Rich Hill Got Good Again

Over the winter, the Dodgers re-signed Rich Hill. They did so because they knew that, when he was able to pitch, Hill looked like one of the best starting pitchers in the game. And then 2017 went and got underway. Through the earlier part of 2017, Hill looked like one of the more frustrating starting pitchers in the game. The stuff, for the most part, remained there, but Hill didn’t have his same control, and his remarkable curveball was no longer working. I wrote about Hill in the middle of June, at which point his curveball had the lowest run value among all curveballs, out of everyone. It made me wonder where, exactly, Rich Hill was. Could someone who reappeared so suddenly disappear with similar speed?

Since I wrote about Hill and his struggles, he’s gone back to looking like one of the best starting pitchers in the game. In three starts, he’s allowed four runs over 19 innings, with six walks and 26 strikeouts. The easy explanation is that Hill has simply regressed to the mean. That is, his newer mean, the one he began establishing a couple years ago. That explanation would ignore the changes that Hill has folded in. Regression to the mean isn’t an automatic process. Rich Hill has simultaneously simplified and grown more complex.

It seemed to me that Hill’s biggest issue was mechanical. That’s not actually tremendous insight — most struggles tend to be mechanical. Some of those struggles are fixed, and some of them aren’t. Hill developed a location problem, and then, between starts in June, it was as if the pieces fell into place. The Dodgers provided Hill with some information showing that he had been pitching better out of the stretch. Hill, in turn, modified and simplified his delivery. He didn’t go all stretch, all the time, but he eliminated some action. Here’s Hill early on:

And here’s Hill in a recent game against the Angels:

It’s easy enough to see the obvious change. The intended effects were and are subtle. Hill wanted to get his timing down, to get every part of his body in sync, and he now feels like he’s better able to do that. I feel like I’ve focused a lot of late on pitchers and their front foots, but you can see something with Hill; below, early 2017 Hill on the left, and more recent 2017 Hill on the right.

These shots are from nearly identical moments in the motion. Allow me to issue the usual caveats that individual screenshots aren’t always representative of regular mechanics, but Hill on the left here already has his front foot down. He’s getting ready to throw across his body, and because his foot is down, his arm is effectively going to be late. This would go toward explaining why Hill had so much trouble missing up and to the arm-side. Over on the right, Hill’s foot isn’t down yet. It’s nearly there, and Hill’s still going to throw across his body, but he’s going to throw less across his body, and his arm and legs are in better sync. Right-side Hill has the timing down, and so better location would follow.

That’s not the only recent simplification. I’ve borrowed some release-point plots from Baseball Savant. For now, ignore the pitch classifications, and just look at where the dots are scattered.

One of the Rich Hill tricks was that he’d throw from two different arm slots. There was the usual three-quarters, but then sometimes Hill would drop down to throw either his fastball or his curve. You know this twist — it’s the one that inspired Clayton Kershaw to mess around with the same thing. Hill, in the earlier part of the season, was dropping down with some regularity. Over the last three starts, Hill has barely done it. It hasn’t been completely eliminated, but Hill has focused his concentration on one delivery, for the most part. It’s tough enough to get one throwing motion down, nevermind two. The low slot could return, but it has been mostly absent.

Hill has modified his delivery and he’s mostly eliminated a second arm slot. Those are simplifications. But there’s also a new twist. In these plots, you see pitch spin on the x-axis, and pitch velocity on the y-axis. Look again at how the dots are arranged.

I suppose there are a couple things here. Recently, Hill has mostly stuck with one variation of his curve. That could be connected to eliminating the lower arm slot. But, also, look at the new clump that’s formed in the mid-80s, underneath Hill’s fastballs. It’s been written about elsewhere, but Rich Hill has added a cutter. It’s not a pitch he used to throw. The two-pitch pitcher has increased the size of his repertoire by 50%, sort of, and Hill had some of his own inspiration:

Because of Hill, Kershaw occasionally started to drop down. And at least in part because of Kershaw, Hill has unveiled a cutter with confidence. He used it a great deal three starts ago; in the last two starts, it’s easily been Hill’s third pitch. It’s likely to remain Hill’s third pitch, but it’s already been good for six strikeouts. Hill feels like he can throw the cutter for a strike, and it gives him something besides fastball and curveball tweaks. Hill has a history of playing with the curveball finger pressure, but the cutter is a whole new toy. Let’s watch it embarrass Cameron Maybin:

That’s a cutter for one strike. Okay. Hill subsequently doubled up:

Maybin was primed for the mid-80s, so he was no match for Hill’s 91 to end the at-bat:

Any three-pitch strikeout is likely to make a pitcher look like the best version of himself. Hill hasn’t struck everybody out on three pitches, and he remains a pitcher who sometimes puts people on base. Hell, just in those clips, you can see that Hill and his team are losing 3-0. But Hill has clearly gotten better, and there’s a link between that improvement and the introduction of the cutter. Batters, in time, could adjust to Hill’s new repertoire, but that’s how baseball works, and no pitcher is ever worse for developing a new and trusty tool.

So, Rich Hill: great, to irritating, and back to great again. Even at 37, he continues to evolve, and by just pointing to basic regression, not enough credit is given to Hill’s ability to tweak on the fly. Regression to the mean doesn’t happen because players just let it happen; it happens, mostly, because players work to return to their true-talent levels, which requires certain focus and effort. Rich Hill has gotten back to being terrific, but along the way, within a matter of days, he changed his delivery and altered his repertoire. There’s likely to be another clunker at some point, but suffice to say that Hill is earning your patience.

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.

Newest Most Voted
Inline Feedbacks
View all comments
6 years ago

An interesting observation here is that Hill has basically converted his repertoire into a clone of Drew Pomeranz (by eliminating the second arm slot and adding a high-80s cutter), who has himself often been compared to Hill. Just by eyeballing it, Hill’s cutter appears to have slightly less action than Pomeranz’s (might just be the off-center camera angle), but his fastball has more rise and his curve is tighter, so they’re not ENTIRELY comparable. But still, interesting the way they have (at least for now) evolved to the same spot.

Meanwhile, Pomeranz has been experimenting with adding a classic four-seam changeup. Hill’s arm slot should be compatible with that concept (unlike Kershaw’s, which is a little too over the top for it), so I wonder if he’s considered a similar line of evolution.