Hanley, Jacobs, Defense
Royals skipper Trey Hillman recently spoke about the defense of his new first baseman, Mike Jacobs, subtly hinting that the numbers deeming him poor are not entirely accurate. Jacobs, according to Hillman, is very good at handling the 3-1 plays in which the first baseman lobs the ball to the pitcher covering the base, which apparently makes him much more of a solid fielder than meets the eye. Whenever I read quotes like this, or stories about players discussing the shortcomings of defensive metrics, I get a bit annoyed because it seems that those doing the talking have a complete lack of understanding of how the metrics work. Hillman may be an innocent bystander in all of this, but his words, for whatever reason, tipped me over the top.
Mike Jacobs, over the last three seasons, has the following UZR numbers: -3.2, -4.1, -11.1. His Dewan + – numbers further confirm the fielding ineptitude. But when managers or players hear numbers like this, they automatically assume it refers to errors made, which is utterly incorrect. Let’s look further at Jacobs, shall we? Thankfully, UZR breaks down where the runs saved/allowed come from, and a more granular analysis at Jacobs shows that he is quite adept in certain areas and equally awful in others.
Jacobs’ double play stats averaged around -0.4 runs/season, placing him right around average. His error runs, which show the runs saved via not making errors relative to the average player at the position, average out to +0.5 runs/season. Essentially, those defending Jacobs because he does not make many errors are correct, but not making errors does not automatically result in a solid fielder. The issue with Mike Jacobs is range, and the fact that he, well, does not have any. Over the last three years, his range has cost the team an average of -6.0 runs.
Put everything together and a player with an average -6.1 runs/season on defense emerges. Certainly not the worst fielder in the history of the sport, but by no means an effective one. Jacobs looks solid in terms of double plays and limiting errors, but his lack of range prevents him from reaching balls that other first basemen can glove. So, just because Jacobs has the reputation of being a poor fielder does not necessarily mean he stinks in every aspect of fielding, and along similar lines, the areas in which he performs well, though important to the mainstream, are not the sole barometers for defensive prowess.
Jacobs’ former teammate, Hanley Ramirez, is another interesting defensive case study, as HanRam has the reputation for being a Jeter-esque fielder despite two league average or slightly worse seasons out of three in the big leagues. That’s right, in 2006 and 2008, Hanley the Manly put up UZR marks of -5.5 and -0.3, respectively, which are much better than the numbers posted by Jacobs given the difficulty of the position. Via positional adjustments, remember that, assuming 162 games for each player, an SS with a -7.5 run defensive mark is equal in overall fielding value to a first baseman with a +12.5 defensive mark, because 1B is very easy to play relative to captaining the infield.
What earned Ramirez the reputation was his -19 run performance in 2007, but getting granular once more shows an interesting tell. Ramirez has been above average in ability to turn double plays in all three of his seasons, and his error runs are not necessarily the ultimate cause of his defensive downfall either, as the -7.3 ErrR in 2007 was not significantly higher than the -5.8 ErrR in 2006. Look at his range runs, though: -1.1, -13.3, +2.5.
In both 2006 and 2008, Ramirez was virtually league average in terms of range, but he lost everything in 2007. The three years may not be a large enough sample from which to draw any conclusions, but my gut instinct tells me that the 2007 season is more fluky than indicative of his true talent level. We have two players here, one of whom is a poor fielder that elicits defense from fans based on an ignorance of what actually goes into evaluating fielding, and another who has garnered a reputation in the mainstream for one really bad year that may very well go down as a fluke among flukes. The bottom line is that more goes into fielding valuations than simply the “ability” to not make errors, and understanding these components is a must in order to ever have intellectual discourse regarding the subject.
Print This Post
I missed the part where Hillman said that the metrics are all about errors. From what I read of the article, he said Jacobs has been better defensively SO FAR IN CAMP. I see nothing worth criticizing there. Perhaps I am misremembering the article though, I’ll go read it again.
Went and re-read the article in the KCStar……yeah, he was just talking about Mike Jacobs improving during the course of ST camp. I don’t know how that equates to Hillman completely botching what defensive metrics are about. I’m no huge Hillman fan and I am not claiming he does perfectly understand them, but I think criticizing him for what he said here is a pretty big stretch. Of course he’s not gonna come out and openly say a player sucks. He just said that he thinks Jacobs is getting better. No harm there. Billy Butler will probably see more time at 1B anyways.
Perhaps I took out my frustration for what every KC fanboy and media member says about Jacobs on Hillman, but the points still stand. Everyone who defends Jacobs does not grasp how defensive metrics are calculated, and there is much more than errors that go into them.
I can agree with that. Anyone trying to defend Jacob’s defense is just wrong. He’s awful, and even if he really is improving, he’s still likely to be below average. The Royals would be best served making him a DH and limiting his ABs against LHPs. Fortunately all they gave up to get him was an injury-prone reliever with questional strikeout rates. Don’t love the deal, don’t hate the deal…as long as he’s used right.
Yeah, you definitely misread the article. Huge swing and a miss by you. It’s a bit sad that you are allowed to post articles on this website when you: (a) do not have the reading comprehension to understand what point a quote, as well as an entire article, is trying to make, and (b) turn around and claim that “every KC fanboy and media member” are saying that Jacobs is better than what the metrics state.
It’s a terrible logical fallacy to claim that every single KC fan and media member are saying this, making you look unbelievably unprofessional at best, and downright biased and spiteful at worst.
This is a great site overall, but petty, childish posts like yours can really drag it down sometimes.
Mentioning Hanley, Uggla had a similar trend of being slightly below average in 2006 & 2008, but terrible in 2007. Maybe it was a small(? – 1 season) sample case of a couple o pitchers getting hit hard and the fielders getting blamed in the metrics.
And Miguel Caberera at 3B:
2006: -4
2007: -8
2008: -4
It is very relevant that you point out what Cabrera was doing in relation to Ramirez’s numbers at shortstop. However, whatever he did in 2008 didn’t affect Ramirez because Cabrera was playing for the Tigers. Cantu was the third baseman for the Marlins and he had much more range than did Cabrera at third in 2007 or even in 2006. This helped to lower Ramirez’s overall numbers as well in 2008. Ramirez is a below average SS, but his 2007 numbers were adversely affected by Cabrera’s horrendous fielding season at third.
“very good at handling the 3-1 plays in which the first baseman lobs the ball to the pitcher covering the base”
Seriously? This is his defensive skill? While I’ve not done a formal study on the matter, I’m pretty sure turning the 3-1 is not a big feat and probably has more to do with the pitcher than the 1B. Also, if he fields fewer GBs in a 3-1 situations, I doubt he does much to make up for this.
However, I think we should all be rejoicing at the level of attention that advanced defensive metrics are getting in the recent press. Okay, it’s not very much attention. But the fact that people are attending is a good thing.
how long before i start patrolling CF for the red sox?
on a serious note, do you have defensive stats dating back 20+ years? I never saw him play but how does someone like robin yount win the SS GG in 82 then play CF 2 years later.
in honor of one of my favorite players rey ordonez, a cool article idea might be to do a list of the best single season fielders who had the worst offensive seasons to see how many runs their gloved “saved”
Rally has TotalZone at his WAR leaderboard.
http://www.baseballprojection.com/war/y/younr001.htm
For reference, Yount moved to the outfield in 1985; the variation in positional adjustment is because of both playing time and DH time.
Does UZR account for runs saved/lost by first basemen based on their ability to receive a throw? In other words, will we see a difference in the UZR ratings of a first basemen who is great at scooping errant throws and one who is not so good? If so, what part of the rating does this go into – range? It generally wouldn’t classify under errors, as the error on these plays almost always goes to the thrower. Given that this is essentially a unique aspect of the first baseman’s job as a fielder, I’m curious as to whether UZR and/or other defensive metrics take it into account. It’s certainly an important part of the picture in analyzing defense at first.
I believe MGL did a study on this, although I can’t recall if it is currently in UZR or not.
Thanks for the heads-up; I found the article in question here: http://www.fangraphs.com/blogs/index.php/first-basemen-scoops.
Short answer – scoops are not included in UZR, and play a minor but not insignificant role in quantifying first base defense – I think MGL estimates it at around 1/4 of the overall equation for them.
While I agree to what the article is saying, what’s a manager supposed to do, throw his guy under the bus?
If a guy stinks, a good manager plays a better player, not bash the crappy guy to the media. And right now, what better alternatives does KC have at 1st?
That being said, Jacobs does kind of stink and will until he can learn how to get on base at better than his career .318 clip. Willy Taveras is a career .331 for Christ sake.
I don’t have faith in UZR yet. It seems that players can wildly swing from year to year with their range ratings. However, it seems to me that range would stay relatively constant. Say your real skill level is a +5 UZR per season. I don’t see you fluctuating much past say -1 and 11 in any particular year. But look at Hanley, he put up scores of -6, -21, -.3. I doubt very much it was that big of a swing per year. Or look at Juan Pierre. He is like +15, +15, 0, 0 ,0 +15. The statistic just seems to0 variable at this point in time to accurately reflect true fielding ability.
More examples include:
Pujols, AROD, Jeter, Orlando Hudson, and basically everyone else in the league
This is one of my pet peeves about the problems with defensive stats. The key to remember is that they are STATS describing what DID happen. They are not meant to tell you a player’s true talent. Offensive stats vary wildly. Does anybody take issue with them on that basis? It’s called statistical variation. In 1980, Ted Simmons had a 141 OPS+. He followed that up with an 87, and then up to 112 in 1982. Does that mean that we shouldn’t have faith in OPS+? Or does it mean that Ted Simmons had an off-year offensively? And isn’t it entirely possible that players have off-years defensively, as well as fluke good years defensively?
Yes, offensive stats do vary just like UZR can, but with offensive metrics, you can explain the variations. Things such as an increase in BABIP, HR/FB rate, LD%, or various other stats can explain why the player varies offensively from year to year. With defensive stats, you can’t, at least not to the same degree.
I think defensive metrics have a LONG way to to go before they can be as legitimate and stable as offensive metrics. Doesn’t mean they are completely useless, but I’m not sure how much stock you can put in them.
I understand that that their will be variation, but defense seems a lot more controllable than offensive and, therefore, should not have violent swings every year. Like I said, I could see a player being +- 6 from a mean every year, but it seems like the stat is +- 9. I don’t think it accurately captures what actually happened during the season.
What makes it “seem” that defense should be a lot more controllable? If anything, I’d expect more variation from year-to-year in defense than in offense.
We can’t always explain the variations like we can with offensive stats, but that doesn’t make “They vary too much” a legitimate criticism of defensive stats.
I agree, and it can wildly swing a player’s value a lot, apparently even more so than hitting.
Even though I think UZR has value as a multi-year average, which is how I always use it, rather than look at one year of work.
Yes, from everything I’ve been reading it takes about 1000 innings to get a pretty good gauge of UZR.
If I remember correctly Yount suffered a shoulder injury. It was thought that playing him in CF would be less stressful on his shoulder than SS… a lot fewer throws. He really was a terrific player.
“The issue with Mike Jacobs is range, and the fact that he, well, does not have any. Over the last three years, his range has cost the team an average of -6.0 runs.”
Not to be picky, but that’s worded wrong. If he “cost his team -6 runs” then he boosted his team by 6 runs.
If this were Sports Illustrated I could accept characterizing Ramirez and Jeter as similar, but we expect a little more insight. Jeter is extraordinarily reliable (sure-handed) on balls he gets to (which isn’t very many) and his range is a disaster. Ramirez has league average range, but gives a lot back with his inconsistency, lack of concentration and poor plays on routine balls.
He is not in any way characterizing Ramirez and Jeter as similar. He just mentioned that Hanley has a reputation for being among the worst defensive shortstops in the game, which would be Jeter-esque.
This was the situation with Raul Ibanez. Many of his supporters claimed that Raul wasn’t as bad of a fielder as many claimed because he had a respectable fielding percentage. Forget that the guy had zero range.
Whereas I can’t dispute the conclusions based upon your statistics, I do watch nearly all Marlin games. No, Jacobs is no Keith Hernandez. However it should be noted (and isn’t) that he played hurt all through 2007. He had a bad ankle but played anyway. In 2008, he hurt his wrist and not only missed significant time, but wasn’t up to par for a period when he returned. He’s a little better than the statistics say.
As for Ramiriz, many errors occur because he can’t get the ball out of his glove. It cost the Marlins a game when he didn’t record the last out because of this. Personally, I think the guy’s glove is too darn big for a shortstop.
I just wanted to chime in, because I watch nearly every Marlins game and wanted to say from what I’ve seen Jacobs is absolutely just as bad as the statistics say.
I agree Hanley has diificulties with getting the ball out of his glove, and also with accuracy on throws. However it seems like he doesn’t take easier plays seriously sometimes and that hinders his overall performance. I still love Hanley even if he is slightly below average as a SS