Is That Curveball Everything Robbie Ray Needed?

Time to sound the New Pitch siren because Robbie Ray is throwing a curveball! And, at least early in the season, it looks like it matters: after a year spent wondering why his balls in play kept finding grass and suffering while his run-prevention marks failed to match his fielding-independent ones, the Arizona lefty finally has the numbers you might expect for a guy who’s been among the top 15 in strikeout rate among starters since he entered the league.

Ray has long had great velocity and a devastating slider, but that doesn’t mean that the player-evaluation community was always of a same mind about him. Opinions varied greatly when he was traded to the Tigers for Doug Fister and then to the Diamondbacks as part of a deal that netted the Tigers Shane Greene. Mostly, the worry was that he would need to learn a changeup.

That changeup never has had above-average movement or velocity gap, and Ray never trusted it enough to throw it even 10% of the time, so those doubters were right, maybe. While watching Ray’s balls in play continually fall for hits last year, despite his great strikeout and walk numbers, I wondered if manipulating a few breaking balls could be his way out. Maybe it would be a cutter or a curve — both pitches effective against opposite-handed hitters. After all, not everyone needs a changeup. I once asked Randy Johnson, now a special advisor to the club, if he did something similar and would preach the same to Ray. “No,” Johnson affirmed. “I learned a split-finger changeup.”

What works for the goose doesn’t always work for the gander — I grew up in a German Jamaican household, not sure that saying works, but let’s use it — so Ray’s using his changeup less and has replaced those pitches (and more) with a new curveball. Here’s what it looks like:

It’s a good pitch! It has good results and is fast and has good drop. Ray’s doing well! End of story?

Maybe not. I remember a few good stretches from Shane Greene, for one. And I remember looking at his blob of breaking pitches once and wondering… if you have five breaking balls, do you maybe actually just have one? I mean, Greene’s slider/cutter/curve combo was so diffuse that it once called me to question what a breaking ball is, anyway.

In related news, here’s a Ray slider from the same game. Huh. Looks kinda similar?

[Edit: Wait, no, PITCHf/x called that a slider (to our point) but this is really a slider. It does look a little more different, more side to side than up and down.]

Existential questions aside, what we’re wondering here is how different a pitch needs to be in order to officially qualify as a new pitch, right? If you have two breaking balls but they morph together, maybe that’s actually even worse than having one. Zack Greinke once had to table the cutter because throwing that, in combination with his slider, made both pitches worse. “Whenever you throw two similar pitches, they end up meshing,” Greinke said back then.

How might we measure this sort of thing? In an admittedly terrible and ham-fisted approach, I took all starters who threw 100 innings last year and removed anyone who threw a changeup more than 7% of the time. Then I took out the cutter-first pitchers because cutters are just a mess to analyze. That left me with 15 pitchers who threw both a slider and curve as their main pitches in 2016.

I then took the movement numbers (in both directions) and velocity readings on the slider and curve and subtracted them from each other to get a measure of how different the two breaking balls were last season. Here’s the list sorted by the “distinctness” of the two breaking balls.

It’s interesting how there’s a general rise to the top when it comes to talent, and that Ray would end up last on this list with a one inch horizontal difference, five inch vertical difference, and 3 mph velocity gap between his two breaking balls.

In one way, yes, we are testing the system. The system — PITCHf/x, in this case — called those pitches different things and maybe it just messed up a lot. I’d argue that this actually helps the case of this as a measure of distinctness. If the pitches were very distinct, the system would easily separate them, and their numbers would reflect that here.

It’s also generally a bad move to add numbers on different scales, but when it comes to the vertical difference in inches and the velocity gap, the scales were so similar I’m not sure it’s a big deal for this little enterprise. It might matter when it comes to horizontal numbers, though. Gerrit Cole and Corey Kluber appear a little low on this list for my taste, and they both have superlative differences in the horizontal dimension. Kluber’s breaking ball is nearly a unicorn, and even here it stands out; he’s basically a sideways Kershaw except for the velocity gap.

In any case, this can serve as an interesting footnote to the New Pitch conversation that’s so tantalizing. What constitutes a new pitch? If the pitch isn’t distinct enough, is it just a wrinkle on an existing pitch? And is a wrinkle not as good as a totally separate new pitch?

We call changeups feel pitches sometimes, but aside from a few pitchers like Kevin Gausman and Kyle Hendricks, pitchers usually only have one changeup. Instead, it’s the breaking pitches that are subject to the whimsy of “feel.” Pitchers often use one grip and slightly altered mechanics to try and manipulate the result on the ball.

That last is accepted fact. And yet, in this context, it has new meaning. Fiddling with the movement on a few pitches that have similar grip — the multiple breaking ball way — walks a much finer line than that traditional pitcher scouts have long lusted after.

Fastball. Breaking Ball. Changeup. Those three, in tandem, don’t have problems when it comes to distinctness.