According to FanGraphs, C.J. Wilson was the ninth-most valuable reliever in baseball in 2009 after accumulating 2.0 wins above replacement (WAR). He appeared in 73 games, tossed 73.2 innings and posted a 2.81 ERA, 10.3 K/9, and 3.9 BB/9. Following the season, the Texas Rangers reportedly agreed to C.J.'s request that he be allowed to compete as a starting pitcher during spring training in 2010.
Most die-hard Rangers fans (including many of you who responded to Joey's Sunday questions on November 8th) believe that Wilson should remain in the bullpen. After all, Wilson proved to be a very reliable pitcher in what proved to be a very effective bullpen in 2009. In theory, removing Wilson from the team's 2010 bullpen would likely weaken that unit. Being a bit of a contrarian, I wanted to get a better look at whether "C.J. -- Starting Pitcher" is as crazy as it sounds. Below are answers to what I believe to be key questions regarding C.J.'s proposed career change.
QUESTION NO. 1 -- IS C.J. AN INDISPENSIBLE COMPONENT OF THE RANGERS BULLPEN?
Wilson was the Rangers' most valuable reliever in 2009, outpacing Darren O'Day (1.4 WAR), Frank Francisco (1.1 WAR) and Neftali Feliz (1.1 WAR) for the honor. The year before, Wilson was one of the bullpen's least effective relievers at -0.4 WAR in 46 innings. He was a 1.0 WAR pitcher in 68 innings in 2007 and a 0.2 WAR pitcher in 44 innings in 2006. Like most relievers not named Mariano, it is difficult to predict whether C.J. will be irreplaceable, useful, or a disaster pitching out of the 'pen in 2010.
Concerns about losing the bullpen's one reliable left-handed reliever seem unfounded, since Wilson was not really used as a situational lefty in 2009. Almost all of his appearances came in the eighth or ninth innings, and he faced almost twice as many right-handed batters (185) as left-handed ones (97). In theory, Feliz (if he remains in the bullpen), O'Day, or Dustin Nippert could fill C.J.'s late-inning role in 2010 without having problems arise due to their being right-handed. In addition, Matt Harrison, Kasey Kiker, A.J. Murray, Zach Phillips, Corey Young, Michael Kirkman, and Beau Jones all appear to be viable candidates to fill in if left-handed relievers are needed in Arlington.
Answer: Making C.J. a starter will reduce the talent available for the 'pen in 2010, but it seems unlikely that Wilson's presence is required for the Rangers' bullpen to be a good one.
QUESTION NO. 2: COULD C.J. BE AN EFFECTIVE STARTING PITCHER?
Adam Wainwright did it. So did Ryan Dempster. And how different would the 2009 Rangers have been without Scott Feldman working every fifth day? Although not a common occurrence, relief pitchers can succeed while making the switch to the starting rotation.
The transition for both Wainwright and Dempster appears to have been simplified by the fact that they were three-pitch relievers. Wainwright's pitch mix (55% fastball/18% slider/20% curveball) remained unchanged when he switched from relieving to starting in 2007. Likewise, Dempster's pitch selection and usage changed very little between his time as a closer (2005-2007) and a starter (50-60% fastball/30% slider/15% change-up).
Feldman's transition from the bullpen to the rotation required that he add a cut fastball to reduce his reliance on his four-seam fastball from 75 percent as a reliever to 55 percent as a starter. In his two years as a starter, Feldman has thrown his cutter roughly 23 percent of the time and his curveball roughly 19 percent of the time while mixing in an occasional slider or change-up.
In 2009, Wilson relied primarily upon his fastball (70%) and slider (19%). Both pitches rated as above average with values of 0.92 wins and 0.42 wins per 100 times thrown. According to FanGraphs, Wilson began throwing a cut fastball in 2009 that rates as an above-average pitch (0.34 wins/100 times thrown). He apparently used the pitch around five percent of the time, which is reminiscent of Scott Feldman's cutter usage in 2008. He will likely need to use his new pitch at least 10 percent of the time in order to bring his fastball usage down into the 60-65 percent range that most starters use.
It is worth noting that Wilson has had success as a starting pitcher. In 403 minor league innings, he posted a 3.53 ERA, 1.26 WHIP, 3.0 BB/9, 7.5 K/9, and 0.6 HR/9 primarily as a starter. As a major leaguer, Wilson has a 4.30 ERA, 1.33 WHIP, 4.1 BB/9, 8.4 K/9 and 1.99 GB/FB. In 2009, his walk rate was 3.9 BB/9, which is not ideal for a starter, but he made up for it with a strikeout rate of 10.3 K/9 and a GB/FB of 2.25.
In addition, he proved to be almost as good against right-handed batters (.249/.329/.373) as left-handed ones (.206/.310/.237). Using the chart shown here, Wilson would likely have an ERA of around 4.25 as a starter if he can maintain the walk and ground ball rates that he notched in 2009 while sustaining a drop in his strikeout rate to around 8.0 K/9. In 2009, that would have ranked Wilson alongside Tommy Hunter for the third-best ERA on the staff.
Answer: Assuming Wilson can increase his use of the “cutter” that he apparently introduced in 2009, then it seems likely that he will have the three-pitch repertoire that most successful starting pitchers feature Based on his career statistical profile, it seems likely that Wilson could maintain the ERA of a solid No. 3 pitcher.
QUESTION NO. 3 - WOULD WILSON BE MORE VALUABLE AS A STARTER OR RELIEVER?
Good starting pitchers are more valuable to their teams than good relievers because starters typically impact two to three times as many innings. The very best relievers in baseball rarely exceed 3.0 WAR in a season (Broxton led MLB relievers in 2009 with 2.9 WAR and Mariano Rivera finished with 2.0 WAR). In 2009, 49 starting pitchers finished with 3.0 WAR or greater. Among the three converted relievers discussed above, Wainwright has averaged 3.8 WAR as a starter after being a 1.1 WAR reliever. Dempster has been a 4.3 WAR starter after averaging 0.8 WAR as a reliever. And Feldman has averaged 2.1 WAR as a starter after beginning his career as a 0.1 WAR reliever.
If C.J. remains in the bullpen, his greatest value to the Rangers would likely result from him being designated the closer. Unfortunately, Wilson has proven to be relatively ineffective when pitching on consecutive nights. In 2009, Wilson pitched on back-to-back days 21 times. In the 18 innings he pitched on the second day, he gave up 20 runs (16 earned) and permitted an opponents' batting average of .325. As noted in the graph below, Wilson's career numbers indicate that he is far more effective when rested. Unfortunately, that is a luxury that closers are not allowed.
Assuming Wilson pitches out of the bullpen but is not the closer, and assuming that he remains healthy throughout 2010, then it is reasonable to expect a 1-2 WAR performance. Fifty-nine percent of the 123 starting pitchers who tossed at least 100 innings in 2009 exceeded 2.0 WAR. If Wilson joins the ranks of starting pitchers and is able to give the team 150 innings of 4.25 ERA baseball, then he will likely produce at least 2.0 WAR with the potential to be significantly more valuable if he pitches more innings.
Answer (and another question): Although he needs to prove that he can be at least moderately effective as a starting pitcher, it seems very likely that Wilson's personal value to the Rangers in 2010 will be greater as a starter than as a reliever. However, the key question is whether the value of “Wilson + his bullpen replacement” will be greater than the “starter that Wilson replaces + Wilson in the bullpen.” The answer to that will depend on how effective the Rangers are this off-season in acquiring major league pitching and how their young pitchers progress before and during spring training.
Conclusion: I like the fact that the Rangers are allowing Wilson to compete for a spot in the starting rotation. The best-case scenario nets the Rangers a starting rotation that includes two hard-throwing left-handers (Wilson and Derek Holland) who will complement three solid right-handers (Feldman, Kevin Millwood, and Tommy Hunter, Brandon McCarthy, or Feliz).
If C.J. fails to secure a job in the rotation, then he either reprises his role in the major league bullpen or the Rangers burn his last option so that he can spend time in the minors either getting his reliever mojo back or working on his starter's repertoire. And on the off chance that Wilson spends a couple of months in the minors, then he postpones his free agency by a year and allows the Rangers to keep the lefty at least through the 2012 season.