Carlos Martinez, strikeouts, and the shift from relieving to starting

It started with a comment on Effectively Wild, as many of my baseball thoughts do. Sam and Ben were talking to Howard Megdal about the Cardinals, and Megdal was opining on freshly-minted rotation piece Carlos Martinez. Martinez was both a highly-touted prospect and a relief pitcher in many of his MLB appearances to date, but for 2015 he'll start the season on the bump every fifth day.

Without getting into Megdal's exact words, he offered up an opinion that immediately set off alarm bells in my head -- that Martinez's filthy stuff would translate to more strikeouts now that he was moving to the rotation from the 'pen. As a good sabermetrician, this suprised me ... Dan Szymborski ran the data in 2003, Tom Tango ran the data in 2009, and Jonah Keri compiled some of the common wisdom back in 2012.

What we think we know is pretty simple: converting to starting pitching tends to make a pitcher (1) less effective and (2) takes a little velocity off their arsenal. It's all very common sense-based: relievers can amp up the velocity in shorter appearances, velocity leads to strikeouts, strikeouts lead to effectiveness. Simple.

However, some relief-to-starting experiences go the other way. In fact, just last season, Garrett Richards logged 41 innings as a reliever and saw his strikeout rate surge the following season as a full-time starter. It's not like a reliever can't see his strikeout rate go up as a starter.

So, I wanted to take a look at reliever-to-starter conversions from the last 10 years, in order to see if there were some case studies similar to Martinez's situation in the past. I used FanGraphs' leaderboards to pull all the guys who relieved at least 20 innings in year one and started at least 50 innings in year two, starting with 2004 as the first relief season. I came up with 128 seasons, and you can find them all in this spreadsheet.

How Many Pitchers Had More Strikeouts as a Reliever?

Of the 128 seasons I looked into, 94 seasons saw the pitcher lose strikeouts in the conversion to starting, 33 saw the pitcher's K% increase, and one season -- Sweet Joel Piniero in 2008 -- saw his K% stay almost exactly the same. That gives us a roughly 26.5% chance that Martinez's strikeout rate will increase, if we just base that on the previous sample. Of course, I'm not so sure we should do that.

Which Pitchers Saw The Largest Increase in K% Upon Converting?

These 15 pitchers did.

Name	Season Y1	Season Y2	Team Y1	Team Y2	K% Y1	K% Y2	K% D	SP / RP Y1	SP / RP Y2
Jeremy Hefner	2012	2013	Mets	Mets	8.7%	17.8%	9.1%	RP	SP
Garrett Richards	2013	2014	Angels	Angels	16.0%	24.2%	8.2%	RP	SP
Tyson Ross	2013	2014	Padres	Padres	17.3%	24.0%	6.7%	RP	SP
Anthony Bass	2011	2012	Padres	Padres	14.7%	21.0%	6.3%	RP	SP
Dustin Nippert	2008	2009	Rangers	Rangers	13.3%	19.2%	5.9%	RP	SP
Chad Gaudin	2006	2007	Athletics	Athletics	13.0%	17.4%	4.4%	RP	SP
Jason Hammel	2008	2009	Rays	Rockies	13.0%	17.1%	4.1%	RP	SP
Luis Mendoza	2012	2013	Royals	Royals	10.3%	14.3%	4.0%	RP	SP
Glendon Rusch	2004	2005	Cubs	Cubs	12.4%	16.0%	3.6%	RP	SP
Edwin Jackson	2006	2007	Devil Rays	Devil Rays	13.9%	17.0%	3.1%	RP	SP
Matt Harrison	2010	2011	Rangers	Rangers	13.4%	16.5%	3.1%	RP	SP
Chad Gaudin	2012	2013	Marlins	Giants	18.9%	21.7%	2.8%	RP	SP
Carlos Villanueva	2011	2012	Blue Jays	Blue Jays	20.2%	22.9%	2.7%	RP	SP
Dallas Keuchel	2013	2014	Astros	Astros	15.4%	18.1%	2.7%	RP	SP
Dan Haren	2004	2005	Cardinals	Athletics	15.7%	18.2%	2.5%	RP	SP

The guy atop the leaderboard, Jeremy Hefner, didn't pitch very often as a reliever and had a wildly poor strikeout rate as a starter. He was just, I dunno, less awful in his few relief appearances back in 2012.

Three of these guys were breakout pitchers from just last year: Garrett Richards, Tyson Ross, and Dallas Keuchel.

Which Pitchers Saw The Largest Decrease in K% Upon Converting?

These 10 pitchers did.

Name	Season Y1	Season Y2	Team Y1	Team Y2	K% Y1	K% Y2	K% D	SP / RP Y1	SP / RP Y2
Carlos Villanueva	2010	2011	Brewers	Blue Jays	29.0%	12.9%	-16.1%	RP	SP
Phil Hughes	2009	2010	Yankees	Yankees	33.7%	19.8%	-13.9%	RP	SP
Brandon Morrow	2008	2009	Mariners	Mariners	33.1%	19.4%	-13.7%	RP	SP
Kevin Gausman	2013	2014	Orioles	Orioles	31.9%	18.5%	-13.4%	RP	SP
Ryan Rowland-Smith	2007	2008	Mariners	Mariners	25.0%	12.3%	-12.7%	RP	SP
Wade Davis	2012	2013	Rays	Royals	30.6%	18.4%	-12.2%	RP	SP
Joba Chamberlain	2008	2009	Yankees	Yankees	30.8%	18.7%	-12.1%	RP	SP
Daniel Bard	2011	2012	Red Sox	Red Sox	25.7%	13.8%	-11.9%	RP	SP
Joba Chamberlain	2007	2008	Yankees	Yankees	37.4%	27.0%	-10.4%	RP	SP
Joe Kennedy	2006	2007	Athletics	Athletics	19.6%	9.5%	-10.1%	RP	SP

Raise your hand if you were surprised to see Phil Hughes and Joba Chamberlain on this list (twice!).

...

I see none of you are raising your hands.

Wait. Hang on. Can We Talk About Carlos Villanueva for a Minute?

Sure. I thought you'd never ask.

Name	Season Y1	Season Y2	Team Y1	Team Y2	K% Y1	K% Y2	K% D	SP / RP Y1	SP / RP Y2
Carlos Villanueva	2010	2011	Brewers	Blue Jays	29.0%	12.9%	-16.1%	RP	SP
Carlos Villanueva	2011	2012	Blue Jays	Blue Jays	20.2%	22.9%	2.7%	RP	SP
Carlos Villanueva	2012	2013	Blue Jays	Cubs	24.7%	17.7%	-7.0%	RP	SP

That's weird, right? The thing about Villanueva is that he's much less of a "conversion" type than a swingman. He had the 50 IP as starter and 20 IP as reliever in three consecutive data sets. And yeah, while you see the wave motion -- loses K%, gains K%, loses K% -- you could more likely chalk that up to a fluke run as a starter in 2012. Regardless, he certainly doesn't fit the profile of a Carlos Martinez. Maybe we can find someone who does.

What Kind of Pitcher is Carlos Martinez?

Martinez is fun -- he's a highly-touted prospect who's still young. He was a starter in the minors and converted to relief not because he failed as a starter, but because of a need and roster issues. What other guys fit that mold?

Chris Sale, Chicago White Sox

Ladies and gentlemen ... this is probably your best-case scenario. Sale made the jump from full-time reliever in 2011 (strikeout rate: 27.4%) to full-time starter in 2012 (strikeout rate: 24.9%). That drop isn't too bad from relieving to starting -- but even Sale took a little while to acclimate and improve his strikeout rate after the conversion.

The thing is that Sale was a better reliever than Martinez was ... probably. Carlos's 21% strikeout rate as a reliever isn't something to write home about in this age of the #strikeoutscourge. Thanks to K% park factors provided by John Choiniere, I place Sale's 2011 K%+ -- his K% compared to league average -- at 149. That's 49% better than league average. Martinez in 2014 ... his K%+ was 101. Not exactly in the same ballpark there.

It's hard to imagine that Martinez could be an elite starter, based on his current K%, unless it does rise as Megdal imagined. And since it's lower than Sale's, perhaps that's possible, or even likely. But Sale's dropped after moving out of the 'pen, and Martinez's might as well.

Justin Masterson, Boston Red Sox

Masterson also had a better strikeout rate during his relief days in 2008 -- but the jump isn't quite as stark down to Martinez's league-average K%. The newly-returned Red Sox had a strikeout rate of 21.5% in 2008, which was about 25% better than the league average, and saw that number dip to a 19.0% strikeout rate as a starter in 2009. That was also a swingman season for Masterson -- he started 16 games while relieving in 28. Not the world's smoothest transition, but it led to a productive career as a starter. Curiously, Masterson's percentage point drop from starting to relieving was almost identical to Sale's.

I'm not sure that Masterson is the best comp, due to the fact that he filled both roles in his 2009 "starting" campaign. Can we do better?

Tyson Ross, San Diego Padres

Here we go. Ross, as I mentioned before, saw his K% jump up from what it was as a reliever (17.3%) in 2013 to a strong 24.0% as a full-time starter in 2014. While Ross is older than Martinez, the baseline for his strikeout rate in 2013 -- which was 13% worse than league average -- was a little depressed compared to his minor-league numbers. The same could be said of Martinez in 2014.

Martinez has more of a prospect pedigree than Ross does, but I feel like there's some possibility here for him to follow a Ross-like path. In his limited time pitching as a starter in 2014, his strikeout rate was higher than his time as a reliever. Given the fact that Martinez wasn't a great strikeout pitcher in his relief time ... maybe that means he can find himself in a position closer to the Ross / Richards / Keuchel group, with almost nowhere to go but up?

I'm not comfortable drawing any sort of definitive conclusion from the data I've run across here -- but I was a little surprised to find as many examples of improved K% upon a move out of the bullpen. Sure, in the aggregate pitchers are more likely to lose strikeout rate than to gain it -- and Martinez probably can't afford to strike out fewer guys than he is already. But there's hope here for him to become a solid, middle-of-the-rotation starter with more Ks, just as Howard Megdal said.

... not that the Cardinals need more hope, really.

* * *

Thanks to John Choiniere and Chris Teeter for research assistance. All statistics from FanGraphs.

Bryan Grosnick is the Lead Writer at Beyond the Box Score and a columnist at Baseball Prospectus Boston. You can follow him on Twitter at @bgrosnick.