Maybe the Last Key for Carlos Carrasco

By this point you might feel like you know enough about the Carlos Carrasco story. Carrasco is coming off what looks like a breakthrough season. It was also his age-27 season, and previously, he was mostly regarded as a bust. He first showed up in the Baseball America top-100 before 2007, when the No. 1 prospect on the list was Daisuke Matsuzaka. Carrasco was one of the headliners of the Phillies’ trade for Cliff Lee. That was before Lee got traded to the Mariners for what’s turned out to be busts. And that was before Lee got traded to the Rangers for whats’ turned out to be busts. Carrasco had a delayed emergence, is the point. He’s why it’s hard to ever give up entirely on a former top prospect.

Yet it’s worth remembering that 2014 wasn’t a total victory for Carrasco from start to finish. It was only down the stretch that he seemed to put all his pieces together in the right places, and before his final stint in the rotation, he looked like just a pretty good reliever. There was something that clicked upon Carrasco’s final return to starting, and it seems to me it bodes well for his future in the role.

Carrasco started four games. Early on, Carrasco started four games. He allowed 18 runs, throwing just 61% strikes. His performance wasn’t good enough, obviously, so the Indians bumped him to relief. Unsurprisingly, Carrasco found more velocity in what we might think of as the sprinter’s role, and all his stuff shot up a couple ticks. He struck out almost a quarter of everybody. He walked fewer than two per nine innings. Carrasco adjusted well to the bullpen — he managed an FIP- of 79. That would’ve been enough; the Indians would’ve been content for Carrasco to be a reliable option for the seventh or eighth innings.

But on August 10, Carrasco started again. He’d start ten times between then and the end of the season. And everything got better. Relative to his time in relief, everything got better, rather unusually. More whiffs. More strikes. And even a little more velocity, with all of it combining for a 47 FIP-. No starter in baseball had a better second-half FIP. No starter in baseball had a better second-half ERA. For good measure, no starter in baseball had a better second-half xFIP. Over that stretch, Carrasco was every bit the equal of Clayton Kershaw. I didn’t think I’d be able to type that sentence, either, but I just double-checked. Either Carrasco was really that good, or I decided to trip balls on a Wednesday afternoon.

So: what? What happened? My suspicion is it has something to do with this:

Francona said Carrasco has raised arm slot to add deception, adding that team would love to see the pitcher “take off” this season.

— Jordan Bastian (@MLBastian) January 22, 2014

That was from January. Carrasco’s throwing motion would go on to be something of a minor story — there was disagreement in the early part of the season, and Carrasco seemed to prefer a different movement with his glove arm. Now, keep that somewhere in mind as you look at the following image, from Brooks Baseball, showing Carrasco’s game-by-game vertical release points:

You see points bouncing around. Then you see consistency. Toward the end, right when Carrasco was performing like a crazy person, his release points remained stable. He was releasing a little lower than he had been. All of his pitches were in line — there was minimal variation, and even though it took some time, what this suggests is that Carrasco and Mickey Callaway were able to get on the same page. By the end, there was no more tweaking and no more fluctuation.

Clearly, based on the consistency, Carrasco was comfortable with the delivery. Clearly, based on the velocity, it was working for him. I don’t think it was all about release points; this is probably the result of some other adjustment, that allowed Carrasco to arrive in a more consistent position. But it sure looks like they got everything about Carrasco all perfectly streamlined, so both his command and velocity played up despite a return to longer outings.

As is pretty much always the case, Carrasco’s improvement seems to stem from a stronger fastball. That is, stronger in terms of Carrasco’s command of it. From Baseball Savant, heat maps:

Maybe the .gif isn’t very helpful. Could be that things seem subtle to you. But, Carrasco unquestionably tightened up. Through August 5, Carrasco threw 61% of his fastballs at least 2.5 feet off the ground, representing roughly the vertical middle of the strike zone. Once he returned to the rotation, he threw just 47% of his fastballs at least 2.5 feet off the ground. Carrasco didn’t fly open nearly so much, and by keeping his fastball closer to the zone, he allowed his other pitches to improve.

The fastball remained the primary pitch — it’s just about always the primary pitch — but Carrasco started to throw more changeups. He also started to throw more sliders. With the fastball working lower than before, it was more difficult to separate from the change. And the slider became a wipeout weapon, especially against righties, as Carrasco would sometimes even bring it up into the zone for a different look. As he was better able to spot his heater on either side of the plate, it got all the more difficult for hitters to lay off secondary stuff breaking in another direction. There were no tells in the arm slot, and there were fewer tells in the pitch locations.

As Carrasco streamlined his mechanics, it all fell into place, better than anyone would’ve imagined. When you have a better fastball, you have a better everything; just ask Corey Kluber. Some quick and dirty splits:

Through August 5

• RHB: 37 strikeouts, 7 walks
• LHB: 25 strikeouts, 11 walks

From August 10

• RHB: 38 strikeouts, 2 walks
• LHB: 40 strikeouts, 8 walks

Better pitches in the zone improved the pitches out of the zone. In the second half, out of 134 starters, Carrasco ranked No. 2 in O-Swing%. He was out of the lead by three-tenths of one point. And he also ranked No. 2 in O-Contact%. The only guy better: Kluber himself. And still Carrasco was also in the top 20 in zone rate, so he wasn’t surviving by getting guys to chase. Because he got so good within the zone, batters became increasingly helpless when he did something else.

The problem with mechanical fixes is they can be kind of like momentum: they’re only as good as the next day’s game. Ubaldo Jimenez looked like he was fixed, until he wasn’t fixed anymore. On the other hand, that’s one example, and unlike Carlos Carrasco, Jimenez no longer has Mickey Callaway as a coach. Probably, Carrasco isn’t going to be as good as Clayton Kershaw and Felix Hernandez going forward. Probably, Carrasco won’t be the Indians’ No. 1b. But then, it’s almost impossible to fake what he pulled off for a third of a season. It’s almost impossible to be that good without being at least kind of close to that good.