Aroldis Chapman‘s bid to join the Cincinnati rotation is fascinating for two reasons:
- He’s such a unique player with very few good comparables
- He was extremely valuable as a closer (his 3.3 WAR was the third-best by a RP in the last seven years)
I’ve been highly critical of Chapman over the last few weeks, questioning whether he had a diverse enough repertoire to make it through a Major League lineup two or three times. As a Red Sox fan I had the displeasure of watching Daniel Bard flame out in grand fashion last year as his mix of electric fastball, devastating slider, occasional change-up didn’t translate well to the starting rotation. Of course, Bard walked 15.5% of the batters he faced — twice the league average of 7.4% — so he wasn’t doing himself any favors, but how can news of Chapman’s transition not remind you a bit of Bard?
On the other end of the spectrum is Chris Sale, and anyone who’s read more than two articles on our site has undoubtedly heard me explain why Sale succeeded where Bard did not. Even when he pitched out of the bullpen in 2011, Sale worked batters with four pitches he threw at least 9.0% of the time. His 95.3-mph fastball wasn’t unhittable, but he used it effectively to set up his slider and change-up, both of which were well above league average.
So now whenever it’s announced that a reliever will try to become a starter, PitchFX is the first place I go. And that’s where I went today to try to better answer the question, “Will Aroldis Chapman succeed as a starter?”
Varying your pitches matters
Using the PitchFX data provided by Fangraphs, I looked back at each season from 2008-2012 and grouped starting pitchers into five categories based on how many different pitches they threw at least 10% of the time. Then I found the ERA- (ERA compared to the league average for each season) for each group to see exactly what effect having an expanded repertoire had on a pitcher’s success.
Only starting pitchers with at least 140 innings pitched in a season were included.
The results are rather straightforward. Pitchers that threw five or more different pitches at least 10% of the time, regardless of the quality of those pitches, had an average ERA- of 91.7. In last year’s baseball climate when the league average ERA was 4.01, these pitchers would have averaged a 3.68 ERA. As you begin to take away pitches though, each group’s performance starts to tail off. Pitchers that lacked variety, throwing just two pitches at least 10% of the time, were 2.7% worse than the league average. Last season they would have pitched to a 4.12 ERA.
In an attempt to look at the data differently, I also grouped pitchers into three groups based on how many pitches they threw at least 20% of the time. The idea here was to eliminate pitches that they throw just to say they threw another pitch, instead focusing on pitches the pitcher actually specializes in.
These numbers show even more variation with the pitchers who evenly mix three or more pitches posting an ERA- of 88.9 (3.57 ERA in 2012) with the pitchers who threw only one or two pitches with great frequency performing significantly worse.
Aroldis Chapman’s repertoire
First, it should be noted that the numbers below represent how Chapman pitched as a reliever. It’s entirely possible — actually, it’s extremely likely — that they will vary once he becomes a starter, but it’s not unreasonable to think that his final 2013 pitch selection will resemble what we see here.
According to the definitions used in the previous section, Chapman is just a two-pitch pitcher. I’d expect that when he joins the rotation he’ll add a third pitch, which will likely be his change-up.
I do have my concerns about his change-up, though. Last year it was a below average pitch (-0.57 RAA per 100 thrown), and that’s because he wasn’t able to get good separation in velocity between his change-up and fastball. It doesn’t matter if your fastball averages 98.0 mph (which Chapman’s did) if your change-up averages a robust 93.2 mph (again, which Chapman’s did).
From analysis that I’ve done on the up-and-down play of Francisco Liriano, I’ve come to the conclusion that a pitcher really needs at least a 7.5-mph gap between their fastball and change-up for the off-speed pitch to become effective. At just 4.8 mph, Chapman is well below this mark. He’ll need to make some serious improvements if he hopes for it to become a plus pitch.
And seeing as he hasn’t even attempted a pitch aside from his fastball, slider, or change-up in the last three years, I find it hard to believe that he’s going to immediately mix in a brand new plus, or even league average, pitch with the required regularity.
Update: Thanks to some educated Reds fans in the comments section, I took a deeper look at exactly what Chapman’s “change-ups” really are — mislabeled fastballs from his dead arm period. In fact, according to BrooksBaseball.net, Chapman has only thrown three true change-ups in his major league career (shown below). That means he’s the definition of two-pitch reliever, throwing exactly two pitches. This doesn’t change any of the conclusions reached in this article, but I appreciate the information!
Does Chapman have any comparable players?
Not really, and that’s what makes projecting him hardest of all. But in my research I did find a few players that closely resemble Chapman, so they’re the best case studies we really have on the subject.
These are all the pitchers who in the last five years have had a season where they fit the following criteria:
- They threw their fastball or slider at least 80% of the time
- Their fastball averaged at least 93 mph
- They only threw 2 or 3 pitches greater than 10% of the time
Those three criteria accurately describe Chapman, which makes these five pitchers (plus an extra season from Jackson) the best comparisons I think we can statistically find. Three of them had below average years while three of them had above average years, and those three above average years (Santana in 2008, Johnson in 2009, Jackson in 2009) were very above average.
If Chapman was to finish the year with an ERA- of 80, he’d end up with an actual ERA around 3.20 (depending on the league average). On the other end of the spectrum, if Chapman was to finish the year with an ERA- of 110, he’d end up with an actual ERA around 4.41. Obviously that’s quite the gap, but what we can really take away from this is that it is possible to be successful with the type of repertoire we expect from Chapman.
Will Chapman succeed?
To a degree, yes, I think he will. I’ve already gone on record projecting Chapman for a 3.69 ERA over 167 innings this year. Since we’ve been using ERA- all day, that 3.69 mark would equate to an ERA- of around 92.0, pretty much in line with the average ERA- of the five comparable pitchers we looked at and noticeably better than the groups of pitchers Chapman would likely fall into (just 2 or 3 pitches 10% or 20% of the time).
I’m giving him the benefit of the doubt because of the talent that he has; his arm is so much livelier than just about any other starting pitcher’s in recent memory. I’ve said several times that Chapman doesn’t have a real comparison because of the velocity he can attack batters with, but after this season that won’t be an excuse any more. Chapman’s breaking the mold, and from now on he’ll be the example we use when trying to determine if high-velocity relievers can hack it in the rotation.