Spin that curve: How Richards uses RPM to baffle hitters

Earlier this week, we looked at "the spectrum of Statcast™," showing how spin rate and velocity intersected for various pitch types. The only pitch in baseball this year to top 3,000 rpm was Garrett Richards ' curveball, which came in at 3,086 rpm, making it among the game's most impressive offerings.
We couldn't help but tweet out what that curve looks like, here destroying poor Scott Schebler...

... and then Richards himself asked a very valid question in response:

It's absolutely the right thing to ask. Most of the first year of Statcast™ has been about identifying what we can measure, and how much of it is useful, and why. Spin rate in particular stands out, because while measuring things like foot speed or arm strength is simply putting numbers to things we've always seen, spin rate is something most of us had never even considered before 2015, much less put numbers to.
So, to Richards' question: Is high spin on a curveball good? To answer that, let's investigate what different levels of spin do to batted balls off of curves.

There are a few interesting takeaways here, the first of which is that spin for curveballs works in opposition to what we learned about spin for fastballs. While high spin for fastballs generally means more fly balls (since the spin helps the ball repel gravity for slightly longer and stay up) and low spin induces grounders, it's the exact opposite here.
Curves with spin in the lower range -- between 1,000 rpm and 1,750 rpm (guys like Scott Kazmir and Kyle Gibson) -- tended to get fewer grounders, in the low 40 percent range. Curves with higher spin were more likely to jump to 50 percent or higher in terms of grounders. Richards, the king of high spin, managed a stellar 75-percent grounder rate on his curve, which is to say that three quarters of the curves that hitters make contact with went into the ground. Only twice all year (Chase Utley and Luis Valbuena) did Richards allow an extra-base hit on the curve.
What makes the spin on a curve so different from the spin on a fastball? In both cases, the spin is doing the same thing, which is doing its best to drive the ball in the direction of the spin. But with a fastball, the pitch is released with backspin, which is another way to say the ball is spinning front to back, from the pitcher's view, and against gravity. With a curve, it's with topspin, or essentially spinning toward the plate, and the ground.
In that case, more spin on a curveball should mean more downward movement, and in Richards' case, that's exactly what it means. In 2015, 207 pitchers threw at least 100 curveballs. Check out the vertical movement leaders:
2015 curveball vertical movement leaders (minimum 100 thrown)
1. Mike Fiers, -11.99 inches
2. Richards, -11.43 inches

3. Evan Scribner, -11 inches
   
4. Chris Tillman, -10.08 inches
   
5. Joakim Soria, -9.96 inches
   Richards' curve drops nearly a foot on average, and it's pretty difficult to get elevation on a pitch like that.
   You might also notice that Richards ranks second behind Fiers, and that might make sense, given that the Astros traded for him alongside Carlos Gomez last summer after being the team that publicly put spin rate on the map when they rescued Collin McHugh from obscurity and turned him into a capable starter because they saw something in his spin rate.
   But while Fiers also has an excellent curve that induces grounders, he's not nearly as high as Richards is on the spin charts. In fact, his average of 2,534 rpm is merely 33rd. What gives?
   We've learned that not all spin is created equal, particularly as it pertains to breaking pitches. A pitch's total spin is comprised of useful spin (which increases the movement of the ball) and what's often referred to as gyrospin, similar to bullet spin, which doesn't really do much of anything. What we've learned as we measure total spin is that it's important next to try to identify how much of the two types of spin make up the total number -- if Fiers and Richards have the same useful spin, but Richards is also adding on additional less useful gyrospin to inflate his total spin number, that would sure explain the movement discrepancy.
   So, back to the original question: Is more spin on a curveball good? It seems like it might be if you want movement and grounders, and while fastballs that induce flies can be useful, slower breaking pitches in the air don't generally end as well for the pitcher. David Hale, for example, had one of the lowest curve spin rates, and he allowed a slugging percentage of over .500 against it, though Coors Field surely isn't any pitcher's friend -- and we shouldn't say with certainty yet that "more is better," because we need more evidence.
   With the caveat that we need to learn more about useful spin, and with the knowledge that speeds and setting up the hitter with other good pitches certainly matter as well, spin certainly has an impact on a curve. In Richards' case, even if he doesn't know or care about the spin, it's served him well. Who could argue with a career .149 batting average against it?