One of the lead stories on the nearly static Post-Dispatch Cardinals page of the last week has been "Kyle McClellan in the mix for a spot in rotation." The Cardinals have been known to encourage their relievers to train "long" in spring training — that is, to prepare as starters — based on the idea that it’s easier to scale back to fewer innings in a relief role when the season starts than to go the other way. But to the extent that McClellan may be doing more than simply going through a routine preparation, what might the team and fans expect from him as a starter?

The most recent analog for a reliever-to-starter conversion project is of course Braden Looper, whom Cardinal pitching coach Dave Duncan helped create almost as much value starting as he did relieving. That was because although Loope had a worse ERA as a starter, he pitched more innings. In his three seasons prior to becoming a full-time starter, he averaged 2.1 WAR; in his subsequent three years as a starter, he averaged 1.9. But that’s only one example. In the last 10 years, nine pitchers have started at least 75 games and relieved in 200 (not necessarily following Looper’s path of exclusively relieving to exclusively starting). Curiously, five of them have pitched for the Cardinals:

And their career splits by role:

And the differences between their stats as starter and as reliever:

The chart shows what most people already knew: That, for the most part, pitchers are more effective in short outings (i.e., as relievers): The group of nine had a 20% lower ERA, 20% lower HR/PA and 28% higher SO/BB as relievers than as starters.

What does that mean for McClellan? First, let’s see how much value he is expected to provide as a reliever. His average projected FIP for 2010 — as a reliever — is about 3.93 over 65 innings. That’s roughly .4 WAR, about the same level he attained in 2008 and 2009 (.4 and .3, respectively). Though no one — including Dave Duncan — can predict what he would do as a starter, it’s reasonable to think that he would perform similarly to those who have gone before him. So if we adjust his FIP upward by that ratio, that yields a Todd Wellemeyer-like 4.91 FIP. Though we’re sure we don’t need to remind anyone what fans thought of the Colonel in the starting rotation, for the sake of argument that comes to about a -0.3 WAR .5 WAR, assuming a modest 120 innings pitched as a fifth starter. The reason appears to be that, with a 4.91 FIP, McClellan would pass the break-even point for returning positive value as a starter reliever, but he is below that threshold as a starter, where the breakeven is more like 5.50.

Looking at the numbers of those versatile hurlers, one has to ask if Ryan Franklin would be a better bet to start than McClellan. Some reasons:

• Franklin has already done it at the major-league level.
• Franklin performed as well as Looper as a starter.
• Franklin doesn’t demonstrate any marked improvement over how this set of pitchers performs as relievers.

None of that is to say that Franklin should be rotation material — just that he makes as much sense as McClellan, which is to say not much.

Perhaps McClellan could take on Brad Thompson’s old role, pitching in relief and taking an occasional spot start. But that seems untenable, since Thompson’s relief duties were limited to long (a.k.a. mopup) outings, whereas McClellan seems to have more value to the team as a frequent-use setup man. But as far as getting more value from him in the rotation goes, La Runcan would have to believe they can accomplish something that is pretty hard to pull off. As George Constanza once asked, "Where do you get the ego?" (punchline is no longer funny in the face of those pesky facts!).

This entry was posted by Pip on Tuesday, February 9th, 2010 at 3:03 pm and is filed under analysis. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.