This thing began life as a comment on the Daric Barton front page article, but it's really big and by the end I realized that what had started out as a conversation about the OF/1B/DH positions now had to include a conversation about the bullpen/pitching staff, so I figured let's just go full fanpost with it.
So let's start with position players. Obviously Coco Crisp and Yoenis Cespedes are in. Josh Donaldson, Jed Lowrie, Alberto Callaspo, and Eric Sogard are in as well. No matter who the catchers are, it'll be two guys. Brandon Moss has to be in, whether he’s a 1B, and OF, or a DH. So that’s nine slots out of 13 filled. In the mix in terms of the remaining outfield slots, the 1B slot, and DHing work, you have these guys: Barton, Nate Freiman, Chris Young, Seth Smith, and Josh Reddick. Five guys and four slots.
First, let me tackle the catchers. You've got Derek Norris, Stephen Vogt, and Kurt Suzuki in the mix here. Norris hits lefties better than Vogt, who hits them better than Suzuki. Vogt hits righties better than Suzuki, who hits them better than Norris. Vogt has a higher CS/SB ratio, but he has a tiny sample size. Norris and Suzuki have about the same ratios. Suzuki’s defensive value by rSB and and RPP averages out to about 12 per full season, while Norris has about enough career innings for one full season and has totaled -3.6 (this was surprising to me.) Vogt doesn’t have hardly any innings at all and has -.3 defensive runs.
I think you have to have Norris, because he hits lefties really well. As in, the only other MLBcatcher remotely on his level against lefties is Buster Posey. As in, tied for the fifth-best hitter against lefties in the American League. Tied with Adrian Beltre. The guys ahead of him on that list? They include Josh Donaldson, Mike Trout, and Miguel Cabrera.
The other selection, to me, you base on defense. Neither Vogt nor Suzuki is anything special as a hitter, but you'd like to have the option of a strong defender behind the plate. Suzuki, you know what you’re getting defensively: a solid catcher at the low end of the middle of the pack in terms of throwing out baserunners and at the high end of the middle of the pack in terms of blocking pitches. Vogt, in terms of the metrics, looks like an unknown quantity, but catcher defense metrics are still much more a work in progress than are other defensive metrics. The eyeball test, to me, says Suzuki is better defensively. The gut/heart/whatever test also says Suzuki is going to mean more to the pitching staff than Vogt will. So at catcher, I go Norris and Suzuki.
So now on to our five guys jockeying for four positions.
By WAR, Reddick is the best of the five for 2013 and by career figures, even when you account for variations in playing time. As a fielder, he’s good for a 19.6 UZR/150 over the course of his career (RF only.) As a hitter, he’s obviously had a down season with occasional stretches of brilliance, but even at his 2012 best, he wasn’t more than a "pretty good" hitter. Last year his strong suit was power; this year (and the rest of his career) it appears to be getting on base. Someday I hope he figures out how to put those two things together because he’ll be a monster if he can do it, but for now, let’s assume that his career figures are more who he is than either 2012 or 2013.
The first thing I notice, aside from the fact that he’s basically an average hitter by wOBA and WRC+, is a very slight reverse platoon split, and this is over nearly 1500 PAs, so it can’t be brushed off. However, I think it can be brushed off due to a slightly elevated influence of slugging; it’s the only thing he does better against lefties than he does against righties, and given that he’s had so many fewer PAs against them, I think the most we can say is that he hits lefties and righties about equally well, which is to say, about league-average. So in Josh Reddick you get a superlative defender—easily the best on the team at any position—and a league-average hitter against RHP and LHP.
As a defender, Young’s slightly above average. In CF, his career UZR/150 is 2.3, and he’s worked negligible amounts of time in the corners. As a hitter, he’s actually third-best among these five options against LHP this year, hitting them either "pretty well" or "above average" depending on whether you prefer wOBA or WRC+. But Young has also been victimized by bad luck this season. His career line tells us he’s a very good hitter against LHP. Like Nate Freiman/Daric Barton good. So in Chris Young, you have a basically average defender who can murder left-handed pitching.
So as a defender, he seems okay. After a little over 2000 innings in LF—which isn’t enough for me to put toomuch faith in UZR/150—he has a UZR/150 of 7.9. If you throw in the 1500ish innings he’s played in CF and RF (really just RF), however, his career UZR/150 drops to 1.9. So let’s say you’d only keep Smith as a LF. As a hitter, he, like Reddick, has had a rough season. Frustratingly, he was basically fine, about league-average, for the first half, but has dropped off a cliff in the second half. Still, against righties, he basically does about as well as Reddick does against lefties. And his career line vs. RHP is about as good as Young’s career line against LHP. So in Seth Smith you get a decent defensive LF and a league-average hitter against RHP.
As a defender, all we can really go off as is what we see, because Freiman doesn’t have enough innings played to evaluate him metrically. And what we see is ugly. So let’s move on to his hitting. He hits RHP about as well as my five-year-old nephew and hits LHP about as well as anyone in baseball. The end. He can’t field and he can murder LHP.
Barton is the only one of these five guys aside from Reddick with any kind of positive impact as a defender. To wit, he is the third-best defensive asset on the team based on raw comparisons. As I have pointed out elsewhere, he is also a historically good defensive first baseman. As in the third-best first baseman in the history of the UZR statistic. And he will probably become the second-best by the end of 2014, assuming Kevin Youkilis keeps trying to play defense, thereby sullying his current career UZR/150 figure.
As a hitter…suffice it to say that we all have our various opinions on the guy, but please note his career line against LHP. That’s right, he has a serious, legitimate reverse platoon split: .282/.389/.438, .827 OPS, .368 wOBA, 131 WRC+ against left-handers, pedestrian numbers against right-handers. In other words, Barton’s career against LHP is slightly better than Freiman’s 2013 against LHP. So what you get in Barton is an absolutely incredible defensive first baseman who can clobber left-handed pitching and hits RHP at a league-average level.
So who’s indispensable here? Barton is actually the biggest no-brainer. You absolutely take a stellar defender who can hit LHP like he does. After that is Reddick, whose defense is indispensable and outweighs a league-average bat. Next up is Freiman, who so far appears to be able to hit LHP just as well as Barton.
Who gets the last spot? Neither Smith nor Young has made a great impression overall. The case for Smith is that his glove is better in LF than Young’s is in CF, which is a weird thing to realize, and that his career tells us he can hit RHP. The case for Young is that his career tells us he can hit LHP, and that he can play center field.
Honestly I don’t know who I pick. On the one hand, I feel like Smith’s glove puts him over the top. On the other hand, I feel like you want a guy who can rake against LHP. On the third hand, you face more righties than lefties, so maybe you’d actually prefer the guy who rakes against RHP.
It was at this point in the original comment that I realized this conversation had to include the pitching staff. Part of me wanted to say drop a bullpen guy and carry a big bench. You possibly could do that for a short series, and you might even be able to do it for a long series.
So let's shift gears to the pitching staff, which usually occupies 12 slots but, if we want to carry both Smith and Young, would only occupy 11. The rotation, which will be a four-man one, has to include Bartolo Colon, Jarrod Parker, and Dan Straily. Who's the fourth guy? It could be Brett Anderson, but I don't know. He might be better in the pen and he might just not be ready yet by that point, although that seems unlikely. Tommy Milone has been unconvincing and A.J. Griffin has been up and down this year.
In the bullpen, you have to have Ryan Cook, Sean Doolittle, and Grant Balfour, who are the team's three best relievers, by FIP and xFIP, among relievers with at least 50 IP. Next, you have to take Dan Otero, who's the team's third best reliever if you lower that threshold to 30 IP and is the best pitcher on the team against LHB. Then you add Jesse Chavez, who also does quite well against LHB.
Next, I think you take both Sonny Gray and Brett Anderson. They can be starters, but they could also be good long men or regular relief pitchers. Along the same lines, you take A.J. Griffin. And that's 11 pitchers. If we want to carry both Young and Smith, we're out of slots, and if we don't, we still only have one.
So could we function with an 11-man staff? In a short series, I think we definitely could. Bartolo starts your first game. He probably goes about five or six innings. Then either Anderson or Gray takes over for two or three innings. Parker starts the second game and Straily starts the third. The fourth game, then, could be started by either whichever of Anderson or Gray did not work out of the bullpen OR even by the one who DID work out of the bullpen. It's three days' rest, but he only worked a couple innings. Then game five is started by Colon if necessary. In fact, I think this strategy could even be repeated in a long series. In the meantime, elsewhere in the bullpen we have a couple of guys who can close, a couple of guys who can do setup work, a couple of guys who can go long, a couple of guys who can be LHB specialists.
But I don't know if Young and/or Smith are so essential that we would need to do that. So if we do want to carry 12 in the pen, who's left? And are any of THEM so essential that they're worth not carrying Young and/or Smith? In other words, while Young and Smith may not be indispensable, are any of the remaining pitchers we could carry? These are the candidates: Jerry Blevins, Pedro Figueroa, Tommy Milone, Pat Neshek, and Evan Scribner.
In his only appearance this year, he faced two batters, walking one and giving up a hit. That doesn't mean all that much, but he wasn't much good last year either, although he was pretty good against LHB in a small sample size.
I remember when Jerry Blevins was a LOOGY. This year, weirdly, he hasn't been, by FIP, but xFIP tells us that he still should have been. But his xFIP against lefties is still over 3.90, while his actual FIP against righties is 3.70. He just plain hasn't been all that great.
Nobody on the staff is worse against LHB than Tommy Milone, except for Pat Neshek. But don't worry, he can't pitch to righties either.
Of these remaining pitchers, Neshek is the best against RHB. And he's the worst pitcher on the roster against LHB.
The opposite of Neshek. He's really good against LHB, can't pitch against an RHB to save his life.
And what do you know, I find myself in the same situation as I did with the position players, not knowing which of these guys to pick. For me, it's between Neshek and Scribner, and all things considered, I want to go with Scribner. I'm not convinced, however, that keeping any of these guys on the postseason roster is worth not keeping Smith or Young. So here's the roster. A couple at the end are multiple choice, and I've bolded my personal preferences.
Kurt Suzuki/Stephen Vogt
Seth Smith/Chris Young
Seth Smith/Chris Young/Pat Neshek/Evan Scribner