Patrick Corbin and Robbie Ray have more in common than pitching left-handed. This season, both have doggedly stuck with a changeup, despite poor results in the past. And while Corbin’s slider had been one of the best in the game and Ray’s had been one of the worst, they’re both using a fourseam fastball, sinker, breaking ball, and changeup, in similar percentages. In that context, the pitchers’ strengths could not be much more different, and yet they seem to be working from almost the same plan.

Their results have also been similar, and not in a good way. Corbin was every bit a top of the rotation starter in 2013, until he tired late in the season, and while he was not as dominant in his partial season last year, he did succeed (3.60 ERA) where nearly everyone else failed. He hit the ground stumbling in 2016, however, and currently has a 4.21 ERA and meager 6.39 K/9, getting stung by the long ball in most starts. Ray has looked worse (4.62 ERA), although he started the season strong with starts of 6 innings (2 ER), 6.1 innings (2 ER), and 6 innings (0 ER). Those three starts are still Ray’s second, third and fourth longest of the season.

Both pitchers have now made 10 starts, and we’re probably at the point at which we can start to draw some conclusions about what they’ve been doing, and what hasn’t worked for them. Some kind of new phenomenon for either pitcher this year may not reflect what’s likely to happen next, unless it also comports with what we know about pitching generally. Basically: if either pitcher were doing something new this season and we had reason to doubt it would work before the season started, we probably have enough of a sample to say it probably won’t work, moving forward.

Lefties are probably not the same as righties, but while we have an exhaustive study about all kinds of pitches and their platoon splits for righties… I’m just not aware of a similar study for lefties, and I’m not capable of that kind of number crunching. So let’s tentatively say that the general principles we derived from the righty numbers are probably relevant to lefties: the more horizontal the movement, the greater the platoon split; the faster the pitch, the greater the platoon split. That fits with convention, anyway. Curveballs and changeups are often used by lefties against righties, and not at all against lefties; sliders and sinkers tend to be reserved mostly for same-handed batters.

Here are the numbers for those pitch groupings for right-handed pitchers. On the left, the numbers inside the pitches is average velocity; on the right, platoon splits phrased as differences in Run Values per 100 pitches (click to embiggen).

12 pitches, velocity and platoon

Patrick Corbin

This season, Corbin has looked to use his slider a little less frequently, in favor of his change; back in spring training, that did look like it would come with a cost. And so it has. Corbin dipped below 20% slider usage in three starts this year. The results in those starts: 5 runs, 4 earned (April 22, Pirates), 7 runs, all earned (April 27, Cardinals), and 5 runs, 4 earned in his last start (May 26). In his other 7 starts, he sports a 3.67 ERA — shockingly similar to his overall outcome in 2015.

Corbin pitch usage breakdown

The differences in Corbin’s approach this season are small, but significant. Against right-handed hitters, he essentially doubled his usage of changeups, almost completely at the expense of sliders; against lefties, he’s also decreased his slider usage, making up the gap mostly with sinkers. The changeup approach should work, in a vacuum; changeups should have a smallish platoon split, although we don’t know how well it would fare against lefties, either (that’s one pitch, in 2016; he threw zero in 2015 to lefties).

There’s something here with fastballs. In general, fourseam fastballs are less prone to platoon splits than sinkers, and sinkers should be thrown sparingly against same-handed hitters. With Corbin, though, that’s flipped — and has always been. In his magical 2013 season, Corbin threw many, many more sinkers to lefties (27.5%), who did little with it (.373 SLG). It was a true primary pitch to righties that year, though, as he threw it 48.3% of the time (.419 SLG). The D-backs’ handling of sinkers is endlessly frustrating; several pitchers threw more sinkers to opposite-handed hitters last year, with thoroughly poor results. Here, in the one case where it’s been demonstrated to work, a pitcher is throwing it dramatically less.

Lefties did fare a bit better against Corbin’s sinker in 2013 than they did his fourseam, and that .125 SLG could easily be a mirage; Corbin hasn’t faced a ton of lefties this year, and he’s thrown just 24 sinkers to them (8 to end at bats). Throwing more sinkers against both righties and lefties seems like a worthy tweak, if just to get him back to his 2013 pitch mix.

The main thing, though, seems to be that it’s time to call an end to this changeups-to-righties mischief, for at least three reasons. First: the thing has had hilariously poor results; David Ortiz is the only hitter with a SLG above .621 this season, and .821 is extreme. Yes, the sample size is small — but not all findings are equal in terms of sample size. Flip a coin 100 times and get heads 70 times, and that’s enough to wonder if something is wrong with the coin or the flipper; get heads 19 of 20 times, and you could draw a conclusion with a similar confidence level. The changeup is bad news, and while Corbin’s fourseam has been unspectacular and there may be health reasons to hold off on the slider, anything looks like a better idea than the change.

The second reason is that the velocity readings support that conclusion. Corbin’s fastballs have had average release speeds between 92 and 93 mph; the average velocity of his change has been 85.5 mph. That’s just 7 mph, and while some pitchers get away with that, the pitchers that do aren’t necessarily getting whiffs, but weak contact. Weak contact doesn’t seem like a working plan in the desert, anyway. The only way to evaluate a changeup with that small of a velocity differential is results, and like we just said, the results are speaking loudly and unequivocally.

The third reason: Corbin’s change was probably never going to be especially helpful against RHH, specifically. The beauty of the Max Marchi numbers in the figures above comes in the greater number of pitch clusters. Yes, some changeups have reverse platoon splits. But the pitches more like his “Power Change” category? Not so much. The velocity is even higher than Marchi’s average for the “Power Change” (which points in the direction of a regular platoon split), and the movement is very horizontal (which also points in the direction of a regular platoon split), at 8.74 horizontal inches, on average, and 5.41 vertical inches of movement. Corbin’s change shouldn’t have worked against RHH, probably; and it also hasn’t worked. Time to move on.

Robbie Ray

I’m on record somewhere as saying I’d like to see how well Ray could do without throwing any breaking balls at all, but, wouldn’t you know, his breaking balls have been surprisingly successful this year; previously, it looked like the breaking balls classified as curveballs were merely cement mixer sliders, but that no longer seems to be the case. Last week, I wondered if they were having a knuckleball-like success, but the “why” isn’t necessarily important. They’re working, and Ray is throwing the curveball version of the pitch more to right-handed hitters — a strategy that seems anchored in those Marchi platoon splits.

Beyond that, though, to the extent Ray’s habits have changed, they’ve changed just like every other pitcher’s had, run in the face of evidence, and haven’t worked.

Ray pitch usage

The vast majority of the D-backs’ sinker experiments did not work last year; they had Rubby De La Rosa, Chase Anderson and other right-handers throwing the pitch much more to lefties, and the results were shockingly terrible. Ray’s sinker was a revelation last season — and it even worked against RHH, who slugged just .386 against it in 70 at bats that ended on the pitch. Last year, though, RHH got just 2 sinkers for every 5 fourseam fastballs, and by increasing sinkers at the same time he reduced fourseam fastballs, that ratio is now more like 3 sinkers to every 4 fourseamers. Ray’s sinker has more horizontal movement; with it no longer functioning as a surprise, right handed batters’ slugging percentage against it has mushroomed. We ought not to be surprised, either.

Ray probably should stop throwing so many sinkers to right-handed hitters. That doesn’t appear to be too complicated. As we said, we’re at the point of the season now when we can start to pass judgment on new experiments if 1) we had reason to think they wouldn’t work, and 2) they haven’t worked. There’s a stronger case here to be made than the one with Corbin, even; Ray’s fourseam has always fared well, and there’s no compelling reason to rachet down his use of that pitch against RHH, other than the foul ball phenomenon. A few extra foul balls seem like an easy price to pay for fewer home runs, and singles, and doubles, and triples.

We’re still in small samples land on Ray’s change and breaking balls, but not quite as much against RHH. Still, just as with Corbin, Ray’s attempts to get changeups by right-handed hitters have been unsuccessful, and it’s probably time to end that experiment. Since the beginning of 2015, the only pitch that RHH have had more success with off Ray is his sinker; we covered that. What little advantage Ray seems to have gleaned in 2015 by having the change be a surprise has now vanished with a fairly small uptick in usage. Like Corbin, Ray only gets about 7 mph of differential between his change and his fastballs, and while it’s not quite as horizontal as Corbin’s version of the pitch, it’s still very much a “Power Change,” a pitch on which pitchers don’t normally gain a platoon advantage anyway.

Against RHH, Ray should probably continue to lean on his breaking balls; we should have guessed a true curveball would probably work against opposite-handed hitters, and it has. He’s now thrown 75 sliders and 39 curveballs to RHH this season, and he’s yielded just 4 hits (all on the slider), while sporting some fun results (21.3% of swings on the slider have been whiffs; just 5.1% of the curveballs have been fouled off). It’s time to get the hell away from the sinker against RHH, and the change, too; bring that sinker down back to the 20% range, at most, and get that change back below 10%. If the team wanted to explore more curveballs, that seems at worst like a worthwhile experiment. Otherwise, it’s a steady diet of one of the best fastballs in the majors in Ray’s fourseam, and that’s not something to be afraid of. A 7 IP per outing pitcher, Ray is not.

And what about LHH? Just like Corbin, Ray is throwing a curiously small percentage of sinkers when facing lefties, and that qualifies as outright bizarre. When we look at the Marchi platoon numbers, we’re seeing one of two things (or both): either a pitch just has much more success against hitters from a certain side, or hitters from the other side flat-out punish it. My unscientific impression has been that with sinkers, it’s definitely both. Lefties have slugged a ridiculously low .206 against Ray’s sinker, which seems criminally underused. Lefties have historically treated Ray’s breaking balls as balls of meat served with al dente pasta and a heaping helping of marinara sauce, and it seems like they’ve just sat on the fourseam, which should have more success than it’s had. There’s no obvious direction to go in with respect to breaking balls and lefties. But with respect to the fastballs, it’s not just that the sinker is supposed to have success against lefties, and has been successful; it’s that by throwing them more often to lefties, the fourseam could get better, too.


Live, and learn. To the extent some Corbin and Ray experiments this season should not have panned out and haven’t panned out, it’s time to turn the page. It is excruciating to see over and over that the D-backs keep trying over and over again to get pitchers to throw sinkers to opposite-handed batters — Corbin is the only exception I’ve seen, and even then, that’s not an excuse for throwing it less to same-handed batters, but a mandate to throw it much more. It’s time for both pitchers to drastically reduce their use of changeups to RHH, and to drastically increase their use of sinkers to LHH. Both pitchers should tweak their fastball approaches to RHH, as well. It’s far from a guarantee that either pitcher would suddenly post ace-level results if that happens, but at least this is an adjustment they can control, unlike health, or velocity, or movement. Repertoire changes like these are simple to make, and when results direct those changes this obviously, it’s a glaring condemnation of the team’s staff to not make them.

Tagged with:

9 Responses to Tweaks to the Pitching Plans of Corbin, Ray Are Necessary, Could Pay Dividends

  1. coldblueAZ says:

    Have you considered sending this to Mike Butcher? Might benefit all of us fans if you would.

    • Ryan P. Morrison says:

      Well… whiffing on something like this is the kind of thing that happens when you have an analytics department charged with finding answers to questions, instead of an analytics department authorized to ask its own questions.

      • coldblueAZ says:

        There is one thing this non-analytic fan can say with certainty: If the team continues to run its staff out with the same game plan and expecting better results, we might as well give up all hope of ever contending.

        • Ryan P. Morrison says:

          The sinkers thing does seem insane.

          Not saying an analytics team should be in the driver’s seat. If they had the power to bring things to others’ attention, though… having a burden of persuasion is probably a good idea, but no one seems to be persuading anyone of anything.

      • David Masters says:

        So, you are in favor of Corbin dumping the change up, which goes against all baseball convention, or you think this article is off center? The only takeaway on Corbin is that he should throw more sliders, whether it is to right handers or left handers, but saying you can project right handed numbers on left-handers is as misguided as saying AJ Pollock needs to heal faster.

        • Ryan P. Morrison says:

          Hello David, and welcome. A good touchstone for the content here: I might be overly defensive about Jake Lamb, but other than that, the goal is to frame questions and explore possible answers. It’s frequently the case that I start out thinking one thing, and end up drawing a different conclusion based on the investigation.

          If you think flipping RHP and LHP results is that misguided, you might have misunderstood what I was doing. With RHPs, we have some fantastic evidence on platoon splits based on a more nuanced grouping of pitches. The results seemed to follow two basic patterns: 1) when pitches’ movement was more vertical than horizontal, they seemed to be less prone to platoon splits (and vice versa), and 2) slower pitches tended to have smaller platoon splits.

          It’s those principles I was looking to port over to LHP, but note that I wasn’t relying on them — what I was saying was that if the lessons those principles teach comport with the results we’ve seen, maybe that should be a reason to trust the numbers we’ve seen significantly more. Doesn’t that seem fair? If not, I’d encourage you to link over to the first post in which we looked at those Marchi numbers.

          I think the takeaways on Corbin are:
          1. Even though he’s had some success with his sinker against RHP, the natural tendency of a pitch with movement like that is to be more effective against same-sided hitters. And as a matter of fact, he’s had tremendous success with the pitch against LHP, too. So why shouldn’t he throw the sinker much more to LHP than he has this season? More sinkers to LHP seems like a worthy experiment.

          2. Quit throwing changeups to RHP. Not only has it not worked, but based on those principles… as it happens, Corbin’s change is a very hard changeup (fitting squarely into Marchi’s “Power Change” category), with a ton of horizontal movement. Most changeups are not like that, and in this case, the difference between Corbin’s change and most changeups probably does make a difference in terms of success against RHP. So no more changeups against RHP.

          Those were the main things; I’m conscious of the fact that Corbin probably can’t throw the slider 30% of the time. But no, I’m not saying dump the changeup completely. The information that makes me think it’s a bad idea against RHP also makes me think that it would probably be a GOOD idea against LHP — but we have no data, because he hasn’t actually done that. That’s worth a try, wouldn’t you say?

          In one way, the D-backs are going hard AGAINST baseball convention by insisting that most pitchers throw sinkers more to pitchers of the opposite hand. That’s never supposed to work. But sometimes a two-seam fastball really isn’t a sinker; when Tony La Russa had Kyle Lohse in St. Louis, Lohse’s “sinker” had about ten inches of “rise.” That pitch should be less prone to platoon splits, and didn’t have much of a platoon split.

          The same idea should apply to changeups. Is this pitcher’s change really similar to changeups that have low (or even reverse) platoon splits? Or is this pitcher’s change really similar to changeups that have not fared well against hitters of the opposite hand? Corbin’s is the latter. That’s all. Okay?

  2. John says:

    Corbin’s slider was why is was dominating before his a injury, don’t need analytics for that. It rides up the barrel of a right handed hitter and look alike a fastball over the middle. It very similar to Madison bumgarners cutter that helped the Giants win…

    Does the slider put to much torque on his arm post recovery? Or is the new coaching staff thinking alternative pitches are better?

    All I know is that I have stopped watching a lot of games because of this coaching staff and front office inability to make these guy grow and reach…

    Escobar pitching is a joke and slap in the face to fans with Shipley in triple A, especially on Memorial day running him out there when the fans have that day off…

  3. […] looked at what Ray was doing and the success of his pitch mix on the last day of May, and the results were stark; Ray was having really great success against […]

  4. […] horizontal movement makes it more like changeups with bad platoon splits than good splits, and Corbin probably shouldn’t rely on his change against righties. A return to a high-and-tight fourseam could also work well, while asking him to go for that low […]

Leave a Reply

Your email address will not be published.