The Yankees need help in the bullpen, and the solution being forwarded by the loudest voices is reinstalling Joba Chamberlain as a dominant setup man and putting Chien-Ming Wang back in the rotation as soon as Wang has proven he's over his early-season woes.
Wang has been effective of late. So if you subscribe to that approach, the time is now because Wang looks just about fixed. But since Wang's pitched well in relief, why not just leave him there as the eighth-inning man and keep Joba in the rotation?
As we all must acknowledge in New York, that loudest voice is Mike Francesa of WFAN. Judging by this transcript, his voice is also the rudest. Let's just say that Francesa feels very strongly that the Chamberlain-as-starter "experiment" has been a failure. The overall stats don't back that up, but that's of no consequence to those who do their thinking via dialog with their talking gut. But let's instead think with our heads and see what the data say.
Before yesterday's game, Chamberlain had a 3.25 ERA as a starter with 120 K's and 50 walks in 110.7 innings. His walks plus hits per innings pitched, or WHIP, is a not-so-good 1.410. But his batting average allowed on balls in play (BABIP) is very high at .336. The average is .300 and Chamberlain as a reliever is .286. I'd expect that to come much closer to average as his sample size as a starter get larger.
Chamberlain's K-rate (9.4 strikeouts per nine innings) and strikeout-to-walk ratio (2.40) are dominant enough as a starter. On Monday, the right-hander went eight solid innings, allowing just six baserunners and two runs in picking up the win. And he was throwing 98 mph in the eighth inning, too.
To be fair, we also must note that his control this year has been poor, and his May ERA was 4.84 despite 11.7 strikeouts per nine innings. In May, his BABIP was .385, which is horribly unlucky.
Overall, the data say Chamberlain has been a good starter separated from being very good (solid No. 2) mostly by bad luck on BABIP.
The other half of the equation here is putting Wang in as a starter. I first looked at this question back in 2007. Unbeknownst to me at the time, Dan Szymborski at BaseballThinkFactory.com had a formula of his own. And since then, Tom Tango on his Web site and Harry Pavlidis at the TheHardballTimes.com have looked at the question, too. They have done it from the perspective of converting a reliever (Joakim Soria) into a starter. We all agree that the same pitcher is most likely to pitch significantly better as a reliever than as a starter. The question is how much better?
Tango and Pavlidis deal mostly with Wins Above Replacement, not simple enough for our purposes. I want to convert the stats that we all understand.
My Wang conversion (based on how he's pitched as a starter for his career) is a 2.89 ERA, 5.71 K/9, and 1.08 WHIP (baserunners per inning). Of course, he'd still be an extreme groundball pitcher. (Note: my conversion formula is based on how a sampling of modern starters turned relievers actually performed through 2006.)
Szymborski says Wang as a reliever would have a 3.09 ERA with 5.4 strikeouts per nine innings and a 1.21 WHIP.
Wang as a reliever to date has logged eight innings with a 2.25 ERA, 7.9 K's per nine and a 1.38 WHIP. The high whip is mostly a product of a .364 BABIP. (Tell Francesa, "It's all Derek Jeter's fault on the balls in play," and his head explodes.) I also note that Wang's velocity of late has been creeping up, another good sign.
If we average all that out, you have a capable setup man who will get into trouble with seeing-eye singles but often get out of it with double plays. Plus, there's a chance we're understating his dominance upside given the early results. For a comparison, think Roger McDowell, who was a one-pitch, ground-ball pitcher who not only set up but often closed for a championship team (the 1986 Mets).
I've gone back and forth regarding what the Yankees will do regarding Chamberlain. I think the odds still greatly favor Chamberlain's being moved in August as he gets near 150 innings. This crazy innings obsession is the perfect excuse for them to make the safe move that will the 50,000-watt second guessing that has been known to cost sports executives in this town their jobs.
But based on Phil Hughes' showing enough at least to be a No. 5 starter and Wang's early success in the 'pen, the Bombers will be a better team with Chamberlain getting more innings in the rotation and Wang pitching more effectively in the pen. Wang may not like it, but his desire to start should not factor at all into this decision. And I would not concern myself at all with Chamberlain's innings because there's not evidence that increasing workload increases injury risk. It's a theory looking for facts when that's really supposed to work the other way around. But that's a story I suspect we'll revisit in the near future.