Curtis Granderson’s Defense and His MVP Prospects

As expected, the campaign has begun in New York. Even prior to his inside the park home run on Sunday, Yankees scribes have started penning the case for their hometown man, Curtis Granderson, to win the AL MVP Award. The case makes plenty of sense from an old school perspective. Granderson is the best player on a playoff-bound team, and has generally outproduced his fellow playoff-bound peers at the plate. With 35 homers he trails only Jose Bautista, who won’t sniff the postseason. That he leads the league in runs and RBI furthers his case among those who actually vote for the award.

The statistically inclined audience tends to ignore most of the above factors. There are plenty of other issues at stake, such as how many runs the player created irrespective of his teammates. There’s also defense. That’s why WAR is often the place a statistically inclined fan will start the MVP conversation. Granderson doesn’t fare as well here, ranking fifth in the AL with 6.1 WAR. Worse, he trails four players in his own division. But WAR does contain a one-year sample of UZR, and we know that one year of UZR can provide misleading results. Is this the case for Granderson?

In The Wall Street Journal today (subscription required), Dan Barbarisi writes about the Yankees’ stellar outfield. It focuses mostly on Brett Gardner, whose UZR, currently 19.0, leads all competitors by a significant margin. Nick Swisher has dramatically improved in terms of UZR this year, producing more defensive runs above average in 2011 than he has in his entire career to this point. Granderson, however, has suffered at the hands of UZR. While he has fared well enough in the past, this year he has dropped all the way to -9.2, his worst mark since 2008.

The nearly one-win ding on defense has certainly cost Granderson on the WAR leader boards. As The Star-Ledger’s Marc Carig wrote later in the morning, no player in the AL MVP race has been more adversely affected than Granderson when it comes to defense. Carig removed defense from the equation entirely, while preserving the other WAR factors: hitting and base running, plus the positional and replacement level adjustments. With defense flattened, Granderson sits behind only Bautista and leads the other contenders by a significant margin.

Why would anyone want to cut defense out of the equation? After all, it plays a role in a player’s overall value, so it seems a folly to omit it from MVP considerations. For starters, there is the sampling issue of UZR. If, as the maxim goes, we need three years of UZR data to properly evaluate a player, then why do we include single-year UZR in WAR? The other issue relates to Granderson’s actual value on defense. While many scouts and observers will confirm the poor defensive abilities of other UZR trailers, such as Raul Ibanez and Lance Berkman, fewer, if any, will find significant flaws with Granderson’s defense. He might not be the best center fielder in the league, but he’s also not the worst, as 2011 UZR tells it. This comes both from the eye test and from multi-year UZR.

One interesting issue Barbarisi raises is that of positioning. The following paragraph is particularly telling.

The way Gardner covers ground allows the Yankees to use different defensive alignments, shifting Granderson more toward right field in some situations because they assume Gardner can cover all of left-center.

Despite the shift, Granderson is responsible for the same zones as all other center fielders. While it’s true that he’s not debited for plays that Gardner makes in Granderson’s zones, Granderson also does not get the credit for those plays. Since UZR compares players on a positional basis, Granderson’s low UZR might simply be the product of him not making the same number of plays as his fellow center fielders. Since poor left fielders flank many of them, they have more opportunities to improve their UZR scores by making plays in the left fielder’s zone. Granderson has no such opportunity. Gardner is responsible for those zones, and he typically makes the plays.

This issue cropped up in the recent past with Andrew McCutchen. When John Russell managed the Pirates he positioned the Pirates in a way that would take away the bigger part of the ballpark. It’s debatable whether this odd shift was effective on the whole, but it certainly took a toll on the outfielders’ UZR scores. This was particularly perplexing when it came to McCutchen, who, by way of every scouting report ever written about him, was a superb defender in center field. Yet in his first two years in the majors he produced negative UZR scores: -1.3 and -13.7. The Pirates replaced Russell with Clint Hurdle this off-season, who presumably ceased the shift. The result: McCutchen has produced a 7.0 UZR this year, which is a bit more in line with expectations.

If positioning does play into Granderson’s heavily negative UZR, then how should that be handled in terms of WAR? Should we actually penalize him for an effective team strategy? After all, the Yankees’ outfield ranks third in the majors in UZR, trailing only the Diamondbacks and the Red Sox. Shouldn’t there be some way of adjusting for this, especially as WAR plays an increasing role among MVP voters?

The truth is that no single stat can perfectly relate a player’s on-field value. While the WAR framework is strong, it’s only as good as the inputs. If UZR is inaccurately assessing Granderson, due to unique positioning, then it won’t properly rate him among his peers. As with every player evaluation issue, we should look beyond the surface and find the truth in every area we can. And who knows: maybe UZR does have it right on Granderson. Maybe he has declined precipitously this year. But with all the questions surrounding that assessment, including observations to the contrary and Granderson’s superb flanker, it’s likely that we have to look beyond WAR to get at true grasp of Granderson’s place in the MVP conversation.

(That said, Jose Bautista, at this moment in time, is the AL MVP. Just to make that clear.)




Print This Post



Joe also writes about the Yankees at River Ave. Blues.


108 Responses to “Curtis Granderson’s Defense and His MVP Prospects”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. everdiso says:

    Given that Granderson has always been a pretty good defender, taking away from offensive accomplishments this year because of some probably outlier defensive numbers seems pretty silly.

    He doesn’t deserve to win over Joey Bats, but he’s got a much better case than any of those Red Socks.

    +12 Vote -1 Vote +1

  2. Jose says:

    Even as a Yankee fan, if Bautista doesnt win I’ll riot

    +5 Vote -1 Vote +1

    • rbenchley says:

      While I agree with your evaluation of the AL MVP race and admire your dedication to reason and rationality over blind loyalty to your favorite team, I’m not sure how much comfort that will be when you’re beaten to death with souvenir Jeter bobbleheads.

      +12 Vote -1 Vote +1

  3. Mike B. says:

    This argument brings up a worthwhile point–should there be a more in-depth review of each outfielder’s “patrol zone”? And what about 3B, SS, and 2B? Some teams use infield shifts far more than others, as well as creative positioning so that superior defenders cover more territory. I’m not sure how much this sort of thing is covered when measuring advanced fielding metrics.

    As for the AL MVP race, the rest of the season should prove fun. As a big C-Grand fan part of me wants to see him go on a hot streak and win the award, which he probably should have won back in ’07. On the other hand, Bautista is just amazing and I have to root for him as well–would like to see him go on another tear (preferably against the Yanks and BoSox).

    Vote -1 Vote +1

  4. Pelly says:

    Granderson is a really exciting player and just a cool dude. I think it’s very possible he wins MVP this year, and with apologies to Bautista, I think it’s a respectable choice.

    Vote -1 Vote +1

  5. Hanzel says:

    FWIW (which is admittedly not much)

    Bats
    Granderson
    Ellsbury
    Pedroia
    Cabrera
    Verlander
    Gonzalez

    Vote -1 Vote +1

  6. Nat says:

    I’m not convinced that the shift in the outfield is what’s giving Granderson his poor fielding numbers. A quick glance at his Defensive Runs Saved shows that he’s at -12 runs this year. DRS is calculated differently from UZR in that it does not use zones, so one would expect that Granderson’s numbers wouldn’t be as negatively impacted, but instead his numbers are slightly worse.

    +12 Vote -1 Vote +1

    • Welp says:

      Don’t understand why this was omitted.

      Vote -1 Vote +1

      • Telo says:

        Because the author is mediocre.

        Vote -1 Vote +1

      • Sultan of Schwwingg says:

        Worse, author is a Yankees fan.

        I hope you all have a nice trip to nowhere; Joey isn’t going to win it. Maybe you all should meet after the fact and hug.

        My vote today is for the Grandy-man.

        Vote -1 Vote +1

    • Richie says:

      If Gardner is taking balls centerfielder Granderson would normally catch, wouldn’t that negatively (and again, falsely) affect his DRS also?

      Vote -1 Vote +1

      • Nat says:

        The answer to that question I think lies deep in the methodology of DRS. The question I have is: If a ball is caught by the right fielder that COULD HAVE been caught by the center fielder, does the center fielder get a negative credit for that? To me, that wouldn’t make sense, but until we know the answer to that question it’s hard to make a judgment about Granderson’s DRS numbers.

        Vote -1 Vote +1

      • Greg says:

        Does DRS account for initial fielder position when the ball is struck? If not, then the shift would still show up in DRS. The question isn’t whether Gardner catching balls in traditional centerfielder territory negatively affects Granderson. (But those outs DON’T positively affect Granderson like they would a normal centerfielder.) The question is instead how Granderson’s postitioning would affect the plays he makes in traditionally safer centerfielder areas. Left-Center will play like far-left-center if Granderson is positioned towards right. This would skew his numbers all over the board.

        Now if he’s making up for it by making plays further into right field, then it should be a wash. But if Swisher is making those plays, then Granderson is essentially squeezed out of tons of should-be-makeable plays while lowering his expected play-making rates in traditionally safer zones or areas.

        Again though, the first question is whether DRS accounts for initial fielder positioning, and if not, then the effect would show up in both statistics, I would think.

        Vote -1 Vote +1

    • Rex Manning Day says:

      DRS still compares a player to the league average. If Granderson isn’t making plays in left-center than 40% of CFs make, his DRS score will be lower.

      Plus, DRS weights plays based on how likely it is that the average player can make it. So the plays Granderson is still making, which are up the middle, are probably relatively low-value because most CFs make them, too. But the plays he’s losing to Gardner are on the edges of CF territory, making them higher value because fewer players can make them.

      Vote -1 Vote +1

  7. Ken says:

    “Nick Swisher has dramatically improved in terms of UZR this year, producing more defensive runs above average in 2011 than he has in his entire career to this point.”

    This single sentence sums up my problem with UZR, especially as it relates to WAR. Is it possible that a consistently averageish outfielder all of a sudden becomes an above average defender in his age 30 season? I suppose it might be. Is it likely? I think not. Should such wild and apparently random fluctuations exert a considerable influence on the way we value players from year to year? I would say no, but that apparently puts me in the minority around here.

    Vote -1 Vote +1

    • Benjamin says:

      Crappy hitters have superlative seasons at the plate all the time (see Bautista, Jose). Why can’t crappy fielders have good seasons in the field?

      Vote -1 Vote +1

      • Frank says:

        Agreed. Fluctuations in baseball are the norm. No one thinks Swisher is all of a sudden a great defender, but he’s having a good defensive season. He should revert back to average/below average next season.

        Vote -1 Vote +1

    • waynetolleson says:

      “Is it possible that a consistently averageish outfielder all of a sudden becomes an above average defender in his age 30 season?”

      Sure. First of all, Swisher was never a terrible defensive player. He had the ability to play the corner OF positions, CF, and 1B, as well.

      When he was in Oakland and Chicago, Swisher was continually rotated between those three positions. When he got to the Yankees, Xavier Nady got hurt, and Swisher became the team’s everyday RF.

      This is now Swisher’s third season playing in the same position, playing half his games in the same stadium. It’s probable that given more and more reps at one position, Swisher has improved his knowledge and technique at that position. The fact that Granderson and Gardner both cover so much ground, combined with the fact that RF is pretty short in Yankee Stadium, also helps to make Swisher’s job easier.

      Vote -1 Vote +1

    • Ronin says:

      I think you are on to something here.

      Both Swisher and Gardner show really good numbers for UZR in the OF which appears to be coming straight out of Granderson’s totals. What seems more likely? Is Granderson just not getting to balls and a whole lot more hits are falling in or does the positioning just mean that Gardner and Swisher are catching the same flyballs that a normal CF would on a different team. If you put three good hitters in a row they all get their chance to hit, put three good OFers together and one is gonna get shorted. UZR compares players against the league average for their position correct? In my opinion that leaves too much chance for error when a player’s teammates are significantly worse or better than league average. Would Zimmerman still be such a great 3B via UZR if Ozzie Smith was the SS or vice versa? Granderson might not be adding as much with his glove as he would if Glenallen Hill played LF next to him but I dont think he is actually costing the Yankees runs with his defense.

      Vote -1 Vote +1

  8. Spunky says:

    Awesome article Joe! I never thought about this before but I’m glad that the Grandyman is probably being underrated by UZR. Now if we can only figure out what was up with Teixeira’s in 2009 and 2010…

    Vote -1 Vote +1

  9. waynetolleson says:

    Granderson’s defense is normally pretty good. The past month has been kinda tough. He has misjudged several catch-able fly balls that all fell in for singles or extra-base hits. A couple games ago Granderson and Swisher got caught staring at each other, and a routine pop-up fell between them for a hit.

    Those four plays probably cost the Yankees four or five runs. Granderson’s UZR was about average the first three months of the season. He has had an exceptionally bad few weeks in the field.

    Vote -1 Vote +1

  10. strongbad56 says:

    I have felt that Granderson has been overlooked for MVP for a while now (on here mostly, not ESPN). He’s been the 2nd best hitter in the American league, plays a tough position at CF, and I just don’t trust those UZR numbers. I didn’t even know about the zoning issues, but just looking at his career UZR numbers (and watching him on the field), it is hard to believe that he is truly this bad of a fielder. With Texas he had three years with an average UZR, two with an excellent one, and one pretty bad. Then he got to New York and put up a good UZR, and now this year it’s bad? I just don’t buy it. No one can fluctuate that much in quality of fielding. My guess would be that he is an above-average fielder (3-4 UZR/150), which would give him 7.3-7.4 WAR, easily good for 2nd in the league and actually pretty close to Joey Bats. If you add in the likelihood that Bautista is a worse fielder than UZR is showing, then it’s not inconceivable to say that Granderson has been more valuable.

    Vote -1 Vote +1

    • strongbad56 says:

      Detroit, not Texas, sorry.

      Vote -1 Vote +1

    • Eric R says:

      “he had three years with an average UZR, two with an excellent one, and one pretty bad. Then he got to New York and put up a good UZR, and now this year it’s bad. No one can fluctuate that much in quality of fielding.”

      I think that has to do with the stat being relative to average. You see a couple average years and a couple plus years and then assume a negative year is bad data.

      If it was a stat that went from zero to one like AVG it would seem pretty different while being pretty similar .

      Here are Grandersons’ UZR/150 with a ballparked relative wOBA

      2006 +13.6 .356
      2007 +14.5 .358
      2008 -11.9 .307
      2009 -1.4 .327
      2010 +7.9 .345
      2011 -12.7 .306

      If a player put up a series of wOBA [or the equivalent stat you'd like] over a six year period, it still stands out, but not quite the same as those ‘extreme’ positive and negative numbers

      Vote -1 Vote +1

    • gabriel says:

      There isn’t much of a likelihood that Bautista has been worse defensively than his UZR numbers – Dewan’s has him at +6 over RF and 3B, and his RZR for RF shows him in the middle of the pack.

      I’m sympathetic to the idea that Granderson isn’t as bad in CF as his numbers this season indicate, but with UZR and Dewan’s in agreement, there has to be something tangible to support the view that he’s above-average.

      Vote -1 Vote +1

  11. GiantHusker says:

    I emphatically second Benjamin. I’m not qualified to judge UZR, but ignoring it because it varies from season to season is clearly not justified. If there’s something else wrong with it, fix it, average it with other defensive measures, crowdsource fielding or something, but defense absolutely needs to be included in the evaluation of a player.

    Vote -1 Vote +1

    • waynetolleson says:

      “I emphatically second Benjamin. I’m not qualified to judge UZR, but ignoring it because it varies from season to season is clearly not justified.”

      I would counter by saying that the severe variance in players’ UZR’s from year-to-year should certainly be taken into account when assessing a player’s defensive value/performance.

      Numbers and statistics don’t tell you everything about a player’s abilities or performance, especially on defense. There are so many factors that determine whether or not a fielder makes a play besides that player’s skill and athletic ability.

      The positioning of every fielder is dependent both on the batter and, more importantly, on the baserunners.

      Who’s on base, and at which bases? Are the runners fast? Are one or more runners in motion on the pitch? Who’s the batter? What’s the score of the game, and in what inning?

      For example, if a hitter scorches a high line-drive down the RF line, right over the first baseman’s head in a tie game in the third inning, the OF would likely be playing straight away; that ball is probably going into the corner for extra bases, even if the RF is Ichiro Suzuki in his prime.

      However, if it’s two outs in the eighth and the same batter hits the same ball to the same spot, the corner outfielders would likely be guarding the lines in a “no-doubles” defense; the same line drive would probably be caught for an out, even if the RF is Bobby Abreu or Hideki Matsui.

      These types of things apply to every defensive position, and every game situation. Stats like UZR can account for many of them, but I doubt they can account for all of them.

      Vote -1 Vote +1

  12. Drew says:

    Do Granderson’s strikeouts deter anyone else from championing him for MVP? I know people have looked at how much Ks actually matter, but I have much disdain for that style of batting. Plus he bends so far over that his strikezone is half of what it should be, enabling him to use that long swing that yields all those home runs. I obviously can’t say that it prevents him from being productive, but it pains me to watch.

    -8 Vote -1 Vote +1

    • 28 this year says:

      I am sorry. Don’t watch him in his best season to date. He makes some minor tweaks and has hit like almost 50 homers over the past calendar year but if that hurts, perhaps, you are watching it wrong. You do know hitting home runs is good?

      Vote -1 Vote +1

      • Drew says:

        I understand that the result of him bending over the plate and hitting home runs is a positive for his team. It’s just an ugly swing.

        Vote -1 Vote +1

    • williams .482 says:

      and this means he is less valuable how… ???

      Vote -1 Vote +1

  13. sc2gg says:

    The problem with someone like Granderson being in an MVP race is mentioned by you – the NY media preaching his case. We are going to get endless insufferable articles like this, courtesy of the ever-accurate Bleacher Report:

    http://bleacherreport.com/articles/817367-mlb-curtis-granderson-is-the-american-league-most-valuable-player-dont-argue

    Obviously the article doesn’t need to be read, but the title is more than enough. Let’s check some hot quotes:

    “Don’t give me Adrian Gonzalez or Miguel Batista or even the great Justin Verlander. The MVP is Curtis Granderson.”

    “Danny Valencia hit a fly ball into right-center field. Granderson and right fielder Nick Swisher each called for the ball, which fell safely for two bases.
    Granderson put his arm around Swisher and took responsibility for the confusion.
    “I think I went too far out of my range,” he said.
    Spoken like a true MVP.”

    Aarrgggg…

    Vote -1 Vote +1

    • waynetolleson says:

      I get that the NY media is ridiculous and out of control. So let’s just forget about it and stick to the topic. At the very least, Curtis Granderson is worthy of being considered for AL MVP. He’s near the major league lead in HR’s, and leads baseball in runs scored and RBI. He also leads the AL in triples and has 24 SB’s.

      This is undeniably a very productive offensive player. He also plays CF, which is one of the most difficult defensive positions to play. His metrics have him as a below-average defensive CF, but I believe those low scores are the result of a short spell of exceptionally bad defensive play on Granderson’s part. Most of the year, Granderson has been a decent-to-good CF, which makes him a great fielder, given the difficulty of playing the position.

      There’s still five weeks to go. We’ll see how all the contenders play down the stretch. If Granderson continues at his present pace, then he certainly warrants serious consideration for AL MVP.

      Vote -1 Vote +1

  14. Telo says:

    I’d call him an average to very slightly below average CF. Still doesn’t put him ahead of Bats.

    Vote -1 Vote +1

  15. vivalajeter says:

    There are a finite number of balls that are hit to the outfield. If Gardner and Swisher are both having great years, doesn’t that mean Granderson’s numbers should go down? Could it simply be a matter of Granderson allowing the corner outfielder to catch balls in the gaps a little more often than other CF’s do? There are plenty of times during a game where the CF and RF are both able to get to the ball, but only one of them can get the out.

    Vote -1 Vote +1

  16. RBIBaseball says:

    Has anyone taken into consideration that Jose Bautista’s WAR has been around one since the break. On top of that he’s only hit about 4 or 5 homers since the break too. Was Jose Bautista the first half MVP? Yes, yes he was. But to give a guy an award for half of season of production is absurd.

    Vote -1 Vote +1

    • gabriel says:

      Yeah, Bautista had a bit of a slump post-ASG, to the extent that his 2nd-half line is .268/.434/.495. Pretty much as valuable as Granderson’s season line.

      Vote -1 Vote +1

      • RBIBaseball says:

        I’m just saying that I rather give the award to a guy whose been consistent in terms of production for the whole year than some guy who had one hell of a first half. Whose more valuable to a team, a guy that puts up great numbers for the entire year or a guy who has an absurdly good first half.

        Vote -1 Vote +1

      • gabriel says:

        I don’t think it matters much to a team if a player is consistent or streaky. One could even argue that teams ought to prefer streaky players, since it would allow them to get the ‘hot hand’ into the lineup more often. It’s Bautista’s overall production that’s superior; how he accomplishes it shouldn’t matter that much. Indeed, if you want context-specific analysis, Bautista’s WPA is +6.98, while Granderson is at +3.00 (less than half of Bautista’s, though still very good).

        Vote -1 Vote +1

    • ACra says:

      Huh?

      Bautista’s monthly splits:

      July – .316/.426/.646
      August – .263/.429/.600

      The worst month he had was June when he hit .258/.405/.430. He’s hit 14 HR between July and August. Not seeing this slump you are referring to.

      Vote -1 Vote +1

  17. Austin says:

    I think what we desperately need here is to use crowdsourcing to compile a list of managers in the UZR era who used unconventional outfield alignments, and see if the UZRs of the players involved tend to improve once the manager is replaced or reverts to a standard defensive alignment. Dedicated fans of particular teams will probably remember any such years and would provide enough information to use if a FanGraphs author writes up a post asking for it.

    Vote -1 Vote +1

  18. Jason says:

    I think Granderson is the clear MVP. The award has traditionally gone to the best player on a contending team. Granderson is the best player on any of the contending teams. He’s amongst the leaders or leading the league in just about every offensive category and he is an above average fielder playing one of the most difficult positions.

    That said, I don’t think he will win the award. The award will probably go to the WAR leader which is a shame. It cheapens the award to hand it out based on such a seriously flawed statistic. Plus it messes with the continuity of the game. Joe D won all those MVPs over Ted Williams not because he was a better player, but because his team won. If you don’t like the MVP award as it’s traditionally been awarded, make a new award (the “WARrior”?). Don’t redefine the awards we’ve already got.

    -10 Vote -1 Vote +1

    • buddy says:

      And the Cy Young award has traditionally gone to the pitcher who won the most games.

      Vote -1 Vote +1

      • Jason says:

        No it hasn’t. The Cy Young has traditionally gone to the pitcher with the best combination of wins and ERA. The Cy Young has always been thought of as the League’s best pitcher. Voter’s still award the Cy Young to the league’s best pitcher even if they don’t regard wins as highly as they used to. I still think wins are important unless a pitcher is far and away better than everyone else (Greinke and King Felix were both good choices I think).

        Vote -1 Vote +1

    • Nitram Odarp says:

      It cheapens the award the hand it out based on a flawed statistic (WAR), but it doesn’t cheapen it to base it on something like team wins which individual players have little control over?

      Vote -1 Vote +1

      • Jason says:

        Not at all. The MVP has always been thought of in that way. We can count the exceptions off the top of our heads (ARod, Ripken, Dawson). Whether you like the logic of the award or not, it is not obviously stupid to think that Granderson’s play is more valuable when he might be the difference between first place and second place in the AL East. ….after all, the Jays could probably hold the Orioles off for fourth place without Bautista. But who cares? It’s fourth place.

        This is obviously out of Bautista’s control but such is life.

        Vote -1 Vote +1

      • Ari Collins says:

        “We’ve always done it that way,” is a very poor argument for continuing to do it that way. You’re in the wrong place if you think that tradition trumps analysis.

        Vote -1 Vote +1

      • Jason says:

        Ari,

        I have no problem with change. But the idea that the WAR leader should get the MVP really does redefine the whole character of the award. If we were to go back through history and re-award MVP based upon WAR, we would probably change the recipient in most years.

        On the other hand, because Cy Young has always been thought of the best pitcher in the league, using better tools to evaluate best pitcher seems like a natural progression of the award. And, indeed, going back in time and reevaluating best pitchers using new tools probably wouldn’t change the recipient that often.

        Vote -1 Vote +1

    • KG says:

      Award is MVP, not who has the best teammates. If you don’t think the Jays would lose more games without Bautista than the Yankees without Granderson, you are crazy.

      Vote -1 Vote +1

      • Jason says:

        I agree that the Jays would lose more games without Bautista than the Yankees would lose without Granderson. So, I guess I’m not crazy.

        However, I also think the value of a single win is greater for the Yankees than it is for the Bluejays. I weight Yankees wins higher because they actually matter.

        Vote -1 Vote +1

      • JDB says:

        You’re right in one sense, Jason: the Blue Jays aren’t going anywhere, so you can reasonably argue that each additional win is meaningless (or close to it).

        But there is an alternative. The award is called Most Valuable Player. When comparing a player to his teammates, maybe we should look at the difference in WAR between the best player and the next best players.
        Does it really make sense to say player A is the MVP when he is only slightly more valuable than the 2nd best on his team?

        Granderson has a 6.1 WAR, while Gardner has 4.4 and Teixeira and Cano have 4.3. Granderson has 1.7 more WAR than his best teammate and 1.8 more than his 2nd and 3rd best teammates.

        Bautista has 7.7 WAR, and the 2nd best is Yunel Escobar at 4.0. After that are Jose Molina (1.4) and Edwin Encarnacion (1.3). Bautista is 3.7 WAR better than his best teammate, and an unbelievable 6.3 better than his 2nd best teammate!

        The Yankees would still be damn good without Granderson. (Same with the Sox and any of their MVP candidates.) The additional wins from Granderson only help with the Yankees’ playoff seeding – is that really valuable?
        But the Blue Jays would be godawful without Bautista – he is the reason they’re over .500.
        Thus, Bautista is more valuable to the Blue Jays than Granderson is to the Yankees.

        Now, I’m not saying the player with the most WAR relative to his best teammate should win. (Someone with 4 WAR shouldn’t win because his best teammate has -1 WAR.)
        But this factor, combined with Bautista’s brilliant play, means he should be the MVP.

        Vote -1 Vote +1

      • Jason says:

        JDB,

        First, WAR is a pretty meaningless statistic because it incorporates a lot of meaningless shit (read my post at the bottom of this thread about UZR). So I wouldn’t recommend looking at WAR at all.

        Second, now you are talking about only giving MVP to players with shitty teammates. I think that is entirely antithetical to the traditional understanding of the award.

        Vote -1 Vote +1

  19. Jason says:

    On to defense:

    I can’t fathom why people put stock into UZR. It’s all over the place. A few years ago Ellsbury was moved out of CF into LF based solely on his abysmal UZR. Everyone applauded Theo’s brilliance. This year people are claiming that Ellsbury deserves the MVP award over Granderson based on his defensive superiority as demonstrated by (you guessed it) UZR.

    Its like a flock of sheep being led around by stochastic noise….

    Vote -1 Vote +1

    • DavidB says:

      Ellsbury to LF was very much like Gardner playing LF in deference to the veteran Granderson. If you have two CF without the arm to play right and you want to get them both into the outfield you play one of them in LF.

      Also Ellsbury has always had plus plus speed but his routes to flyballs have gotten much better with experience. He may not have gotten this much better, but his defense has definitely improved over his career.

      Vote -1 Vote +1

      • Jason says:

        This is incorrect. Ellsbury to left was totally based on his dreadful UZR and Cameron’s good one. Ellsbury and Cameron was not a comparable situation to Gardner and Granderson. You don’t move your centerfielder of the future out of position to make way for a 38 year old with no possible future. Granderson was in the prime of his career and Gardner had never played a full season. Additionally, Cameron has the arm to play right field (and he did play there for the Mets when he moved over for Beltran).

        Vote -1 Vote +1

      • DavidB says:

        So what defensive alignment of Ellsbury/Cameron/JDDrew/Darnell Mcdonald/Bill Hall do you think is more compelling than what the the Red Sox chose to do? The similarity is that you let the older more established player play his most comfortable position and if Cameron and Ellsbury had stayed healthy that would have been a reasonable plan. I think having young guys play out of position while backing up their natural position may not be ideal for them, but managers do seem to value keeping veterans happy.

        Also, when healthy, Mike Cameron played a pretty spiffy CF even into his late thirties.

        Vote -1 Vote +1

      • DavidB says:

        Also what makes you think Epstein and Francona just sit in front of their computers looking at UZR scores rather than talking to scouts and using their own eyes?

        Vote -1 Vote +1

      • williams .482 says:

        The sox front office has built their own computerized defensive statistics, using HITf/X data which is not available to the public now. They also have a very large number of scouts. Ells to LF was NOT “totally based on his dreadful UZR.”

        Vote -1 Vote +1

    • EDogg1438 says:

      I’m sure Ellsbury moving to LF for a season had nothing to do with a perennially elite CF like Cameron being placed there.

      Would it instead be better to assume that fielders never change from year to year?

      Is it impossible for fielders to improve or decline?

      Just like a lot of good hitters can randomly have a down season, it is not unreasonable to believe that a good fielder can have a down fielding season.

      Vote -1 Vote +1

      • Jason says:

        Fielders really don’t have “off” fielding seasons. Fielders can improve by making fewer errors, but they can’t improve reaction time or foot speed, etc. They have to work with the ability they have. The exception to this is psychological problems like Knaublauch or Sax, or Utley.

        Do players ever have “off” running years where they just don’t run as fast as one year as all the rest? Of course not (unless they were injured).

        Hitting is different. It is a very complex skill and it also involves a fair amount of luck once a ball has been put into play. Good hitters really do go through slumps where they are temporarily not good hitters (timing issues, swing issues, etc.). For example, Alex Rodriguez had as much chance of getting a hit in the 2006 postseason as I would have. I could never suddenly field as well as Ichiro (nor could I suddenly field as poorly as Adam Dunn).

        Granderson isn’t fielding any different this year than he did last year. Similarly, Ellsbury isn’t fielding any different this year than the year the Sox management thought he couldn’t cut it as a CFer. It is all just error in a really shitty statistic.

        Vote -1 Vote +1

      • williams .482 says:

        How does an outfielder catch a fly ball? They watch it (like they would a pitch, in a way) and try to judge here it will land. If they are unsure, they make their best guess and reevaluate as necessary. Once they have done that, they run to the place where they think it will go, and if they were right, try to catch it.

        Why do we think that the above process is completely without any chance of varying from year to year? If someone is “not seeing the ball well” and slumping because of that, might they also have trouble tracking flys for the same reason? also, players have so many minor, unreported injuries, that they could very well be slower or faster year to year, based off of heath.

        Vote -1 Vote +1

  20. noseeum says:

    How is team UZR calculated? I can see on the fangraphs team leaderboard that the Yanks are third in outfield UZR. Is that a “top down” number? Meaning it looks at all balls in play to the Yankees outfields and rates them on the whole?

    Or is it a “bottom up” number, meaning it simply is a combination of the underlying individual UZRs?

    Vote -1 Vote +1

    • Nitram Odarp says:

      Pretty sure the result should be the same either way. The position that catches the ball makes no difference in how much credit is given. As long as the ball is caught, no one is punished. If a ball is dropped, the deduction is split up based on the percentage of each position that would normally make the play.

      Vote -1 Vote +1

  21. gabriel says:

    If we are going to discount Granderson’s measured defensive value, shouldn’t we also discount his positional adjustment? It seems ridiculous to give Granderson a pass on his defensive numbers but full credit for being an everyday centre fielder. This seems to me particularly to be the case in New York, where Gardner may well be the superior defender in centre. For the Yankees, there shouldn’t really be any value placed on Granderson’s ability to occupy CF, just as we shouldn’t place any value on Jeter’s ability to stand between 2nd and 3rd in the early years of A-Rod’s tenure in NY. Positional scarcity, particularly in MVP discussions, is relative to the potential replacements.

    Vote -1 Vote +1

    • Jason says:

      Jeter was always a better defensive player than Arod.

      -10 Vote -1 Vote +1

      • Jason says:

        …Arod also made more sense at third than Jeter since he has a far superior arm.

        Vote -1 Vote +1

      • gabriel says:

        Not even close. Arod was a good defensive SS before he was moved to 3B while Jeter is a historically bad defensive SS. A great player, but a horrible defensive SS.

        When Arod came to the Yankees, their 2B was Miguel Cairo. Moving Jeter to 2B and putting Rodriguez at SS would have been the best defensive alignment, and would have saved Yankees fans the Miguel Cairo era.

        Vote -1 Vote +1

      • Jason says:

        Is there any actual evidence that Arod was a better defensive shortstop than Jeter? Is third base a harder position than SS? Arod has never played a decent third base. Even when he was young.

        Jeter is certainly not an all time bad shortstop. That is laughable.

        Vote -1 Vote +1

      • Jerome S. says:

        You have to be trolling…

        Vote -1 Vote +1

      • Jason says:

        No, I’m perfectly serious. My eyes tell me that Jeter is the better defender. The only supposed evidence to the contrary, as far as I can tell, are other people’s opinions or the silly fielding numbers. I don’t think there is any real evidence that Arod is a better defender than Jeter.

        -8 Vote -1 Vote +1

      • Helen Keller says:

        My eyes agree with this assessment.

        +12 Vote -1 Vote +1

      • williams .482 says:

        Trolling, I am sure. you really feel that every defensive metric but webgems and FLD% is lying to you when it comes to Jeter? Seriously, when the ALL agree over a NINE YEAR sample, I think you can trust them.

        Vote -1 Vote +1

  22. Domenic says:

    I tweeted you this question on August 10th (http://twitter.com/#!/DomenicLanza/status/101374700973850625). Can I pretend that that helped inspire you to pen this post? I would be entirely heartbroken if it went unnoticed.

    Vote -1 Vote +1

  23. Jaysfan says:

    Yankees writer’s and fans make me laugh. They praise UZR when it single handedly boosts Brett Gardner’s value to a 4-5 WAR player, yet poke holes in it when it shows the foibles of their plateauing center fielder Granderson. PS, anyone who thinks Jose Bautista shouldn’t win MVP is senile. If A-Rod deserved to win in 2001 with last placed Texas, Bautista does now.

    Vote -1 Vote +1

    • Jason says:

      I don’t know any Yankee fan that actually believes Brett Gardner is one of the best players in the league. I know it’s a myth that UZR people created.

      Who say’s A-Rod deserved to win in 2001? Thinking Bautista is the league MVP doesn’t make you senile. But thinking that any other choice is unthinkable probably does…

      Vote -1 Vote +1

      • Jerome S. says:

        Uhm, Brett Gardner isn’t like “elite” but he’s a top-20 player in the league, easily. I should point out that his UZR is over multiple seasons, while Granderson’s disputed sample is over one.

        Vote -1 Vote +1

      • Jason says:

        There are probably 20 players in the AL East better than Gardner (5 on the Yankees alone!). He’s not even close to top 20 in the league. Not even close.

        You can’t throw away UZR in the players it varies in and keep it in the ones it doesn’t!!!! ….sometimes when I throw darts they all end up in a tight pattern. Sometimes they are all over the board. If I ignore the times they are all over the board I’m super consistent!!!

        Vote -1 Vote +1

      • Sultan of Schwwingg says:

        Not that I disagree about Grandy being a good CF or Gardner being a wonderful LF, but Dom specifically mentioned Granderson’s declining defense after his trade to NY, giving some validation of his poor numbers this year, and Gardner doesn’t have enough opportunities in LF for us to conclude anything. We assume he’s very good, but there’s no data proving that.

        Vote -1 Vote +1

    • Sultan of Schwwingg says:

      I’m not senile, and I don’t think he will win it. For one, I fully subscribe to the you have to blow away all competitors to win an MVP on a mediocre team. He has not blown Grandy away, nor Pedroia, for that matter.

      Two, since no one has ever (to my knowledge – although B Anderson was close) done what Joey has, there will be too many voters skeptical about his doing it clean. The present debate about Jays batters stealing signs will not help.

      Vote -1 Vote +1

    • Rex Manning Day says:

      Nobody’s suggesting throwing out UZR. Heck, if you just use a 3-year average instead of a 1-year sample, Grandy’s WAR jumps to 6.8, second only to Joey Bats’ 7.5.

      As for Gardner, his UZR/150 over the last 3 years: 20.5, 31.8 (!), 28.8. If a player fluctuates significantly in year-to-year UZR, it’s tough to pick any one year and say “yes, that’s how good his fielding is”. Just like if you sometimes throw darts well, and sometimes throw them poorly, it’s tough to point to any one showing and say “yes, that is how good he is at darts”. But if a player consistently puts up huge defensive numbers year to year, it’s pretty safe to say that he’s pretty good defensively.

      Also, I don’t think “plateauing” means what you think it means. This year, Granderson has the highest OBP, SLG, ISO, wOBA, wRC+, and wRAA of his career. He’s on pact to double his WAR total from last year, which will mark the second highest WAR of his career (behind 2007′s 7.8). All this on the back of a BABIP 3 points lower than his career mark.

      If Granderson *does* plateau from here, that just means he’s going to be a 6-7 WAR player for the next few years. That…doesn’t seem all that bad, really.

      Vote -1 Vote +1

  24. TK says:

    As I understand it, UZR for a season (or 3/4 of a season) is not a good predictor of future value, but is a good evaluation of the player’s defensive value that season in terms of results (much like a 25% HR/FB ratio would not be a good predictor of future production but would mean that the player hit a ton of HRs). This difference needs to be understood. So while the flaws in UZR that come from better flanking fielders or smaller/larger stadium outfields are legit, an award that rewards individual season accomplishments should not factor in other seasons.

    Vote -1 Vote +1

  25. jklender says:

    Show us the data!

    Joe, your article again raises some good questions that have been asked before, but it is still all just speculation without taking the extra steps of testing these hypotheses against the data — hopefully someday leading us to more refined metrics.

    Vote -1 Vote +1

  26. Jake says:

    “This is incorrect. Ellsbury to left was totally based on his dreadful UZR and Cameron’s good one. ”

    1. The Red Sox don’t look at UZR, they have their own proprietary system for rating defense.

    2. According to Epstein, it told them that Ellsbury was “above-average” in 2009.

    Vote -1 Vote +1

    • Jason says:

      haha, take that with a grain of salt. Ellsbury was and is a trade chip for him. Do you think he’s going to tell us he’s terrible even if he thought it?

      Vote -1 Vote +1

      • Ari Collins says:

        Do you really think that they’re going to trade Ellsbury, he of the 7 win season?

        I’ll give you your main point, but Theo saying he’s “above average” doesn’t tell us that he’s above average OR that he’s below average. We really don’t know what the Red Sox think of Ells.

        We DO know that they don’t simply use UZR, though.

        Vote -1 Vote +1

  27. Ben Hall says:

    The positioning argument doesn’t make that much sense to me. He may not be getting to balls in left center that he would if positioned normally, but presumably he would be getting to balls in right center that he wouldn’t if positioned normally.

    It seems most likely that he’s having a bad year defensively. Probably not as bad as UZR has him, but still, likely below average.

    Vote -1 Vote +1

  28. Tom says:

    I love this idea that it’s Joey Bats and Granderson and then a dropoff.

    Just to make everyone happy, let’s use a THREE YEAR SAMPLE of UZR.

    Oh, look, Pedroia still gains a full WIN AND A HALF on Granderson just on defense alone.

    And of course we know that his .390+ wOBA this year is legit, as the one-year-UZR detractors will be the first to tell you.

    Looks like Pedroia is the best player on a contender. Whoops.

    Vote -1 Vote +1

  29. pft says:

    If you are going to use 1 year UZR for WAR, why not follow the instructions of UZR’s creator, and regress it by 1/2 toward league average. This is in the UZR primer on this board.

    I watch most Red Slox and Yankees games. Granderson is not as bad as UZR says and positioning has a lot to do with that, Ellsbury is not as good as UZR says but is still good, and Pedroia mught even be better than UZR says this year.

    Vote -1 Vote +1

  30. black_knight101 says:

    Anybody that still consistently relies on UZR for OF (even over large samples) needs to watch Carlos Gonzalez.

    Vote -1 Vote +1

  31. Thought says:

    Come on now, no one is buying into every MVP’s “this award is really a team award not a recognition for individual achievement” line?

    Vote -1 Vote +1

  32. Justin Verlander says:

    I am the REAL MVP. I would OWN any of these other suckas.

    Vote -1 Vote +1

  33. squelchtoad says:

    I’m sure someone has considered this at some point, but doesn’t it make sense that a player with better defensive teammates *actually is less valuable (in runs or in wins) to his team on defense,* thanks to a declining marginal returns argument?

    Lets say your team has an outfield of Ichiro, Brett Gardner, and a replacement level defender. Add Franklyn Gutierrez to this outfield (replacing the replacement level defender), and you team will save more runs on OF defense, say X more runs.

    Now suppose your team has an outfield of Hideki Matsui, Manny Ramirez, and a replacement level defender. Add Franklyn Gutierrez to this outfield (again replacing the replacement level defender), and your team will again save more runs on OF defense, say Y more runs.

    I think it’s pretty clear that X < Y (possibly X << Y). In the second outfield, there are a lot more uncaught balls for Gutierrez to step in and catch. In the first, a large share of the balls that would be catchable for Franklyn are already being caught.

    The upshot of this is that (if we're talking about actual marginal defensive value above replacement to the team rather than trying to predict future defensive performance, it may be TRUE (rather than simply an error in the statistic) to say that the good defensive play of Granderson's teammates makes him less valuable (above replacement) defensively.

    Whether it means he is actually *costing the Yankees runs,* and whether he shouldn't get some of the statistical credit that is currently going to Gardner and Swisher is a different story. I think we should probably regress his poor numbers (and Gardner and Swisher's superlative ones); that is, the "real" credit for the Yankees high UZR outfield is probably shared more evenly than individual UZR shows.

    Furthermore, if the MVP is truly the Most Outstanding Player, then we should no more penalize Granderson for having good defensive teammates than traditional voters want to penalize Bautista for having bad teammates all around. Granderson has played well defensively, even if he has not been worth as many marginal defensive wins as he would be in a worse defensive outfield.

    As a side note, this argument could also be applied to offense. A double hit by a Red Sox player is worth fewer marginal wins than a double hit by a Mariners player (though the difference may well be small). oWAR, because it is based on linear weights, does not reflect this. oWAR is useful as an assessment of outstanding-ness of performance, independent of team (though, as others have noted, some team effects like extra PA on a good offense do creep in). dWAR, because of the credit-sharing nature of UZR, ends up being a truer measure of marginal value, for better or for worse.

    Vote -1 Vote +1

  34. Andrew says:

    I don’t get it. Bautista’s 2011 line is

    GP AB R H 2B 3B HR RBI BB SO SB CS AVG OBP
    115 400 91 126 20 2 36 80 103 81 6 3 .315 .456

    GP AB R H 2B 3B HR RBI BB SO SB CS AVG OBP
    124 466 115 130 20 10 35 98 68 133 24 10 .279 .375

    Now Bautista has a huge lead in Avg and walks, but Grandy has a huge lead in runs, RBIs, triples, and Steals (and caught stealing and Ks). I can’t see how Bautista is “clearly” the MVP.

    Vote -1 Vote +1

  35. Jason says:

    The blind use of UZR to compare fielders reminds me of a famous paper in evolutionary genetics called “Reading the entrails of chickens: molecular timescales of evolution and the illusion of precision” (Graur and Martin 2004).

    http://www.ncbi.nlm.nih.gov/pubmed/14746989

    Evolutionary biologists want to know the time when species last shared a common ancestor and they can estimate these times using differences in DNA sequences (the “molecular clock”). For a long time after the technique was developed scientists went about calculating all sorts of divergence dates from DNA sequences. Graur and Martin showed that by stripping the parameters they use in the calculations of error they result in divergence estimates with false precision. When the error is properly included in the calculations they result in error bars that encompass the entire age of the universe!!! In other words, the seemingly precise numbers they calculated were actually entirely meaningless. They were worse than random because the false appearance of precision made them misleading.

    UZR is like this. UZR is a very complicated calculation with many parameters. Each parameter in UZR has an associated error. We don’t know the magnitude of the error, because its not published, however the subjective nature of the enterprise leads me to believe the error on each of the parameters is quite large. However, even if the estimates are individually quite precise, you still have to compound those errors to calculate the final error on the statistic.

    We don’t know, because it is not published, but it is likely the case that the error involved in UZR is actually greater than the differences they are trying to measure. If this is the case, then there is no reason to believe that even the player with highest and lowest UZR are, in reality, different from each other.

    At any rate, until we know the error associated with estimating UZR, absolutely no one should use UZR to compare players. There is simply no good reason to believe that if player X has greater UZR than player Y, player X is in fact a better fielder. It is incumbent upon the authors to provide this information, and until then, the statistic should be ignored entirely.

    Please note, that denuding estimates of error seems to be a problem inherent in all of SABR. However, the problem is especially egregious in fielding statistics because the estimates are inherently less precise (its not hard to tell if a batter ball was a homerun, etc.).

    Vote -1 Vote +1

    • pft says:

      You are right, but as they say, the “science is settled” If you do not believe, you are a denier, or a dinosauer.

      Numbers are a beautiful thing, but they can be used convince people the world is flat if the people do not take the time to understand their limitations and not accept them with blind faith (like a religion).

      Vote -1 Vote +1

  36. J says:

    If the NY media is so powerful than how did Morneau beat out Jeter in 2008?

    Vote -1 Vote +1

  37. Jeff says:

    That’s a disgrace that UZR has him being so bad at defense this year. I like Baseballreference’s defensive metric, and their WAR.

    Vote -1 Vote +1

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>