The Curse of Jerry Hairston, Jr./Eric Hinske:

Wednesday, February 6, 2013

Hardball Talk: Yankees still deciding between Curtis Granderson or Brett Gardner in center field

Granderson has been exclusively a center fielder since joining the Yankees in 2010, but does have some previous experience in left field for the Tigers and has drawn some criticism for his defense recently. Meanwhile, Gardner has always played left field alongside Granderson, but he was a center fielder in the minors and would have stayed there for most other teams.

At this point I’m certain that Gardner is a better defensive player than Granderson, so he should be in CF since he’ll see more chances there.  I’m also certain Granderson will be gone after this year so for 2014 it’d be smart to get a feel for just how good Gardner can be in CF. 

When you flip two defenders the impact overall isn’t going to be huge, maybe a handful of runs over a full season, but with the apparent decline in the Yankees’ overall projection as of right now they really need to take every single incremental chance to improve themselves.

As far as what the impact might be, let’s consider these numbers.

The average CF sees about 462 chances per year.  The average LF sees about 365.  Granderson’s got a zone rating of .853 over the last two seasons which means he’s converted 85.3% of the chances he’s seen into outs.  Gardner’s got a career zone rating of .920 in CF in his career, although that only encompasses 183 games.  He’s got a .904 career zone rating in LF.  We can split the difference which ends up at .911. 

Not all chances are equal and the angles of the balls that CF see vs. LF may skew this somewhat, but let’s assume that Granderson would have a .85 zone rating in both LF/CF and that Gardner would have a .900 zone rating in both LF/CF. 

Granderson CF/Gardner LF: .85 x 462 + .9 x 365 = 721 outs
Granderson LF/Gardner CF: .85 x 365 + .9 x 462 = 726 outs

So you’re looking at a difference of about five outs, or four runs if both stayed healthy and played every game this year.  The real-world impact is probably closer to three runs.  It’s worth doing, but it probably won’t have a huge impact.

--Posted at 9:53 am by SG / 32 Comments | - (0)


Page 1 of 1 pages:

SG- Is there anyway to consider the difference in their arms?  Neither have a great arm, but Granderson’s strikes me as better (I think he has a quicker release). It might be only a run or two difference in preventing 1st to 3rd’s or tag ups from 2nd to 3rd.

I think Granderson throws better than Gardner, but Gardner’s been more valuable according to the arm component of UZR (7.5 runs vs. 5.2 runs) over the last three years and that includes a 2012 where he essentially didn’t play.  I think part of that is that he cuts off the balls in the gap with a frequency that surprises a lot of runners and that might get mitigated some in CF where he won’t be the one chasing all those balls in the gap, but I wouldn’t attempt to try and quantify the impact of that.

Because of the depth of LF in DNYS, I don’t think you lose all that much by flipping them.

Incidentally, the bulk of Gardner’s arm value came in 2010 (6.6) so it’s possible runners have wised up on trying to take the extra base on gappers against him.

Did you factor in the fact that Gardner’s arm could very well fly off while making a throw from CF? That has to factor in somewhere, right?

The good news is Gardner’s throwing arm is not the arm he needed surgery on.

[5] previously.

Resident lawery types: is this ( reasonable?

[0] Is it possible that ZR is underrating the difference between them?  That is, Gardner gets to a lot of balls outside his zone, so his ZR will be higher.  But Granderson doesn’t, so while it doesn lower his ZR, the fact is Gardner has a larger zone.

IOW, Gardner may have 475 balls in his zone in CF, where Granderson has 462, making it another 11 runs.  Of course it cuts both ways b/c Gardners LF zone is also bigger and they’re losing that advantage.  Plus all numbers are POOMA, of course.

Basically, I think that the analysis above may understate the difference in moving Gardner/Granderson.  Not saying it’s going to add a TON of value, but perhaps more like 6-8 runs instead of 2-3, and that half-win is pretty big.

[8] How does that work then for Granderson’s zone? Does UZR count balls that are hit into a players zone that *are* converted into outs, or that *he* converts into outs?

If its the former, and is combined with Gardner’s ZR not improving due to balls turned into outs outside of his zone, then Gardner’s ZR doesn’t improve when it should and Grandersons does improve when it shouldn’t (or should be at least neutral.)  Outside of Gardner’s zone is inside Granderson’s zone.

[7] - I’m not a “resident lawery type” but to me it is plausible he just contacting him for his MLB hearing. I won’t deny that.  Then again it is also plausible he used the guy who was giving him the steroids as a consultant on its effects and how to argue his innocence.

But all of that is pointless to me.  If you fail a drug test and get off on a procedural technicality, you still failed a drug test.

[0] Is it possible that ZR is underrating the difference between them?  That is, Gardner gets to a lot of balls outside his zone, so his ZR will be higher.  But Granderson doesn’t, so while it doesn lower his ZR, the fact is Gardner has a larger zone.

Sure, it’s possible.  Comparing just their CF time as Yankees, Granderson’s averaged 0.298 chances per inning, equivalent to 434 over a full year.  Gardner’s averaged 0.308 chances per inning, equivalent to 450 over a full year. 

Sample size concerns with Gardner apply since we’re comparing 1186 innings to 3832.  If Gardner’s getting to 15 more balls out of zone that Granderson wouldn’t have gotten to then the gap between them is much bigger, but at the overall team level you can assume you’re losing a commensurate, if slightly lower, amount of left field plays out of zone.

6-8 runs seems realistic if that all holds.

[8] How does that work then for Granderson’s zone? Does UZR count balls that are hit into a players zone that *are* converted into outs, or that *he* converts into outs?

I’m using ZR, not UZR.  The way ZR works is it defines the zone for a position as plays that are converted into outs at least 50% of the time.  So ZR = plays made divided by plays in that defined zone.  If you had a player who caught 80 balls in 100 chances all in his zone, his ZR would be .800.  If you had a player who caught the same 80 balls in 100 chances in zone but also made another 5 plays out of zone his ZR would be 85 divided by 105, or .810.  If you were to give the first player 105 chances instead of 100 and penalize him for those 5 additional plays not made his ZR would actually be .76 and you’d have a larger obvious gap between them.

[9] UZR looks at plays that the player makes and splits the (weighted) difference on plays no one makes. UZR doesn’t really have out-of-zone plays in the way ZR does, since it compares you to the typical fielder of your position for each zone, even if that zone is on the other side of the park.

[9] If I understand ZR correctly…

The field is split up into zones (obviously).  An individual player’s zone is any zone in which a ball hit is converted into an out at least 50% of the time (I believe).  So ZR is (outs-on-balls-in-zone)/(balls-in-zone).  This is where we start with the caveats.  If a ball is in a fielder’s zone but another player fields it - e.g. ARod cuts in front of Jeter to get a GB - the ball is not considered in the zone.  Also, there are balls out of zone.  If a field makes a play on a ball out of zone, it’s added both to the numerator and demoninator.

So, Gardner’s .904 in LF career, he’s probably something like .900 on BIZ (365*.900 = about 329 outs), but then adds like an extra 11 OOZ plays (so 340/376 = about .904).  A .904 ZR fielder with 365 BIZ (and no OOZ) only adds about 1 out, but we can see by my example Gardner was actually 11 outs better.  Again numbers are guesses and we may be able to get exacts using FanGraphs data.

Just visually - and I’ve seen a range map that confirms this - Gardner gets to a lot of plays that other LF won’t.  It doesn’t seem visually to me (haven’t seen any graphics) that Granderson gets to a lot of OOZ plays.

Edit: either I type too slowly, or you guys started answering way before I did.

I’m starting a clinic called Abiogenesis to create life from the cold, dead stuff of the universe.  In case there are ever any legal questions, I have made notes indicating that you all owe me money.

[15] Khan Noonien Singh will surely find you and take your technology so that he can use it to destroy planets. Hope you are read up on your Shakespeare so you can dialogue with him.

[15] I don’t have any money, I do have debts, can I give those to you instead?

[13] Hung up on “plays no one makes”...think you have it but…

Instead of zones they have buckets; a bucket is a combintation of location and speed (and possibly hangtime for FB, not sure if that’s added in yet).  But if a ball in play is converted into an out 100% of the time (other than that play) and the fielder misses it, he’s deducted for the full portion of the play (-1 play).  If he makes the play he gets no credit (+0).  However, if it’s hit into an area where the play is made only 20% of the time, missing the play debits you only -.2 plays, and making it credits you with .8 plays.  When all is said and done you add up the totals and that’s your UZR plays, and then apply the correct multiplier (like .8) and you have runs.  That’s the simple explanation…there are adjustments for shifts, ballpark, etc in there as well.

You know what’s confusing?  Data was built by Dr. Noonien Soong.  Why two characters with such similar names?  Grrr, Roddenberry.

yes, by “plays no one makes” I meant balls in play that are not fielded for outs.

[11] Thanks.  We really won’t know for sure until it happens of course.  It’s possible that age has slowed Gardner, or it’s also possible he’s still at his peak for speed and learning LF will allow him to be a better CF.  It’s also possible that Granderson’s UZR number last year was real (that is, he’s declining quickly as a fielder) and using 2 years of data underrates how poorly he’s played.  Or he suits LF better (good at going to his left but not right) and will be better.  I’d like to think the move puts the Yankees from an 88-89 win team up to an 89-90 win team though!

[19] Wonder if there was a story arc where they would be related somehow?  If you think about it, it would make some sense.  Khan was genetically engineered(In the 1990’s!  Whoops!) to create a better human, and help humanity.  His relations (brother, descendents, whatever) saw how giving a human that kind of power could corrupt, so set about learning how to build a machine for the same purpose (Brent Spiner was in an episode of Enterprise as a predecessor to Noonien Soong).  Along the way they changed their name slightly (Singh to Soong), and finally succeeded with Data.

I should do some Google searching and see if there’s been a book written yet…and then start writing!

[22] Doesn’t lightning bolt holder Keith write Star Trek novels?

[22]  According to my very cursory search, not related, but I think we all know how reliable I am.

[23]  Yes, he’d know!

Also, SG any word on the weird Matt Harrison projections I found yesterday?

[19,23,24] I’m guessing that it was purposefully made similar to drive Trekkies into a frenzy of speculation. Sort of their version of “the Walrus was Paul.”

Also, SG any word on the weird Matt Harrison projections I found yesterday?

Yeah, for some reason the WAR formula on his line was off.  His projected WAR is around 4.1 now, which still seems a bit high to me but CAIRO likes him more than me I guess.  It must think he’s a Yankee.

[27] Ok yeah that seems a bit more reasonable, although I agree it is a bit higher than I expected, since he doesn’t really limit hits or miss bats.

[24] I think I should get started on my novel before he does!  Okay step 1…take writing classes…

[26] Fortunately there are no Trekkies here!

(discretely removes Vulcan ears)

[30] Aren’t the ornery ones “Trekkers” who get annoyed when called Trekkies? 

It’s sort of embarrassing that I know that.

Anyway… Go-go Gadget Gardner for CF.  Make it so.

Jhonny Typo Peralta also named, also firmly denying ever using PEDs (though I can’t tell if he’s also denying ever going to the clinic).

Page 1 of 1 pages: Yankees prep for non-tender deadline
(45 Comments - 12/4/2016 3:50:18 pm) Peace & glove: Owners, players reach CBA deal
(12 Comments - 12/1/2016 8:33:36 pm)

Fox Sports: Sources: Baseball’s 21-year run of labor peace could be in jeopardy
(75 Comments - 12/1/2016 3:10:24 am) Eovaldi among three pitchers released by Yanks
(29 Comments - 12/1/2016 1:05:06 am)

Fox Sports: The Yankees are positioned to make major moves
(30 Comments - 11/26/2016 12:53:32 am) Yankees deal McCann to Astros for pitching prospects
(78 Comments - 11/20/2016 10:26:57 pm)

KickStarter: Help the documentary “Candypants: A Fairy Tale” come true!
(46 Comments - 11/17/2016 6:15:43 pm)

NYDN: Why Yankees and their fans shouldn’t care about Gary Sanchez’s Rookie of the Year snub
(11 Comments - 11/15/2016 4:34:07 pm)

TGS: MLB Rumor Central: Yankees, Mets show interest in Boone Logan
(26 Comments - 11/15/2016 9:11:12 am) Gardner wins first career Gold Glove Award
(22 Comments - 11/13/2016 12:30:47 pm)