ALCS Coverage: The Decision

Writing about yesterday’s Yankees-Angels game is harder than it sounds. There are just way too many angles to cover, and because it’s the Yankees, most of them have already been beaten into the ground. But, let’s try anyway, and let’s focus on the one decision that stood out the most – replacing David Robertson with Alfredo Aceves in the 11th inning.

First off, let’s look at the two relievers. They’re obviously very different types of pitchers, with opposite strengths and weaknesses.

Robertson is a strikeout guy who uses his movement to induce swinging strikes at the expense of giving up some walks. He’s hard to hit, but if you let him, he can get himself in trouble with walks.

Aceves is the classic command guy who pounds the strike zone and pitches to contact. He’s much easier to make contact off of, but he’s not going to beat himself. He’ll let you get yourself out at the risk of putting the ball in play.

There are certainly scenarios where you would prefer one to the other. With a runner at third base and less than two outs, you want Robertson’s ability to get the strikeout and strand the runner. With the bases loaded and two outs, you want Aceves’ ability to throw a strike and not walk in the winning run.

But the bottom of the 11th didn’t present either of those scenarios. Instead, it was bases empty, two outs, with Howie Kendrick coming to the plate. Is there any reason to prefer Aceves to Robertson in that situation?

Not unless you’re relying on a number that doesn’t mean anything. Aceves’ ability to avoid the walk is essentially useless against Kendrick, because he refuses to walk on his own. Given how aggressive he is at the plate and how hard he hits the ball (the career .354 BABIP is no fluke), you’d prefer to have a pitcher who can make him swing and miss. He’s going to chase pitches. You don’t have to force him to swing by pounding the strike zone.

Robertson was the good match-up for a Kendrick type of hitter, not Aceves. Going beyond Kendrick, you’d still prefer Robertson to Aceves even if you assume that he’s not going to get Kendrick out, because now you have the winning run on base, and a ball in play is not what you want in that situation.

Girardi has consistently over-managed the bullpen in this series, using guys for one batter here or one batter there in order to try to get the best possible match-up. In this case, he already had it, and he ended up making a move that was neutral at best and a downgrade at worst.

Sometimes, less is more. The Yankees’ best chance of winning the World Series will come if Girardi is more willing to trust his relievers rather than making a move every time he thinks he can incrementally improve his odds by a couple of percent.



Print This Post



Dave is the Managing Editor of FanGraphs.


Sort by:   newest | oldest | most voted
Mike K
Guest
Mike K
6 years 10 months ago

It seems the matchup Girardi was going for was Kendrick’s ability to hit the fastball and curveball, vs his (relative) inability to hit the slider. FanGraphs Pitch Values seem to back this up – though they have Aceves throwing the cutter instead of the slider Yankees’s staff have called it a slider. Is the data just not reliable enough to make those kinds of calls? By FanGraphs, Robertson’s only plus pitch is the fastball, of which Kendrick does best against those. Aceves has 3 plus pitches (FB, CB, CT). One of the debates is whether this was good analysis taken too far, or Girardi just tinkering for the sake of tinkering.

The biggest problem of course was that Aceves came in and immediately got behind in the count, so that any advantage he may have had with his slider/cutter was gone.

Mitch
Member
Mitch
6 years 10 months ago

I hate to defend Girardi, but in this case I think you’ve ignored one key element that suggests that going with Aceves WAS the right play. Don’t focus on the relative BB/9 rates, instead look at pitch type win values: Kendrick is a dead fastball hitter (+9.4 wFB in ’09, +8.6 wFB in ’08) who fares relatively poorly against offspeed stuff (+1.2 wCB and +2.3 wCH in ’09, -1.9 wCB and -3.4 wCH in ’08). Since Aceves has an excellent changeup (+6.8 wCH in ’09) I have no problem with the call, aside from the general theme of over-managing that is often evident in October November.

BIP
Guest
BIP
6 years 10 months ago

Except you don’t throw changeups to same-handed hitters.

Mitch
Member
Mitch
6 years 10 months ago

Ooops, I took too long writing that post, but great minds think alike?

Brian Recca
Guest
Brian Recca
6 years 10 months ago

I was telling people the same thing as Mike and Mitch, it wasn’t a bad move, it just didn’t work out.

mike
Guest
mike
6 years 10 months ago

I agree with you, Mitch, but do you think Girardi was adhering to the numbers you laid out? I don’t know the answer; I’m simply wondering.

Mitch
Member
Mitch
6 years 10 months ago

Dave makes a good point about pitch type values being more luck-volatile than some other stats, but since most of the trends quoted above hold for several years in a row, I’m comfortable in using them.

Mike, I’m fairly certain that Girardi has no clue about pitch type values, but I’d expect him to be aware of both (a) Aceves’ above-average offspeed stuff (especially as an ex-catcher?) and (b) Kendrick’s reputation as “a fastball hitter”. Even Joe Morgan wouldn’t object to that kind of “eyeball” analysis.

Mike K
Guest
Mike K
6 years 10 months ago

Mitch, I doubt he was using FanGraphs values, yes (though it would be cool if he did). I think they had some data behind it though. Probably more along the lines of, Kendrick makes contact on a FB x% of the time, CB y% of the time, and slider z% of the time. Or something like that. Even if Girardi didn’t have that directly, I believe the Yankees have a team that uses stats to put together Girardi’s book. The “book” probably had some rating for Kendrick on pitch-types. Letter grades, scale of 10, etc.

Andrew
Guest
6 years 10 months ago

Girardi’s binder is the “Ask Madden” of baseball management.

Mitch
Member
Mitch
6 years 10 months ago

You’re right about “the book” Mike. What’s interesting is that, if that’s the basis for the decision, why not cite the relevant stats to defend yourself? Instead JoeG offers a vague non-explanation:

“It’s just different stuff against those hitters,” Girardi said, without offering specifics. “We have all the matchups and all the scouting reports, and we felt that it was a better matchup for us.”

Daniel
Guest
Daniel
6 years 10 months ago

Even though that data is presumably available to anyone with the internet and spreadsheet software, I would imagine Girardi would want to be vague about the specific conclusions drawn from that data. If it means that much to his in-game managing strategy, giving it away might give Scioscia a clue to Girardi’s tactics. A small thing, sure, but I don’t blame Girardi for wanting to keep that stuff vague.

Carl Robinette
Guest
Carl Robinette
6 years 10 months ago

Pinch hitting Jerry Hairston Jr. for Brett Gardner irked me even more.

Girardi burns so many roster spot with these minor moves, (a downgrade in this case) and then theres just 1 tool players left on the bench.

Kevin S.
Member
Kevin S.
6 years 10 months ago

The Yanks have been keeping Gardner away from lefties all year.

Jack Moore
Member
Member
6 years 10 months ago

I was thinking about this as well, as I thought Gardner might’ve actually been able to catch that ball Mathis hit last night (Hairston was damn close.) Hairston has a comparable UZR/150 career and recently, though, and playing splits makes sense there.

Still, I’m not completely sold on the move.

BD
Guest
BD
6 years 10 months ago

I’m confused. If Aceves had been on the mound, and Girardi brought in Robertson to face Kendrick, instead of the other way around, would that have been the RIGHT move because “Robertson was the good match-up for a Kendrick type of hitter, not Aceves”? Or would that have been the WRONG move because “the Yankees best chance of winning the World Series will come if Girardi is more willing to trust his relievers rather than making a move every time he thinks he can incrementally improve his odds by a couple of percent”?

The point is, Girardi wasn’t wrong to make a move per se; his mistake was in making the move he made.

Ron
Guest
Ron
6 years 10 months ago

The risk of replacing someone who has just gotten two outs with someone whose stuff may or may not be there balances and maybe outweighs the benefit of the possible better matchup.

BD
Guest
BD
6 years 10 months ago

So Ron: Does that mean Girardi shouldn’t bring Mo in for 4-out saves?

Ron
Guest
Ron
6 years 10 months ago

No, it means that when the pitcher you’re bringing in is not clearly more talented than the one you’re removing, and the one you’re removing has not allowed anyone on base in that inning, you’re matchup data should be very strong.

SteveP
Guest
SteveP
6 years 10 months ago

The really confusing part about this was that Robertson looked good and was not in trouble at all during that inning. He was throwing strikes and both his fastball and curveball were effective. Why take the hot hand out? The matchup stats had Kendrick a whopping 1 for 2 against Robertson and no history against Aceves. As was said before, Aceves came out of the pen and immediately got behind therefore negating any “offspeed” stuff advantage that he may have had.

Girardi seems to have a need to use every man (almost) out of the pen. He used Marte for one out against Figgins and then the next inning brought in Phil Coke, another lefty, to face the left handed Abreu. Now, Coke is better, but why use Marte for only one batter. If he wanted to turn Figgins around with a lefty, why not go with Coke in the first place? Or leave Marte in to pitch to Abreu. Then to bring in Aceves you leave just Gaudin in the bullpen. Again, with Robertson in no trouble, why not let him try and finish off the inning and then you can start off the next inning with Aceves.

It just didn’t make much sense.

Daniel
Guest
Daniel
6 years 10 months ago

I don’t think Robertson looked all that great out there. Rivera hit a pretty sharp ground ball, but right at Jeter (IIRC). Then Morales came up and absolutely smoked a curve ball that almost took Teixeira’s hand off. It’s not like he’d gotten a strike out and a weak pop up. Guys were hitting the ball hard.

Not that that should ultimately change the conclusion, but I think it’s false to say that Robertson looked just fine. He was lucky that there wasn’t already a guy at second base.

SteveP
Guest
SteveP
6 years 10 months ago

Also, this was Howie Kendrick were talking about. It’s not like he’s trying to find the best matchup to neutralize, say, Albert Pujols. Kendrick’s just not good enough of a hitter to justify taking out Robertson.

fire jerry manuel
Guest
6 years 10 months ago

it wasn’t a bad move. it was just strange. aceves has been very good all year. it’s not you expect him to give it up.

robertson wasn’t going to start the next inning anyway. so if you prefer the scouting match-up for aceves over robertson for kendrick, and this is going to be robertson’s last out/inning anyway, is it really BAD to bring in aceves?

it was weird/strange/nonsensical/vexing. that doesn’t mean it was bad.

OkulaFan67
Guest
OkulaFan67
6 years 10 months ago

How about this one? Eric Hinske is off the roster in favor of Freddy Guzman. Girardi pinch-ran Guzman (for Matsui) in Game 2 and then had to let him hit! Yesterday, same instance only this time he uses Gardner to pinch run. Again, Matsui is out of the game. Now, Gardner is in Matsui’s spot but Girardi burns another position player by pinch hitting Hairston. The Angels have now realized there is no incentive to pitch to Rodriguez with such poor hitters due up. Removing Matsui effectively removes Rodriguez from the lineup also. Instead of Guzman on the roster, Hinske could have been available to take over at DH instead of Guzman/Hairston. Girardi cannot think more than one move ahead!

Wally
Guest
Wally
6 years 10 months ago

Here’s another thing which I hadn’t noticed until now, as I read the game log:

“Mariano Rivera replaces Johnny Damon (LF) pitching and batting 2nd; Jerry Hairston moves from DH to LF (team loses DH)”

So, by moving Hairston out of the DH spot, Girardi is now forced to have the pitcher bat in the second spot, or PH for him. If this game goes past 12th inning that spot likely comes up, and with the way Girardi was managing the bullpen he’d be forced to either have Gaudin bat or start throwing position player or starters on their rest days out on the mound. And all these moves are originally set in motion because Girardi PR Gardner for Matsui and had him caught stealing. And of course, Cervelli, who had to PH for Mariano (which of course also limited Mo to pitching one inning or having to bat for himself), has a .682 OPS.

The one move of Aceves for Robertson isn’t so bad, but looking at the game as whole, Girardi isn’t just over managing the bullpen, he’s over managing the whole damn team. He’s burning through relievers attempting to find every possible small advantage and burning through his bench. He’s ending up taking some of his best hitters out of the game, and in this case was forced to take his best pitcher (Mo) out because of line up choices. It nuts.

Rob in CT
Guest
Rob in CT
6 years 10 months ago

I didn’t think about it at the time, but the pinch runner should have been Guzman. Then Gardner comes in later as a defensive replacement. Melky slides to LF, and the DH spot is intact (inhabited, most likely, by Hairston).

As for Cervelli… I like Frankie C a lot, actually, but I’d have let Mo hit at that point :)

Rob in CT
Guest
Rob in CT
6 years 10 months ago

It was an odd move, to be sure. Robertson looked really good. Then again, Aceves really should be able to get Mathis out, yikes. And certain members of the Yankee lineup had awful games (hat tip to Angel pitching, of course).

Matt
Guest
Matt
6 years 10 months ago

Bottom line here is not Girardi:

Johnny Damon: .214/.241/.357/.599
Mark Teixeira: .120/.241/.240/.481
Robinson Cano: .200/.259/.280/.539
Nick Swisher: .136/.208/.182/.390
Melky Cabrera: .200/.259/.200/.459

The Angels backup catcher got a game winning hit while five guys on the Yankees are on vacation.

Wally
Guest
Wally
6 years 10 months ago

Matt, I wouldn’t say Girardi lost this game for the Yankees, and of course more fault is always going to be on the players, because they are the guys that actually play. However, Girardi made some stupid choices that hurt his team’s chances of winning. There is no excuse for stupid things like losing the DH. Especially in a playoff game that is tied.

walkoffblast
Guest
walkoffblast
6 years 10 months ago

I wish you would quit harping on the DH thing. There was one out and the bases were loaded when he made the move. It was a desperation move, who cares if the pitcher is going to come up later if the game is about to end and having Damon’s arm out there increases that chance. Was it really worth it or necessary? Who knows? Was it inexcusable? Hardly. Same thing applies to the relief pitching move in question.

Is he over managing? Probably. Is it what cost his team? Not on paper.

Wally
Guest
Wally
6 years 10 months ago

“It was a desperation move, who cares if the pitcher is going to come up later if the game is about to end and having Damon’s arm out there increases that chance.”

Exactly, “IF the game is about to end.” The game didn’t end and Damon’s arm figures to be a very small factor in if that game ends. Regardless of arm strength, very few outfielders gun down runners at home on possible sac flies, and any hit wins the game anyway. Then in the next inning frame one of your batters is as close to an automatic out as you’ll see (except for pitchers of course) AND you can no longer bring out your best relief pitcher for a second inning. It was one tiny improvement at the cost of a larger future disadvantage. It was stupid.

“Is he over managing? Probably. Is it what cost his team? Not on paper.”

Well yeah, on paper his choices haven’t added up to costing his team a full win, but that’s pretty much impossible over a 7 game series (unless you do something horrific like not play A-rod, Tex and CC or something). But these moves have reduced his team’s chances of winning. Given the probabalistic nature of baseball, one of these choices may just have been the difference. Maybe Matsui hits a HR in his next AB, or Damon gets on and Tex hits a HR in the 11th. No one can say. But taking your best hitters and pitchers out of the game for tiny momentary advantages is dumb.

Seideberg
Guest
Seideberg
6 years 10 months ago

Now that I’ve had the day to stew upon this, I have sort of changed my tune about it.

I am still annoyed that the Yankees blew a winnable game with a move that I still view as inexplicable. With that said, the move was clearly driven by some sort of sabermetric principle that Girardi referenced vaguely. Clearly Girardi is not bright enough to come up with this idea on his own, or to make the infamous “binder,” so that must mean there is someone in the Yankees organization with enough clout to tell the manager what to do, who is pushing this agenda. While Girardi may not know which are the important stats to use, or when to use them, the idea that the Yankees might actually adopt some statistical analysis in their operations thrills me.

Also, I’m pretty sure that Joe Morgan hated the move (I’m heading over to ESPN to read his chat from today, with quite a bit of excitement), so that’ll help me sleep tonight…Unless, of course, he pinch hits for A-Rod tonight because his binder tells him that Freddy Guzman has a better batting average in the 6th inning in a road game on a Tuesday during a waxing moon in an even-numbered month and …..

Colm
Guest
Colm
6 years 10 months ago

Yes.

This looks like Girardi doing what numbers-obsessed web-geeks are always accused of: placing TOO MUCH faith in statistics, and ignoring his own eyes. Although, as many have pointed out, whatever stat’s Joe G was looking out weren’t significant enough to warrant pulling a pitcher 2/3 of the way through an inning after he’d recorded two quick outs.

There’s very very little advantage to be gained by switching Robertson for Aceves in that situation. If the roles had been reversed the same would still be true. But there’s value to be lost, in an extra inning game, in burning one more pitcher from your bullpen, and throwing another pitcher into the game cold.

Kevin S.
Member
Kevin S.
6 years 10 months ago

Weird, considering you’re commenting on an article written by a numbers-obsessed web-geek (and I mean that in the nicest way possible, DC) that absolutely killed him for his decision.

MPC
Guest
MPC
6 years 10 months ago

He said what stat-geeks are ACCUSED of doing.

Kevin S.
Member
Kevin S.
6 years 10 months ago

The issue isn’t Girardi’s brightness – the man has an BS in Industrial Engineering from Northwestern. Perhaps he doesn’t understand how to properly apply his statistical database, but he’s not stupid.

MPC
Guest
MPC
6 years 10 months ago

Haha, Industrial Engineering isn’t a hard major. It’s not even close to the level of civil/electrical/chemical.

wpDiscuz