Joba Chamberlain is at a crossroads between two different career paths. On one side, there’s the path the Yankees have publicly set for him–become a reliever to help the team reach the playoffs this year, and then return to pitching every fifth day. On the other side, there’s the plan Chamberlain is creating for himself, by allowing one hit and striking out eight hitters in his first five innings. This path would involve preparing Joba to become their future closer, sort of like the Tigers with Joel Zumaya.

Personally, I’d like the Yankees to take their cue from Earl Weaver and re-convert Chamberlain to the starting rotation. While Nate Silver did some of the math on the differences between starting and relieving from a value perspective earlier this year (during Jon Papelbon’s spring starting trial), I think Joba needs to be given the opportunity to become an ace instead of a fireman. He doesn’t have a problem with lack of diversity in his pitching arsenal; Chamberlain throws four average-or-better pitches for strikes. He also has a frame that should be able to sustain 200 innings of pitching per year. But too often, teams are willing to typecast players once they find any success in the Major Leagues.

In 1999, the Blue Jays called up former first-rounder Billy Koch to fill a hole in their bullpen, and while Koch had started in all 35 minor league games in which he had previously appeared, his future in relief was assured after he allowed just two baserunners in his first three appearances, spawning 5 2/3 innings. Billy Wagner had 70 minor league starts under his belt when called up by the 1996 Houston Astros. Wagner struggled in his fourth inning, of work so Houston began keeping his outings short, and when he had success doing that, his past success as a starter were forgotten. Now I can’t definitively say that Koch or Wagner or the dozens others who were forced to turn their backs on starting would have made better starters than relievers, but they weren’t even given the chance.

I do believe the Yankees can follow through with their plan of moving Joba back to the rotation and still look internally for a future closer candidate. They even already have one at hand–Edwar Ramirez. It should come as no surprise, however, that Joe Torre didn’t give Ramirez the time of day in his brief big league trial, but is nevertheless willing to hand over the set-up spot to Joba so quickly. It’s a Prospect Evaluation 101 old school axiom: major league relievers were once minor league starters, but dominant minor league relievers struggle to make the final step.

Between 2000 and 2006, I found twenty-seven different major league closers who posted WXRL figures of five or better. I added Joel Zumaya into the study because my thought process started with Chamberlain, and in career path B, Zumaya is Joba before Joba was. Of this group, only six players were relievers throughout the minor leagues, representing just 21.4 percent of the study. The other 78.6 percent began as starters–and whether in Double-A (Francisco Rodriguez) or 13 years into their major league career (John Smoltz)–and were moved to the bullpen.

The Yankees will be the sole deciders between these two career paths for Joba, but thinking about a potential career as a reliever for him, I began to be curious if Chamberlain fit into the profile of the usual minor league ace turned major league closer. I also wonder if a minor league reliever like Edwar Ramirez (or Kevin Whelan or David Robertson) fits as the rare pitcher who continues his trade between the minors and majors.

Lifetime Relievers

As mentioned, six players in the study were relievers from A-ball to the major
leagues. Chad Cordero spent the shortest time in the minors, pitching just 26 1/3 innings in High-A before ascending to the Major Leagues. Four more were products of the draft: Danny Graves (fourth round, 1994), Troy Percival (sixth, 1990), Trevor Hoffman (11th round, 1989), and B.J. Ryan (17th round, 1998). However, Hoffman and Percival represent special cases–both started off as position players, Percival at catcher, and Hoffman at shortstop. All five of those players came from solid college programs: Cal State Fullerton, Miami, UC Riverside, Arizona and Louisiana-Lafayette, respectively.

Cordero was the only member of the group drafted in the first round, but I think
that represents a bit of a generational gap between him and the others, more than a comment on his relative talent. College closers have been trending up on draft boards over the last decade, and while we’re a few years away from studying if they represent first-round value, it certainly appears the most likely minor league relievers to turn into major league closers were once college relievers. That’s a mouthful, but you can bet Cordero can thank Troy Percival and friends for his first round-bonus check.

The other player who never really made an attempt at starting was Armando Benitez, who did start three games at age 18 in the Gulf Coast League, but would relieve in his next 155 contests. Benitez was a dynamic talent, struggling with his control as a teenager in the Appy League before putting it all together in full-season baseball. Benitez would pitch in 160 1/3 innings over all four levels before reaching the Orioles‘ roster, in which time he struck out 255 batters and allowed just 89 hits and 38 earned runs.

The group’s minor league numbers, by level, as a whole:

Level      ERA       H/9       K/9    BB/9    HR/9
A-        1.68       5.3      13.8     3.3     0.2
A+        1.81       6.4      11.5     3.6     0.1
AA        2.37       6.1      11.0     4.0     0.4
AAA       3.80       8.1       8.9     3.9     0.7

(Note: The Double-A numbers do not contain a brief 1992 six-start experiment by
Trevor Hoffman, who would relieve in 37 games in Triple-A that season.)

Better minor league hitters began to take more pitches and hit more home runs at
higher levels, resulting in the gradual increase in ERA. Overall, this was a
group that put up some ghastly minor league numbers, and I’ll let you only imagine their group K/9 if Danny Graves were omitted from the study.

Getting back to the Yankees, Edwar Ramirez isn’t the perfect fit, because he’s older–Ramirez’ minor league dominance came after struggles at the ages the players above were successful at, so it’s not a perfect fit. Ramirez is actually more like Jeff Zimmerman, a journeyman signed by Texas after dominating the Northern League. Zimmerman then put up ridiculous minor league numbers, allowing just 11 earned runs in 81 innings before becoming the Rangers‘ closer. So, instead of Ramirez, who is there? I think the answer is David Robertson; the guy hasn’t allowed a home run yet in the minors, and A-ball hitters haven’t touched him. If you give Mariano Rivera another couple of years before his inevitable plummet into normality, Robertson could be there to catch his fall.

Minor League Starters

Most often, closers come after years of starting in the minor leagues. This study has 22 closers that began as minor leaguers, which I divided into three groups: those moved to relief shortly after reaching the big leagues, those moved to relief in the minor leagues, and those moved to relief after starting in the major leagues didn’t work out.

The latter group consists of six names: John Smoltz, Joe Nathan, Ryan Dempster,
Jason Isringhausen, LaTroy Hawkins, and Eric Gagne. Smoltz and Dempster were, to
varying degrees, accomplished starting pitchers who moved to relief as a catalyst for extending their careers. Jason Isringhausen was still a starter when he missed all of 1998 with injury, but he hadn’t been great before, so Oakland faced no gripes when they moved him to reliever full-time in 1999. The other three–Nathan, Hawkins, Gagne–were just horrible as starters. In a combined 175 big league starts, the trio had a 5.45 ERA and allowed 152 home runs.

The group, overall, has diverse biographical baseball beginnings. Smoltz was a
late-round local high school pick by the Tigers, and Hawkins and Dempster also came straight from high school. Joe Nathan was a sixth-round pick from Stony Brook University, Isringhausen a 44th round draft-and-follow, and Gagne an undrafted signee from Seminole State. Despite little in common in that department, there were more similarities in their minor league numbers. A look at how the six did as minor leaguers:

Level      ERA     H/9    K/9   BB/9   HR/9
A-        2.82     6.9    9.5    3.3    0.6
A+        3.61     8.2    7.3    2.8    0.8
AA        3.66     7.4    7.8    3.8    0.9
AAA       3.31     8.8    6.9    2.7    0.7

(Note: A repeat trip by LaTroy Hawkins to Triple-A after he had already lost his
rookie eligibility was not used. Hawkins had already begun his descent as a starter.)

These are some very interesting results, with particularly noteworthy numbers in the strikeout column. This group was striking out players at a much lower rate than the group that relieved throughout the minors.

Getting to the players that have similar career paths to how Joba Chamberlain moved up, we arrive at a list of 16 players who began as minor league starters but didn’t get the opportunity to start many major league games. Five of these players were moved to the pen prior to even reaching the major leagues, and were given more time to do so than Chamberlain, who was just given a quick Triple-A relief tune-up. Francisco Rodriguez had battled injury problems and ineffectiveness, but dominated in 2002 when moved to the bullpen at Double-A, relieving 50 games before his memorable playoff dominance with the Angels. J.J. Putz wasn’t moved until his second trial in Triple-A, though Putz hadn’t been particularly bad in his first run through the league. The other three players–Brad Lidge, Todd Jones, Brian Fuentes–were moved to the bullpen when
reaching Triple-A. Overall, this group is also heavy on college pitchers, with all four American-born players drafted from colleges. Lidge and Jones were first-round picks (from Notre Dame and Jacksonville State, respectively), Putz went to Michigan, and Fuentes was a steal from Merced College.

The other 11 players all started in the minor leagues and moved to major league
bullpens upon or shortly after their being called up to the majors. Here’s a list of the players, and where they hail from:

Keith Foulke: Lewis-Clark State (9th round)
Derek Lowe: Michigan HS (8th round)
Billy Koch: Clemson (1st round)
Billy Wagner: Ferrum College (1st round)
Mariano Rivera: Panama
Jon Papelbon: Mississippi State (4th round)
Antonio Alfonseca: Dominican Republic
Bob Wickman: Wisconsin-Whitewater (2nd round)
Jose Mesa: Dominican Republic
Joel Zumaya: California HS (11th round)
Tom Gordon: Florida HS (6th round)

Perhaps surprisingly, just three players were taken in the first three rounds in the group, with Wickman and Wagner being drafted high out of very small schools. Here’s a look at the numbers of these players in the minor leagues, as well as the previous group until they began relieving:

Level     ERA      H/9     K/9   BB/9    HR/9
A-        2.92     6.9    10.0    3.7     0.3
A+        3.65     7.7     8.2    3.8     0.5
AA        3.55     8.0     7.7    3.5     0.7
AAA       3.67     8.3     7.7    3.3     0.7

Important point: Joba is a much better pitcher than these averages. Looking through the numbers, the best comp for Chamberlain is Billy Koch. Like Koch, Joba was taken in the first round from a big program, and progressed quickly in the minor leagues, though Joba was a bit more dominant. He’s also a bit like Brad Lidge, but the latter is a reliever mostly because he was injured too often to stick as a starter. My guess is that this isn’t good enough for Yankees fans, and that we all might agree that the best thing for everyone is that Joba winds up as a starter.