Carlos Carrasco’s Change Doesn’t Really Have a Comp by Jeff Sullivan February 20, 2015 I’m back again to close out your week by talking about pitch comps. I’ve talked about pitch comps a lot lately, looking at Henderson Alvarez, and Mariano Rivera, and Cole Hamels, and so on. Pitch comps love Marcus Stroman. Not coincidentally, I also love Marcus Stroman, but this is going to be about a different guy — this is focusing on Carlos Carrasco. And while I’ve written about Carrasco already in the recent past, I want to add something to that. Carlos Carrasco throws a changeup. No one else really throws Carlos Carrasco’s changeup. To quickly review the methodology, I use the Baseball Prospectus PITCHf/x leaderboards and look at velocity, horizontal movement, and vertical movement. For this, I looked at right-handed starting pitchers who, last year, threw at least 50 changeups, according to the page. For each of the three categories, I calculated the z-score difference between a given number and Carrasco’s changeup’s number. Then I simply added up the three absolute values, yielding a comp rating. That’s all the boring stuff. Below is the more interesting stuff. The top comps, that I found: Pitcher Pitch Velocity Horizontal Vertical Comp Rating Carlos Carrasco CH 89.5 5.3 -0.3 – Felix Hernandez CH 89.9 6.3 1.4 1.5 Henderson Alvarez CH 89.9 6.6 2.2 2.0 A.J. Burnett CH 87.2 7.2 1.4 2.8 Zack Greinke CH 87.7 7.0 2.1 2.8 Yu Darvish CH 87.6 5.5 4.0 2.9 It’s never a bad thing for a pitcher’s closest comp to be Felix Hernandez. And Felix’s changeup is arguably the greatest changeup in the game, with Henderson Alvarez’s similar changeup being too inconsistent. Driving this, in part, is that Carrasco throws his changeup really hard. Those guys after Alvarez — those don’t make for great comps. The way I think of it, a comp rating of 2 or greater isn’t much of a comp at all. Yet there’s something you might point out about this. A changeup isn’t necessarily a weapon on its own. A changeup really needs to be considered relative to a pitcher’s fastball, because that’s the whole illusion. Carrasco’s change might be closest to Felix’s based on the three traits above, but what does that really mean if we don’t have fastball information? So now here’s more stuff, including just Carrasco, Felix, Alvarez, and the league average. First, a table of changeup traits relative to four-seam fastball traits: Pitcher Velocity Horizontal Vertical Carlos Carrasco -6.7 -1.0 8.0 Felix Hernandez -3.7 3.8 7.2 Henderson Alvarez -4.7 2.6 6.4 League Average -8.0 2.7 4.2 This might be a little difficult to understand at first. I’ll walk you through it. The average changeup is eight miles per hour slower than the average four-seamer, within the pool. Carrasco has a separation of almost seven ticks. Felix is under four; Alvarez is under five. Meanwhile, the average changeup has almost three inches more run than the average four-seamer, and about four inches more sink. Carrasco’s changeup, interestingly, has less run, and a lot more sink. Almost double the average sink, relatively speaking. You see the comps break down a little, here; Carrasco’s changeup is different, relative to his four-seamer. Now, the same stuff, relative to two-seam fastball traits: Pitcher Velocity Horizontal Vertical Carlos Carrasco -6.0 -3.5 5.1 Felix Hernandez -3.1 -1.3 4.7 Henderson Alvarez -4.4 -1.1 3.8 League Average -7.3 -0.6 1.7 Carrasco’s changeup has a lot less run than his sinker. And he gets plenty of relative drop — even more than Felix does, by a small margin. I should note that Carrasco threw nine times as many four-seamers as two-seamers, or thereabouts. So the most important separation is the separation between Carrasco’s changeup and his four-seam fastball. This is where we get to go back to comp ratings again. Only this time, instead of comparing velocity, horizontal movement, and vertical movement, we’ll compare separations in those three categories between changeups and four-seamers and changeups and two-seamers. I fully understand this might be hard to follow, but I can explain more after the numbers. Here are changeups, relative to four-seamers: Pitcher Velocity Gap Horizontal Gap Vertical Gap Comp Rating Carlos Carrasco -6.7 -1.0 -8.0 – Juan Nicasio -6.7 -0.7 -4.8 1.9 Scott Carroll -7.7 -1.2 -5.2 2.2 Jake Peavy -7.5 1.3 -6.8 2.3 Bronson Arroyo -7.1 1.8 -6.8 2.4 Matt Cain -6.3 2.1 -6.8 2.5 The only somewhat decent comp would be Juan Nicasio, whose changeup has an identical velocity separation, and a similar horizontal separation. But the vertical separation isn’t the same. Carrasco gets a lot more drop. No one here is all that close. Here are changeups, relative to two-seamers, and people are even less close: Pitcher Velocity Gap Horizontal Gap Vertical Gap Comp Rating Carlos Carrasco -6.0 -3.5 -5.1 – Matt Cain -5.6 -1.8 -3.5 2.2 Trevor Cahill -8.5 -3.1 -3.4 2.5 Jhoulys Chacin -6.3 -3.2 -0.6 2.8 Johnny Cueto -8.8 -1.8 -4.6 2.9 Carlos Villanueva -6.6 -3.6 -0.4 3.0 The point of all this: no one really throws a Carlos Carrasco changeup. Not in terms of its raw traits, and not in terms of how it compares to Carrasco’s fastballs. Though it’s all shades of gray, you wouldn’t be wrong to think of Carrasco’s changeup as being unique. Or exceptional, if “unique” makes you uncomfortable. And Carrasco’s changeup is good. That’s the other thing. He doesn’t just generate interesting overall averages: he’s figured the pitch out. He figured a lot of things out in 2014, and changeup command was one of them. From Baseball Savant, here’s where Carrasco’s changeups used to go: And here’s where they just went: This is what the changeup looks like, basically: Last year, Stephen Strasburg led righty starters with a 53.6% contact rate on his changeup. Carlos Carrasco finished at 54.0%, which is effectively the same percent, and Carrasco wasn’t pitching in the National League. Even more, in the second half, hitters whiffed at the changeup more often than they made contact, showing that Carrasco only got stronger. A consequence of this? The changeup, as you know, is mostly a weapon against opposite-handed hitters. Before 2014, against lefties, Carrasco had 67 strikeouts and 57 walks. In 2014, against lefties, he had 65 strikeouts and 19 walks. He turned a K-BB% of 2% into a K-BB% of 18%, with a corresponding massive drop in wOBA. Add in better pitching against righties as well and you get Carrasco looking like one of the best starters in the league. That’s what the numbers show, and Carrasco has the stuff to back it up. I don’t need to write more about how Carrasco was excellent. Already done that. This is about the changeup, specifically. Carlos Carrasco throws one. Nobody else does — not quite like him.