In 2007, the Royals moved a talented but struggling starting pitcher to the bullpen. Two seasons later Zack Greinke won a Cy Young. Wade Davis is no Zack Greinke, but the Royals are hoping he, too, can successfully transition back into a starter after a year in the pen.
When Wade Davis was acquired as part of the Wil Myers-James Shields trade, many of us who follow the Royals had to at least give passing thought to the emergence of Zack Greinke as a dominant starter after an interim stint as a reliever.
As Craig detailed a few days back, Davis struggled in his first two seasons as a starter, but enjoyed great success working out of the bullpen in 2012. His strikeout rate doubled, Wade's home run rate was cut almost in half and his fastball velocity increased by two miles per hours. In addition to the fastball increase, Craig points out that Davis' overall approach seems to have changed as well.
The question, of course, asked over and over here and other places is, can Wade Davis transition back to the starting rotation and bring with him, if not the increased velocity, at least the new found effectiveness he enjoyed as a reliever who faced more than nine batters in an appearance just three times in 2012?
Sure, it worked for Greinke, but the gods gave Zack a thunderbolt for a right arm. Pretty much anyone who manages to make it to the majors has a good arm, but not everyone has a thunderbolt. When interested and right, Greinke is phenomenal. That he went from starter to reliever to dominant starter may not be a fair comparison for most pitchers.
We discussed some other pitchers who made a similar journey to Greinke to a small extent in the thread to Craig's article and I decided to take it a touch further and review the careers of the 88 starting pitchers who threw enough innings to qualify on the Fangraphs leaderboard for 2012. That runs the gauntlet from Justin Verlander down to Ervin Santana. (Yes, Nick, we are aware Ervin was the worst starting pitcher according to fWAR in baseball last year).
The theory of grooming a young future starter by giving him work out of the bullpen is not nearly as prevalent as it was in the good ole days, but it does still happen. Among our 88 pitchers, Chris Sale, Adam Wainwright, Jeff Smardzija, Lucas Harrell and James McDonald all worked predominately out of the bullpen when first called up to the majors. While they were mostly starters in the minors, I don't really classify this development as being the same as 'the Greinke Maneuver'.
You also have a decent sized group that includes Jake Westbrook, Bronson Arroyo, Mark Buehrle and Max Scherzer, who did some work as swing-men early in their career. Again, this type of developmental plan is not really the same as what happened with Greinke or what the Royals hope will happen with Wade Davis.
In addition, there is a portion of our 88 starters that kind of defy classification. I mean, R.A. Dickey started and relieved and started, but then became a knuckleballer, so we cannot really learn anything from him (other than it would seem worthwhile to try the knuckleball if you are on your way out of baseball). Kyle Lohse was on a downward spiral with the Twins in 2006, mangling 8 starts before moving to the bullpen for 14 more appearances. He was then traded to the National League (Reds) and has pretty much started and started effectively since. While this technically probably does deserve to be included in this discussion, Lohse's career path just doesn't feel like what we are looking at. Maybe more than anything else, Kyle Lohse is really just not very exciting to me at all and I did not feel like spending much time on him.
There are also guys like Bruce Chen and Chris Capuano, whose trips to the bullpen were mostly injury based. You also have Ryan Vogelsong, who started for two years, relieved for two years, went to Japan for three years and came back an All-Star. I would give that path a try with regard to Luke Hochevar, but only because I just don't want to see Luke pitch anywhere that I might end up having to watch.
Where does that leave us? There are some examples of pitchers who have followed the Greinke/Davis path (or more appropriately, went down that path before or at the same time). Simply working down the leaderboard, sorted by fWAR, I found seven pretty decent examples.
First up is Matt Harrison. He started 15 games in 2008, 11 more in 2009 and 6 at the start of 2010, before making 37 relief appearances to finish out the 2010 season. Harrison's average fastball velocity in 2008 was 90.3 mph and in 2009 it sat at 91.1 mph. He was not a very good starting pitcher: 5.49 and 6.11 ERAs while striking out less than five batters per nine innings. Matt's six starts in 2010 were more of the same and, frankly, early on as a reliever there was little discernible difference. However, Harrison did not pitch in the majors between June 27 and July 15 of that year and when he came back, Matt's average fastball per appearance ranged from 91.4 mph all the way up to 95.0 mph. His strikeout rate went up about half a batter per nine innings and his control disintegrated (over 5 walks per 9 innings). Harrison was different, but not any better.
Since then, however, Matt Harrison has enjoyed two effective seasons as a starting pitcher. Featuring his two seam fastball, cutter and curve more than his four seamer and changeup, Harrison has regained his control (2.76 and 2.49 BB/9 in 2011 and 2012) and manged better, not great, strikeout rates (6.11 and 5.61 K/9). In 2011, Harrison's ERA was 3.39 with an xFIP of 3.85, while in 2012 he was 3.29 and 4.13 respectively. For the last two seasons, Harrison's average fastball velocity has been between 92.4 and 93.0 and helped him account for a total 8.2 fWAR. Statistically, there are some similarities between the left-handed Harrison and the right-handed Wade Davis.
A name brought up in the earlier discussion thread and well known to all is Ryan Dempster. Between 1998 and 2003, Dempster started 123 times (out of 128 total appearances). He was pretty good in 2000, but was a starter in steady decline through 2003. We don't have velocity numbers prior to 2002, but Ryan's K/9 went from 8.31 in 2000 down to 6.54 in 2003.
Moving to the Cubs and bullpen in 2004, Dempster would spend the next four seasons working in relief. His fastball velocity went from 91mph as a starter to between 92.0 and 92.7 mph out of the pen. While, Ryan's strikeout rate rebounded, his control (a problem as a starter) remained an issue. The control, however, did get better with each year in the bullpen, going from 5.66 BB/9 in 2004 to 4.05 BB/9 in 2007.
Pitch type data is a little sketchy prior to 2007, but it appears that Dempster actually used his slider and changeup more often in relief than he had as a starter. He kept up that pattern upon his return to the rotation in 2008 and over the subsequent 160 starts since. The changeup was either ditched or reclassified as a splitter, but no matter the cause, Dempster went from a starter who threw his fastball as much as 68% of the time, to a reliever who used it as little as 45% of the time, to a reinvented starter who threw the fastball between 44% and 57% of the time over the last five years. His velocity as a starter dropped back to the 90-91 mph range, but Dempster's walk rate the last five seasons are the five lowest rates of his career (starting or relieving).
Dempster's five year-four year-five year cycle is obviously a much elongated version of what occurred with Greinke and Harrison (or Davis), but it does demonstrate an improvement that was not necessarily contingent on retaining higher velocity. Of course, in this case, it could also simply demonstrate a veteran pitcher figuring out how to pitch, as well.
Lance Lynn is our next subject and, using my logic above, he might not even belong in this discussion. Lynn was a starter throughout the high minors, but pitched out of the bullpen when called up to St. Louis in 2011 before becoming a starter once more in 2012. This puts him more in the Sale-Smardzija group, but one thing did catch my eye: strikeout rate. In two AAA seasons as a starter, Lynn struck out 7.74 and 7.68 batters per nine innings. As a rookie reliever, however, he fanned over 10 batters per nine innings and posted a 9.20K/9 rate as a starter in the majors in 2012. I did not delve any deeper than this, but here is a pitcher who was not a big strikeout guy in AAA, butywho strikes out a ton in the majors as part of and after a stint in the major league bullpen. His velocity, by the way, was basically the same, 93.2 versus 92.8, relieving to starting. Make of this example what you want.
Onto the oft traveled, seldom appreciated Edwin Jackson. Over pieces of three seasons with the Dodgers, Edwin started 14 times in 19 major league appearances. His average fastball velocity in the last of those three years (2005) was 92.1 mph and Edwin was not very effective. After being shipped to Tampa, Jackson made one April start in 2006, but otherwise split time between starting and relieving in AAA before getting the recall to Tampa in late June.
Working exclusively in relief the rest of 2006, Jackson's fastball jumped up to 94.9 mph. His strikeout rate was a mediocre 6.69 and his walk rate an atrocious 6.19 BB/9. While Edwin was a pretty awful reliever, he has been a pretty average starter (and I mean that in a good way) in the six subsequent seasons. The fastball velocity has averaged between 93.5 and 94.5 mph and Jackson uses his slider three times as much as he did when coming up: a trend that started in his bullpen season and has increased each season since.
Our next subject is C.J. Wilson. One could argue that Wilson does not really fit the profile as he started six times as a rookie, appeared as reliever 21 times that year and then worked exclusively in relief for the next four seasons. To be honest, I included him more for the dream that Aaron Crow could someday become a starter more-so than as hope that Wade Davis can become a reinvented starter. Long story short, Wilson threw his fastball 75% of the time and at 93 mph as a reliever. As a starter, Wilson throws the fastball 91 mph and just 50% of the time. Like Lance Lynn's resume, I am not sure what this tells us other than it gives the 'Crow to the rotation' crowd a little ammunition - which is not a bad thing.
Okay, we are down to our final two pitchers: Justin Masterson and Phil Hughes.
Masterson is another guy who does not exactly fit the mold. He started his first nine major league games, then relieved in 25 more games in 2008. His average fastball velocity that year was 90.8 mph. In 2009, split between Boston and Cleveland, Justin started 16 games and relieved in 26 more with an average fastball velocity of 92.9 mph. Once with Cleveland, Masterson has been a full-time starter with fWARs of 2.5, 4.7 and 2.3. During those three years, the fastball has sat between 92.4 and 93.1 mph. A lot of this could simply be maturing as opposed to anything related to the bullpen and truthfully Masterson's career probably fits more in with the swing-men.
In the case of Phil Hughes, he started 13 games in 2007 (7.18K/9, 3.58 BB/9) and 8 more in 2008 (6.09, 3.97), with a fastball sitting right at 91 mph both years. In 2009, the Yankees used him 44 times out of the pen and 7 times as a starter. Hughes, with a fastball now sizzling in at 93.7 mph, struck out 10 batters per nine innings and walked 2.93 over every nine frames.
In 2010 and 2012 (we will ignore an injury riddled 2011) as a starter, Hughes's velocity was 92.5 and 92.0, his walk rate was 2.96/9 and 2.16/9 and his strikeouts remained decent at 7.45/9 and 7.76/9. It is interesting to note that after using a cutter quite a bit as a reliever and almost 19% of the time as a starter in 2010, Hughes reverted to his pre-bullpen mix of fastball, curve an changeup in 2012.
While this group of pitchers each found relative success in a variety of different ways, they all did move from being starters to the bullpen and then back into the rotation. While there is not a Zack Greinke in the group (other than Zack himself), they are all pitchers who can help good teams win games. It was not all about retained velocity, although in most cases each had at least a little extra speed in their second starting stints. Almost all exhibited some measure of increased control and most featured, at least superficially, some change in their pitch usage/mix.
Hope for Wade Davis? Yes, I have some, but I am also aware that there is a likely a larger group of pitchers who never made it back to the rotation or were not effective once they did. Color me cautiously optimistic, if only because spring is coming and hope springs eternal.
Even for Royals' fans.