's why the Giants should be looking hard for a CF, and why Brown probably don't have to do a lot offensively to start in CF)
- Pence: -7 (I was surprised by this, but might be a quirk of the system as his Rtot was positive, as well as UZR/150; I might average the 3 in the future, but for now I like DRS more)
- Sandoval: -5 (I wish I could see how this changed during season, if he improved late, when he started his brother as chef phase)
- Scutaro: -7 (we heard about all his difficulties last season...)
- Belt: +4
- Blanco: +10
- Bumgarner: +5
- Perez: +10
- Posey: +4
This season, got a mix again, but most negatives are low, and got a couple of big contributors. Here are the negatives: Arias (-2), Belt (-1), Crawford (-2), Hicks (-1), Morse (-1), Petit (-1). It is early season, yes.
Positives so far:
- Adrianza: 1 (38 DRS/yr, as an FYI)
- Casilla: 3
- Pence: 5
- Posey: 2
- Sandoval: 2 (17 DRS/yr, he was around 25 in 2011, if I recall right)
- Vogelsong: 1
Early, but some interesting things so far that I would want to point out. Alphabetically:
Adrianza is among the positives, in just 5 games played, 3 starts, 31.2 innings or about 3.5 games. He did that last year too, in 6 games played, 4 starts, 45.0 innings or 5 games, and had a 27 DRS/yr rate (but that was for SS, while this year for 2B). Not as great by UZR, 5.5 UZR/150 last season at SS, but 18.0 UZR/150 this season at 2B, or roughly 9 UZR/150 over the two seasons.
What this means is that he doesn't have to hit for much for his superior defense at 2B and SS to be much more valuable MI than Arias (poor on defense everywhere but 3B for his career) or Hicks (poor in MI, average in CI). Both costs the team at a -10 to -20 seasonal rate there, if Adrianza is +10 seasonal rate, that's a 2 to 3 wins swing defensively between Adrianza and Arias and Hicks. It's no contest if Adrianza is as good as SSS for DRS. And I still believe that Adrianza will eventually hit for us, making his defensive value all the better.
Hicks is going to need to hit at his high OPS to be more valuable than Adrianza. And after his hot start in his first four games, in the 8 games since (7 starts), he's been hitting only .158/.407/.158/.565 since then, with a horrible 10 K's in 19 AB's (but 7 walks for that hefty OBP), so it appears to be just a matter of time before Arias and Adrianza start getting more starts at 2B. Of course, Ehire's 3 hit game probably earned him an extra start or two, starting with tonight's game.
Pence has thrown out a number of runners so far on the bases, compiling three assists already. The Giants have done well in recent seasons in turning defensive duds like Burrell and Huff to be productive in the OF, but Pence has not changed as dramatically, he was basically as bad defensively in 2013 as he was in 2012, per DRS. Given how hard RF is to field because of the oddities and quirks (vs. LF for the water buffaloes), perhaps it took Pence a little longer to get the hang of playing there. Also, as I noted, the other advanced defensive metrics didn't have him as bad there, so it could be a quirk of DRS that don't recognize stuff that Pence did.
In any case, hopefully he can continue at a good rate there, great start there so far, he could coast and be average the rest of the way and still end up one win better defensively than last season, earning his raise just from that (he made roughly $13M last season, getting $18M this season, roughly one win increase in salary). And he has five more years of playing there, so learning now is a good time, contract-wise, for the Giants.
Posey when he came up and before he was injured, was a defensive god, a man among men, running up at least one win above average on a seasonal basis (since he only played parts of both seasons), but after his injury, in 2012 and 2013, he merely slightly above average at best. So it is good to see him again among the leaders in defense.
Most probably the injury affected his defense a lot more that couldn't be fixed in one off-season or two, or perhaps it set him back not playing for nearly a year while rehabbing, or a combo of all that. In any case, he had a total of only 4 DRS in 2012-2013 (0 in 2012, 4 in 2013) and already has 2 DRS so far this season in 13 starts, 119.0 innings. It would be great to have him back as a defensive stalwart on the team and so far, so good.
There was also an article noting how he is throwing out more base stealers this season. That is probably a combination of his improvement as well as the pitchers, who after last year's face plant, felt the urge to do anything, everything, to improve themselves to win this season. Still, every last bit helps the team win more games.
Sandoval has only been really good at 3B in two seasons before 2014, both where he was in reasonably good shape: 2008, his first season in the majors, and 2011, when Operation Panda got him in good shape for the season. He had a 16 DRS/yr in 2008, 20 DRS/yr in 2011, otherwise, has been either pretty negative (2009, 2012, 2013 he had at least a -5 DRS) or about average (2010, with a 1 DRS for the season). This season, with his 2 DRS, his seasonal rate is 17 DRS/yr, right in line with his two good seasons. Perhaps the Panda is back!
I know his hitting hasn't, but given all the walks he's been taking, I think he's still working out his batting mechanics in order to take more walks, probably something he's been working on since spring training, maybe at the behest of Barry Bonds, who knows why after all these years he suddenly realized that he should take more walks. I think he'll be fine once he gets that into his muscle memory. But yeah, it's frustrating to see him hitting like this from the 3-hole. Maybe he should bat 7th for a while and reward Crawford for his good hitting by moving him up into the top of the lineup. Bochy had talked about hitting him second, so maybe the Belt/Pence experiment there can be moved to batting third.
Defensively, we have missed the big contributions from other players, like Schierholtz, Stewart, and Ishikawa, who provided stellar defense for us at their positions, even though they were never truly full-time starters. The team for a number of years were among the top teams in defense, never the top, but generally among the leaders, but last year was a huge decline defensively, hurt as you can see from above, from a number of key players.
I covered the pluses, but I would be remiss to not note that Pagan, who was very negative last season, is currently at 0, which is both a huge improvement as well as a good area for him to be in, as he has been strongly in the negative for CF since 2011. He must have had a pretty bad injury to go from being a great defensive CF in 2009-10 to a very poor one in 2011-13. It was pretty much a 3 win swing, a huge decline. By just being average, Pagan would have improved his defense by a win. That with him hitting the way he has during his last few years (don't count on him hitting .900 OPS :^) should more than earn his $10M salary, he'll be easily over 2 WAR he needs to do to earn his salary, his hitting has always been very good for a CF, it is his fielding that has held him back the past few years, from being among the good players in the league.
I haven't looked at this blog in over 2 years, back then you wouldn't stop discussing Ishikawa. I check in today, mentioned again to help make some point.
ReplyDeleteAn idly pointless comment from an idly vagrant reader making his sort-of biennial drop-in! What would be to the point here, for others besides stray hecklers, is the comparative strengths of the defensive metrics, DRS vs UZR, for example. Should one assume that DRS is a leveraged rating, reflecting the game situation in which a play occurs, whereas UZR isn't? Is that why you choose to look at DRS, ogc, rather than more abstract metrics?
Delete"Idly pointless" means Anonymous's crack, obviously, not o/c's (as always) helpful, reasoned analysis.
DeleteActually, it's been more like 4 to 6 years and the reason I wrote about Ishikawa was because his struggles at Dodd Stadium illustrated perfectly the struggles our hitting prospects were having in AA, plus brought up the ugly fact that our pitching prospects in AA were nowhere near as good as they appeared to be there.
DeletePlus, to me, more importantly, this quirk of the park might cause hitters and pitchers to change the way they did things successfully in the past in order to be successful in Dodd, which would most likely lead them to be unsuccessful when they get to AAA or the majors.
But if you don't think those are important things to worry about, then you are probably right not to read my blog, I only care to write about things that affect the success of the Giants team. If you don't care about those things, then you don't really care about the Giants success.
campanari, always with the insightful questions.
DeleteAdmittedly, part of it was laziness: BB-Ref has the DRS stats for the team easily accessible on their team page. Instant Post!
Luckily for me, I like the DRS stat. UZR seems to be much more of a black box type of stat, punching in numbers, spitting out numbers, I kind of understand the number and kind of don't. DRS is produced by the company that publishes the Fielding Bible, and the founder of the company works closely with Bill James. It involves people viewing each MLB BIP and rating it based on their Plus/Minus system. It is also an evolving system, as new technology is added or new concepts (it is incorporating techniques for valuing a catcher's DRS, something that I don't think that UZR will ever be able to measure very well, another reason I hew to DRS).
Not that black box scares me - I'm a former programmer, and a sabermetrician - just that I like the Plus/Minus system more, even though there is that subjective aspect to it since humans are rating each BIP. It is a system covering the different aspects of each fielding position, whereas UZR focuses only on productivity within zones, from what I understand. It is a system that is evolving with technology; UZR probably will evolve with Field/FX being added, but, again, I would note that it will probably always be lacking relative to DRS with regards to catchers.
Neither rating is leveraged, that I'm aware of, for the most part. It is based on making the plays that a player should, plus giving him credit or not (subjective) for good/bad plays that end up as an out. Each system does it in its own way.
I believe that DRS is leveraged a bit in that when an OF steals away a homerun, that adds to his runs saved, though it should not include any runners on base. I tried to read up on it, but did not find anything related, so I could be wrong on this.
Total hilarity. The Giants just signed Ishi to a minor league deal. Go get a handle already you anonymous coward.
DeleteI watch the games, I see the defense, and I think many of the defensive metrics around, are just terrible. Seriously flawed. Belt seems to me to be a very good defensive first baseman, yet I have checked with some defensive ratings, and see Belt below replacement. What is this? Hicks looked terrible for a few games, and his range looked bad and his arm looked bad, but the last few games he looks like a new man. I wonder if metric's change significantly throughout the year. Pagan has looked better in center so far this year than he has since he came to the giants. Morse deserves a negative fielding metric, otherwise the team should be OK.
ReplyDeleteI can certainly understand your position and agree that to a strong extent, they are terrible. How can one system purport to show one player as a great defensive player and another system rates him as only average or below average? That has bugged me a lot too.
DeleteBut that's how things are early in an evolution of any metric. It takes time and adjustments to get any tool to work the way you want it. And sometimes it has to evolve with the times as well. Like with accounting itself.
At this point, while there is a lot of question marks about which is better and whether either is entirely valid or not, I think it is worthwhile to see what the system is saying about defense and about who are good defensively.
It is not like the system is totally off on an overall basis, teams that are viewed as good defensively show up as leaders (like the Rays), and players who are viewed as good generally show up on the leaders list as well. I think that shows that there is merit to all the systems, but that when some players don't pass the smell test, then that shows that the system still is lacking in some way. That's why I like the DRS system, it has been adding techniques and data each year that improves the system, while I don't think UZR is changing much, season to season. Plus, UZR is currently severely lacking in terms of rating catchers, which is very important to any team's defense, and based on the way it works, I don't think that it can ever make up for that lacking.
Per data right now that don't make sense, part of the problem right now is that most positions have not racked up a lot of chances yet, and thus a good or bad play, here or there, can have a significant effect on a players defensive worth, so early in the season.
Or maybe a player is battling some physical ailment that he can play with but hampers his defense some? Like Scutaro and his back last season.
And yeah, I'm sure that metrics have changed significantly during the season, though I have no specific example since the database is updated daily but no splits are available. I think that's the nature of that particular data beast.
But sometimes your view matches the data. Pagan was horrible last season per DRS but much improved this season. Maybe that leg problem that finally fell him after his homerun dash was just the final straw that knocked him down, but was hampering him up to then (I would note how poorly he had hit up to then as well, perhaps he wasn't getting the infield hits he normally would get with his speed). Or maybe multiple seasons, he went from day to night with his defense while with the Mets, maybe a nagging injury had been bothering him for years which he didn't really take care of until his DL last season. Will be interesting to watch what he does.
So I see your point, but at some point you need to use something as a way of comparison, even if there is still some warts in the system, because otherwise, you can never reach the point where you have THE perfect system, you need these work-in-progress metrics now to ever reach a system that evaluates defense in a proper way, because of the complexity involved in rating defense, period, then trying to create nuance by positional differences.
Oh, I should add, terrible on an overall basis, as there are wide disparities for too many players for anyone to feel that we are there yet in terms of defensive metrics. We are still in the BA era of defensive metrics, it will take a ways to reach the OPS era, and a while more to get to wOBA.
DeleteThanks for the DRS analysis OGC. I think its a nice complement to "our eyes" and the various dWARs. It is a lot more subjective.
ReplyDeleteI hope you do one every month, to see what kind of progress is happening throughout the year.
That's a great idea, perhaps around the same time I do my PQS analysis. I was going to revisit it anyway because of my interest in how Pablo is doing relative to 2011, which I'm using as a gauge of how fit he is, but had no timetable for that.
DeleteI was also interested in how Adrianza was doing, as that would be a huge component of his value and whether he would be kept by the Giants or not on the bench.
This also got me interested in how Pence is doing, since he had not done so well before, which didn't make a lot of sense to me, he was hugely positive defensively early in his career, he still has some speed, the Giants are known for positioning their players well batter by batter, so one would think he would be improved some, and that seems to be happening this season.
Another good reason to do it on a regular basis is because BB-Ref does not provide any type of splits on the defensive stats, so one cannot track the ups and downs of any player over a season. And I understand a bit why they don't, a player's value could shift up and down based on how all the players handle their position, as the idea is to center the overall rating to be 0 as average for a position, without the player getting substantially better or worse. Still, I think looking at this could be informative. So this proposed series could be a chronicle of those ups and downs.
I'll try to remember, give me a kick in the rear should I forget...
And yes, it is a nice complement to all the other ways defense is being measured. DRS is a nice blending of subjective and objective, whereas UZR is purely objective, and our eyes is almostly purely subjective with no overarching standards across the eyes. At this point, we need all the various measures to have a good overall view of any player today, as each measure, I'm sure, has its strengths and weaknesses in terms of evaluating any player's defensive abilities.
DeleteSo if I can find a way to get the UZR data in a team table, I'll include that in my future posts on defense, in addition to DRS. Plus I'll include the other defensive metric that BB-Ref provides, which gives a very rough idea of defensive worth, as it is designed to provide a defensive value to players back a hundred years ago where we had very minimal information available defensively, until recent years. See all of them would be interesting to contrast and compare.