Friday, June 29, 2007

Memo: Pierre still attempting to bat against LHP

When Juan Pierre was signed, I suggested that it could have been an okay signing if Pierre were platooned. I said this because a) Pierre, though his split may be less pronounced since he's a contact hitter with almost zero power, still is a better hitter against RHP, b) Pierre might play better if he was getting a day or two of rest per week, and c) starting any LHB against a southpaw needs to be evaluated in light of the opportunity cost of letting a RHB play that day - even if you had a LHB with no true platoon split, starting a generally inferior RHB is often preferable because your RHB gets the platoon advantage.

As it turns out, Pierre has hit a semi-tolerable .291/.328/.352 against RHP and a miserable .217/.233/.229 against his fellow lefties. Is there a reason that this hasn't been considered a scandal? Now, you should know as well as I do that this is a small sample that doesn't fully represent his true platoon split, but the point is that Pierre isn't a good enough hitter that you'd ever want to start him against LHP unless you were painted into a corner by injuries and/or roster problems. Of course, were the Dodgers to defend themselves against this charge (of course, they aren't since as far as I know no one in the media is advancing the Platoon Pierre cause), they'd probably spout something about not having an RHB centerfielder. And yes, the Dodgers have three starting OF who are left-handed (Gonzalez and Ethier the others). But those other two are much better hitters than Pierre, and in Ethier's case we have reason to believe that he has an atypically small platoon split. So why haven't the Dodgers been sitting two of their southpaws against southpaws? For one, they've largely scorned having a fifth outfielder, which for reasons beyond merely the title of this blog I consider to be a lousy strategy, especially if you're going to couple it with the leftrinity. Why no fifth outfielder? We could chalk it up to the Repko injury, but he was replaced with Brady Clark, who was DFA'd for essentially no reason. Plus, the Dodgers have a fine righty platoonmate at Vegas already in Delwyn Young, who is not good enough that you can really argue they need to keep him there to play everyday. The real reason boils down to the Dodgers decision to resign Nomar Garciaparra, which gave them two right-handed first basemen; now that they've finally given Loney the job they should have given him to begin with, Garciaparra and Saenz are fully redundant as RHB pinch-hitters slash first base platoonmen. Of course, the Dodgers aren't using Nomar as such, which instead means that Betemit is relegated to a bench role, meaning the Dodgers have 40% of their bench dedicated to no-field corner infielders.

I call shenanigans on the whole thing; I was arguing that the Dodgers shouldn't re-sign Garciaparra back when he was having a great 2006, going so far as to argue that they should try trading him. That signing is looking downright albatrossic, which is pretty tought for a two year deal under $20 mil. But despite knowing that they had Loney waiting and major league ready, the Dodgers still shelled out the bucks to Nomar and gave him a no-trade clause. Yuck.

But in any event, even the fallout from Garciaparra isn't enough backdoor justification for the Dodgers to screw around on this Pierre affair. Put Kemp in center against the southpaws, already. They can't possibly want to continue to make the Pierre signing look this awful, can they? Might as well boost his rate stats ... and, you know, the team's ability to win.

Bizarre Stat Use of the Day

Why should anyone take something on MLB.com seriously enough to write about it or even research it? No reason. I just did.

The site's frontpage promo for tonight's Diamondbacks-Giants game reads "Bonds faces one of his favorite opponents: Of his 749 career homers, Barry Bonds has hit 38 against the Diamondbacks." 38 out of 749 hardly seems impressive for a team in the Giants' division, but then again the D'backs were an expansion team, so maybe it's more significant than I thought. Here are Bonds career numbers against Arizona compared to everyone else:
Opponent        Avg  OBP  SLG  HR/AB  HR/(PA-IBB)
Arizona .311 .493 .689 .100 .081
Everyone else .298 .442 .605 .763 .063
Ahh, maybe MLB.com is onto something. Except that, you know, Bonds has been a much more productive hitter since Arizona entered the league. While only 40.3% of Bonds' PA have come since the D'backs entered the league in 1998, the majority of his home runs (375) have come during this period. Here's how Bonds' performance against the Arizona squadron stacks up against everybody else since 1998:
Opponent        Avg  OBP  SLG  HR/AB  HR/(PA-IBB)
Arizona .311 .493 .689 .100 .081
Everyone else .317 .498 .705 .103 .082
So if Arizona is in fact Bonds' favorite opponent, it is not because he has had more success against them; indeed, he has had less success against Arizona.

Wednesday, June 13, 2007

Hong-Chih Kuo

I must say, Kuo is, to me, among the most enigmatic players in baseball. Kuo now has a total of 85 IP at the major league level and 169.2 in the minors. He's spent most of his career sidelined by injury, and while the backstory is very different, he's similar to Josh Hamilton in that it's very difficult to feel at all confident in projecting his performance. Kuo obviously has talent, as he's struck out more than a batter an inning at every stop in his career outside his present cup of coffee. At the same time, he's had very high BABIP's at every level. While I've generally known a lot about most players coming up the Dodgers' system, Kuo kind of slipped through the cracks because of his injuries, and so he wasn't on my radar, even after his September call-up in 2005. It wasn't until he showed up in the Dodgers bullpen in April 2006 that I really took notice, just kind of dawned on me in 2006 that and even then, he didn't seem all that noteworthy. He wasn't even doing all that well in Vegas, and while I probably should have noted the obscene number of strikeouts he was accruing, I didn't. Then all of a sudden he was converted to starting, which sure as heck didn't make a lot of sense - why convert him to a tougher role? But suddenly, Kuo's overall numbers improved, and he was back in LA, this time as a starter.

It has now become commonplace among LA fans to assert that Kuo can only start, not relieve. One does not often hear this said of pitchers, and normally when it is said the logic is off. In this case, though, it may be more accurate. Kuo's major league numbers as a starter are 47 IP, 2.68 ERA, 41 H, 50 K, 13 uBB, 1 HR, for a 2.18 FIP of and a szERA of 3.06 as well as a .317 BABIP. As a reliever, he's got 38 IP, 6.16 ERA, 37 H, 46 K, 27uBB, 3 HR, a 3.80 FIP, a szERA of 4.10, and a .347 BABIP. Obviously, the outcomes have been considerably better for him as a starter. And since his split has been the opposite of the typical - essentially, his split would be about normal if his starting numbers had instead come in relief and his relief numbers came as a starter - it's certainly fair to believe there's a good chance he's better as a starter.

However, I'd argue that Kuo probably is not somebody who's much better as a starter than as a reliever - his split would probably be smaller than typical, but I suspect that over a suitable sample he'd actually do better out of the bullpen. Others will surely protest that he's just better suited for the rotation, biorhythmically or in some such way. I think, however, that that's a post hoc quasi-scouting observation. I think it's more likely that Quo improved when he converted to starting because he was so inexperienced. Kuo obviously has quite a bit of talent, but was inexperienced enough that he had a tough time above AA and was walking more batters and perhaps setting up batters to hit the ball better. But maybe Kuo simply became a better pitcher when he started because it was the first time he was logging a considerable amount of innings, and he was learning much more than he had been as a reliever. His starting stint was all at the end of the season, and for someone with such a tiny amount of minor league work, we would expect the in-season experience to matter quite a bit. Or, rather, I think we would expect that. While some will point out that Kuo has pitched better as a starter this season than in relief, that kind of ignores that his relief stints were when he was just coming back from a shoulder injury and that the sample size is close to meaningless.

While we might possibly have sufficient data to conclude that Kuo is better as a starter, that's not the data that most are using to paint Kuo as starter-only material. That is, the very granular data might indicate that he doesn't gain the typical relief advantage and in addition gains other advantages by having a better approach while starting. Tentatively, I think that that's just over-reacting to a little bit of data. In reality, I think Kuo is probably a good starter who would be a good reliever also, and hence, like essentially any pitcher, Kuo probably in reality would put up better numbers in a full-time role than in the rotation.

That's not to say Kuo should be in the bullpen. I think it's probably best for all involved parties to have him in the LA rotation. He seems as if he's going to be (and is already) a very good pitcher, and he should be better than Tomko and Hendrickson. However, I caution against the Dodgers (and their fans) deciding that Kuo is only good as a starter. What if Kuo, pitching in the rotation, goes through a rough patch, and starts to look like he's more of a 4.00 ERA pitcher for the time being? And let's further assert that Billingsley continues to pitch brilliantly in relief. I fear that the Dodgers might stick with Kuo in the rotation while choosing to use Billingsley in a smaller role - and, given the strength of their bullpen, a relatively low-leverage role. I fear they would decide that Kuo only has value in the rotation, and conclude that Billingsley is as valuable in the pen as he would be in the rotation, when in reality both of these propositions are unlikely. It's obviously not a terrible problem to have, but there's no point in making rules for Kuo based on inadequate data, and I fear that the offshoot of such roles would be additional bizarre rules for Billingsley.

While I am curious to see whether Kuo really is much better as a starter, there's a good chance we'll never find out. Ideally, he will do so well as a starter that no one will ever think to try him in relief. Then again, he could flame out in the rotation this year, end up as a reliever and never being considered good enough to merit a starting chance again. He could also get tossed from the rotation for a poor stretch, catch fire when demoted to the bullpen, and we could end up with the rationalization that he was better in the pen all along - and, as with Gagne, we'll never find out if he could have made it as a great starter because people will be so attached to his relief performance. And unfortunately, given Kuo's history, we have to consider that a future injury will mean that we're about to learn even less about Kuo than the small record he has already accrued.