The Terrifying Comp for Lance McCullers’ Best Weapon

Lance McCullers is starting for the Astros on Thursday, and while he’s going to have his innings closely monitored down the stretch, potentially reducing his availability, there’s no doubt he’s played a big role in getting the Astros to where they are at present. McCullers has followed an up-and-down 2014 with an incredibly successful 2015, and in just 11 big-league starts, he’s been worth about 2 WAR no matter which formula you prefer. The Astros are said to be in the market for a starting pitcher, but if it weren’t for McCullers’ presence, the situation would be a lot more desperate.

McCullers throws three pitches, though the changeup is simply coming along. He has a good fastball that sits in the mid-90s, but the best pitch here is the curve, thrown a third of the time and responsible for the majority of McCullers’ strikeouts. It’s never been a secret that McCullers throws a good breaking ball, but in talking with David Laurila, Brent Strom tossed out a heavy comp:

“I’d say McCullers’ breaking ball is Kimbrel-like at times,” said Strom. “That’s as good as you can get. I haven’t seen everybody’s curveball, but I would say the young kid McCullers has a curveball as good as anybody in this game.”

That’s a direct comparison between Lance McCullers’ breaking ball and Craig Kimbrel’s breaking ball. That’s coming from a big-league pitching coach, so it carries some weight. But why not put numbers to this, to try to find the best comp, really?

It’s been a few months since I played around with pitch comps. It was a useful tool to pass the time during the offseason, but this is a good opportunity to bring it back. The pitch comps are made possible by the Baseball Prospectus PITCHf/x leaderboard. As a reminder, what’s considered are just pitch velocity, horizontal movement, and vertical movement. I find the standard deviation for each category, and using that, I calculate the z-score difference between a given pitch and the target pitch. So, for each pitch, I get three z-scores, and then I just add them up for a comp rating. For purposes of this analysis, I compared all the current curveballs to Lance McCullers’ curveball. I sorted the comp ratings in ascending order.

Strom, off the top of his head, compared McCullers’ curve to Kimbrel’s. By comp rating, the runner up here is David Robertson. Usually I keep starters and relievers separate, but I decided to combine them today. So Robertson’s curveball is the second-best current comparison. And, the pitch at No. 1? The curveball most similar to Lance McCullers’ curveball, by this method? It’s…Craig Kimbrel’s. Strom and the numbers agree. The two pitches are essentially brothers.

By horizontal movement, they’re more or less identical. By vertical movement, they’re more or less identical. By velocity, Kimbrel has the edge over McCullers, by a couple miles per hour. Still, it’s the closest comp, and this is where I note that, while Kimbrel’s curve is harder than McCullers’ is, McCullers’ curve is the hardest in baseball among starting pitchers. The starter with the curve that compares the best is Jimmy Nelson, but his curve is slower by two ticks. The only starter with a curve within a tick of McCullers’ is Matt Harvey.

People wonder if and when McCullers will develop that reliable change, but as long as his breaking ball looks like Craig Kimbrel’s, he should be able to survive. Kimbrel’s is one of the most dominant pitches in the game, and McCullers works several innings at a time. To this point, the curve has been wildly successful against major-league opponents. Perhaps in time it’ll be over-exposed, but perhaps it’s just an excellent weapon.

Righties have seen the curve almost 40% of the time. Lefties have seen it almost 30% of the time. Lefties have actually had more trouble hitting it, but that doesn’t mean it’s been easy for righties, and it’s both a swing-and-miss pitch and a groundball pitch. In that sense, it could be compared to A.J. Burnett‘s curveball. And Felix Hernandez‘s curveball. And Cole Hamels‘, and Madison Bumgarner’s. There have been 91 at-bats ending on a curve. Half of those have been strikeouts. Against the pitch, opponents have slugged .121.

From Baseball Savant, this is where the curveballs have gone:

mccullers-curve

This, also, is where the whiffs have been:

mccullers-curve-whiffs

Those locations aren’t surprising. Let’s check out the pitch in action. For a called strike:

For a grounder:

For a whiff:

I mentioned that McCullers has made just 11 big-league starts. Keep that in mind as you consider the following. We can refer to the pitch-value leaderboard, available on this site. As always, I should remind you that it’s more difficult than this to evaluate a pitch in isolation. Pitches really can’t be evaluated in isolation. Pitching is more complicated than that. But what we find is that good pitches — pitches we think we know to be good — tend to get good ratings, and worse pitches get worse ratings. So there is real signal here. Sort the leaderboard by curveball value. The top:

  1. Lance McCullers, +12.5 runs
  2. Felix Hernandez, +10.2
  3. Corey Kluber, +10.1
  4. Clayton Kershaw, +9.7
  5. Yusmeiro Petit, +6.8

By this one measure, McCullers has thrown the most valuable curve in baseball. He hasn’t been in the majors all season long. His curveball is also the fastest among starters, so this isn’t a total shock, and it turns out it really does compare well to Craig Kimbrel’s breaking ball. Everything is in place. It seems established by now that this curveball is terrific. Its top comp is terrific, and terrifying.

Which gives McCullers a strong foundation. He’d like a better changeup, but right now he’s doing okay. If the Astros get their wish, McCullers will remain sufficiently effective to stay a starting pitcher, long-term. And if he were to ultimately end up in the bullpen? Craig Kimbrel wound up in the bullpen. Never started a single game.





Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.

14 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
youngmichael
8 years ago

Holy F that was an awesome article.