The Texas Rangers are considering moving reliever C.J. Wilson into their starting rotation, and he has done nothing but impress them this spring. Of course, one shouldn’t put undue (i.e., any) weight in spring training performances, but there are other ways of gauging whether Wilson might make a decent starter.
I first want to make clear that I’m not going to address two important issues: 1) How much of an increase in workload Wilson may or may not be able to handle, and 2) whether or not Wilson is more valuable in the rotation than in the bullpen. I will just be focusing on his chances of making the transition performance-wise (other than endurance).
How good would Wilson be as a starter? CHONE projects him to have a 3.70 FIP in 2010, and ZiPS projects him for 4.17 — an average of about 3.94. Both of those projections are based on Wilson’s performances as a reliever, and as a general rule one estimates that a reliever will be one run per 9 innings pitched worse as a starter. A 4.94 FIP isn’t replacement level, but even in Texas’ hitter-friendly home park, that’s nothing to get excited about other than as a stopgap or back-of-the-rotation type.
There are other reasons to be doubtful. For his career, he has pretty big platoon splits, posting a 3.39 FIP (3.36 xFIP) versus lefties, and a 4.58 FIP (4.26 xFIP) versus righties. While relievers can be put into games so as to maximize their platoon advantage, this is not the case for starters, and most of the time, hitters would have the advantage against Wilson.
Finally, while pitch-type linear weights don’t tell us everything about the quality of a Wilson’s repertoire (for example, sequencing), they do tell us something. Of Wilson’s primary offerings the last few seasons, only his fastball has been clearly above average, and this is important for starters, who have to face batters more than once.
The factors cited above tell against the likelihood of Wilson being very good as a starter. However, projection is always a tricky business, particularly in the case of pitchers, whose true talent is generally subject to more changee than that of hitters. Moreover, there are some interesting recent developments in Wilson’s case. So, with all the usual caveats, two things stand out to me from Wilson’s 2009. First, while Wilson has always been a favorable groundball/flyball ratio (a good thing anywhere; even if sometimes the ball goes towards Michael Young that’s better than it going out of the park), in 2009, he took it to a new level, at 2.25. Second, in 2009, Wilson’s platoon split was remarkably even: 2.81 FIP (3.60 xFIP) versus lefties, and 2.94 FIP (3.05 xFIP) versus righties.
Normally, I’d be pretty skeptical, primarily because it relies on just one year of data (and from a reliever-sized sample, at that). But there might be something more at work here. Some of the Rangers’ pitchers (including Wilson) have added a cutter. Again, usually I’d be wary — “adding a new pitch” is right up there with “in the best shape of his life” for Spring Training stories. But in Wilson’s case, it has some substance. According to his pitch types, his pitch use and selection changed in 2009. While he threw about the same number of changeups as before, he seems to have thrown his fastball less in favor of more sliders and cutters. While earlier I said that over multiple seasons the fastball was his only consistently effective pitch, there is the possibility that adding a cutter and throwing more sliders might have increased the overall effectiveness of his pitches (other than the still below-average changeup). The cutter, in particular, might help explain his greatly improved platoon splits in 2009.
I don’t have a firm conclusion. At first glance, we wouldn’t expect Wilson, a capable reliever, to be that effective as a starter. However, if the new cutter has not only helped his platoon issue but has expanded his repertoire so that he can to get through the order more than once (and perhaps get even more groundballs), then — provided he can handle a starter’s workload over a full season — the Rangers might have something more than just a stopgap starter on their hands.*
* Thanks to David Appelman and pitch f/x expert Mike Fast for their helpful responses to my emails on this subject. Neither of them, of course, should be held responsible for any mistakes and/or misguided analysis in this post.