by Ryan Potter
The Reds intentions concerning their flame-throwing Cuban lefty, Aroldis Chapman, have never been much of a secret. Entering Spring Training in 2012, Chapman was on pace to become Cincinnati’s number one starter before an injury to incumbent closer Ryan Madson forced the Reds to keep Chapman in the bullpen. The 25-year old southpaw acquitted himself well in the closer role, posting a 1.51 ERA and striking out a video-game like 15.32 batters per 9 innings on his way to saving 38 games. During the offseason, the Reds signed former Dodgers and Royals closer Jonathon Broxton to a 3-year, $21 million deal, giving them the flexibility to make Chapman a starter. I can’t blame the Reds for going all in this season, they have a very talented team capable of winning the NL Central, but I don’t think taking Chapman out of bullpen and putting him in the starting rotation is a good idea.
I went back and examined the all the available data on starts Chapman made in the past, both in the minor leagues and in Cuba. What I found leads me to believe if he is made a starter, Chapman will be pretty close to a major league average starting pitcher. From 2005 to 2009, Chapman started 63 games for Hoguin in the Cuban Series Nacional, a league that is roughly equivalent to High-A in the American minor league system. In his final season, 2009, Chapman’s ERA was 4.03, fairly pedestrian considering the low level of competition. Opponents posted a .252 batting average against him in 20 starts. Yes, he was only 21 that season, but he was already throwing nearly 100MPH fastball and high 80s, low 90s sliders. He should have been mowing down the competition at that level.
Chapman’s final season in Series Nacional, was also the season in which he pitched the most innings of his career, 118.1. Since moving stateside, he hasn’t throw any more than the 95.1 innings he put up during his first season in the Reds organization. Moving him to a starting role will dramatically increase his workload; most Major League pitchers throw 200+ innings per season. A sudden increase in the workload for a pitcher like Chapman will create an increased risk for a major arm injury, which is already a concern given his mechanics.
To avoid putting stress on the arm, most teams prefer to gradually increase a young pitcher’s workload. The Reds have yet to make it clear if Chapman will have an innings limit this season, but it would be a good way to protect their investment. If Cincinnati decides to put a limit on Chapman’s innings in 2013, it could really hurt them during the stretch run, as it did the 2012 Washington Nationals when their ace starting pitcher, Stephen Strasburg, reached his agreed upon innings limit in September.
It shouldn’t surprise you to find out that Aroldis Chapman’s mechanics differ from those of other pitchers. But the reason he is capable of throwing much harder than the average pitcher is the same reason he’s more susceptible to a major shoulder injury. To throw a baseball 100MPH, you have to generate a tremendous amount of torque. In addition to having excellent lower body strength, Chapman is able to rotate his humerus (the bone that connects the shoulder to the elbow) much more quickly than other pitchers, creating the added torque that allows him to throw harder than anyone else in the game. This is both good and bad news for Reds fans. The speed at which he makes this rotation makes it less likely he’ll sustain a major elbow injury, but it also increases the likelihood of him seriously injuring his shoulder. At the very least, repeating his pitching motion frequently will cause the ligaments in that connect the shoulder to the humerus to lengthen which means he’ll slowly lose velocity on his pitches. This regression is somewhat natural in pitchers, and it’s the reason why you’ll never see a pitcher at the end of his career throwing just as hard as he did the day he came up to the big leagues.
There is a decent amount of evidence that this may already be happening to Chapman. I went through all of his Fangraphs’ PitchFX game logs (which record the velocity and type of every pitch thrown) to find out for sure and pulled out a couple samples to illustrate my point. Compare the average speeds of Chapman’s fastball in each of his appearance from June 10th-June 19th 2012 (prior to which he had thrown 30 innings) to the same data from August 31st 2010-September 10th, his first 5 appearances for the Reds (prior to which he had thrown 94.2 innings at AAA-Louisville)
6/10/12: 99.3 MPH 8/31/10: 100.7 MPH
6/13/12: 97.4 MPH 9/01/10: 102.1 MPH
6/15/12: 96.2 MPH 9/04/10: 100.4 MPH
6/16/12: 96.5 MPH 9/06/10: 101.0 MPH
6/19/12: 95.8 MPH 9/10/10: 102.2 MPH
That’s a pretty significant difference in average fastball velocity in just two seasons. It’s even more significant when you consider Chapman was a reliever during this period and his arm was not being taxed like it will be as a starter.
This brings me to my biggest problem with Aroldis Chapman becoming a starter. As a reliever, a pitcher doesn’t need much of a pitch selection. Relievers can get by using two pitches and only two pitches. Aroldis Chapman is one of these pitchers. His fastball and slider are both what scouts had in mind as the best possible offerings when they began applying the standard 20 to 80 scouting scale to pitching (you’ll occasionally hear it as 2 to 8, those people are wrong. They are not your friends. I’m a purist and also prefer to have the ability to grade by 5s if I need to).
Chapman mixes his fastball and slider to create a devastating effect on the batter. In my opinion, this is not only because they have a great deal more velocity than the offerings of most pitchers, but also because his slider often comes on the same plane (See! Your high school geometry was right, after all. You will need to use geometry in adult life, mostly to read shoddily written articles about lanky Cuban lefties) as his fastball, giving the batter with little indication of which pitch is approaching the plate until the slider bites sharply to the right, often leaving the batter with nothing but a case of the “jelly legs” to slow for his efforts.
The problem is that no starter will get by using just two pitches. Yes, Chapman does possess a change-up, but it’s purely a “show me” offering that he rarely deploys. Here’s the breakdown of the PitchFX data regarding the types of pitches thrown by Chapman during the 2012 season, per Fangraphs:
Pitch Type Percentage Thrown
The figure on changeups seems a bit mysterious to me. It seems to me that’s possible a indeterminable amount of pitches tallied by PitchFX as changeups were actually no more than fastballs in disguise during one of Chapman’s “dead-arm” periods of the season, in which his average velocity dipped considerably due to his inability to adjust to the workload expected of him. I can’t prove that, and I’d speculate that it would be impossible to prove anyway. Considering that his average changeup came in at 93.2, which is only 4.8 MPH less than his average fastball velocity of 98.0 MPH, if he does have a changeup he might as well leave it in the bag.
I feel it necessary to mention that I’ve heard talk of Chapman also possessing a cut-fastball. I never saw it when I watched him live, and the scouting reports I’ve read on him were conflicting on the subject. The scouts that did claim he threw a cutter didn’t rate it very well. I’m going to go ahead and assume it’s a poorly thrown slider that doesn’t bend like it’s supposed to, because the Reds unfortunately did not feel that 4FanSports was reputable enough to grant me an interview with Chapman.
When I began researching this piece, I assumed that pitchers with more pitches in their repertoires fared better as starters. I was dreading a long night of going through several years of data (to create a sample size large enough to be meaningful) to find out if my assumption was true. Luckily, Bryan Curley over at Baseballprof.com already compiled this research and condensed it into a handy graph. Bryan looked at Fangraphs’ PitchFX data from the 2008 season through the 2012 and grouped pitchers by the number different pitches in their arsenals. For the purpose of this exercise, if a pitcher threw a type of pitch at a rate of 10% or more, he was considered to have the pitch in his arsenal. He then compared that data to the ERA- of the group of pitchers. Note: ERA- is the standard ERA (Earned runs multiplied by 9, divided by the number of innings throw) when compared to the league average.
What we can see here is that pitchers with 5 or more offerings perform at a level better than league average, but as you take away a pitch, the groups begin to regress toward the league average or end up below it.
Given all the information provided above, is it really worth it to the Reds to give up an All-Star level reliever to gain a major league average starter? I suppose it really depends on how the Reds organization values relief pitching compared to starting pitching. But to me, it isn’t worth the risk.
Bonus: At some point later today, I’ll post an additional section of this article on regression and fly-ball rate. It couldn’t be included in this post because the University of Kentucky selfishly demanded a paper on the American educational system from me.