When the Detroit Tigers acquired Phil Coke in early 2010, they floated the idea of using him as a starter. Although he had appeared exclusively in relief as a major leaguer, Coke worked as a starter in the minor leagues until his promotion to Triple-A in 2008. Ultimately, the Tigers decided to kept him in the bullpen for the 2010 season (although he did make one start at the end).
Coke pitched well enough as a reliever (3.76 ERA/3.23 FIP) for Detroit to re-visit the idea of moving him to the rotation again this spring. This time, the Tigers followed through on their plan and Coke made 14 starts sandwiched around a brief stint on the disabled list. After dropping to a 1-8 record with a 4.91 ERA following his last start, Tigers manager Jim Leyland announced Coke is moving back to the bullpen.
Coke’s record and ERA are pretty ugly. Meanwhile, his 3.73 FIP is more than admirable — however, fueled by an extremely low home-run rate. With 4.75 xFIP this season and a 4.34 career mark, Coke appears to be one of those pitchers who may out perform his xFIP because of home-run rate — especially if he remains in the bullpen. In 216.1 career innings, he has allowed just 15 home runs. It also helps that has pitched most of his big league career in one of the more spacious American League parks and his usage until 2011 was largely dictated by match-ups.
Prior to 2011, Coke averaged just under 7.5 strikeouts per nine innings out of the pen. As a starter, his K/9 dropped down to 4.46. This coincides with slide in swinging strikes as well (12.0% in 2010, 6.9% in 2011). On his four-seam fastball in particular, his whiff rate dropped from 7.8% a season ago to 3.1% in 2011. You could point to a few reasons why the strikeouts disappeared including: loss of velocity as a starter, facing a lineup multiple times a night, league wide exposure, as well facing the platoon split more often.
Last season, Detroit actually used Coke more against right-handed batters. On the other hand, his career numbers profiles more as a LOOGY than a neutral reliever (5.05 xFIP vs. RH, 3.50 vs. LH). I’m not sure what Detroit saw in his profile, but the larger workload versus righties did not help Coke’s chances as a starter. While his 2011 FIP of 3.90 against righties looks acceptable, it is once again propped up by a strong home run rate. He is holding lefties to a .185 average and a .219 BABIP, but righties have hit .315 against him with a .335 BABIP. That’s a rather significant split considering he faced more than double the amount of RHB (235) than LHB (104). In addition, a large chunk of Coke’s strikeout issues are attributed to facing more right-handers – who struck out just 19 times against him while taking an equal amount of walks.
Converting from a relief pitcher to a starting pitcher is not an easy task. C.J. Wilson made it look that way, but he is more the outlier than the rule. Coke was able to maintain a solid walk rate and continued success at keeping the ball in the park as a starter, but his stuff – most notably his fastball – did not hold the test and he was vulnerable to opposing managers loading the lineup with right-handed batters.
Still, the end result of the experiment are as bad as Coke’s 1-8 record indicates. Considering the lack of options that Detroit had for the rotation coming in to the season, it was likely a worthwhile effort. For all the good he did as a reliever in 2010, he was worth 1.1 WAR. Although he stumbled in 2011, he narrowly passed that mark (1.2 WAR) in half a season as a starter. The updated ZiPs projections have Coke with a 3.74 FIP in nearly 150 innings of worth. That would not be C.J. Wilson good, but a solid back-end of the rotation piece. Meanwhile, actually maintaining that FIP as teams continue to exploit the platoon advantage – or even keeping up the starter’s workload – is another story; one that appears to have ended before we have answers.