The high fastball and the big curve

Late this past season, Padres righty Andrew Cashner came back from a shoulder injury with a new twist on his repertoire -- again. This time, he featured a few more high fastballs and big curves than he had in the past. You'd think those two pitches are often linked across baseball, but the numbers aren't as clear.

The last time Cashner came back from injury, he focused on throwing more two-seamers to get quicker outs, altered his changeup grip, and changed his grip on his breaking pitch. These changes were made with his health in mind, but they also served to make him a more complete pitcher.

This year, when he came back from shoulder inflammation that sidelined him for two months, Cashner again came back with a wrinkle. "I started throwing the four-seamer more in order to establish the high strike," Cashner said before a game against the Giants in late September. Of course, the pitcher knows best about his approach, but it's worth noticing that he threw an average of only three more four-seam fastballs per game when he returned compared to the same timeframe before his injury. And that his heat maps before and after his injury aren't conclusive on the subject of high four-seamers.

He pointed out that he threw more curveballs when he came back, too. He'd thrown nine in his first 12 starts before he got hurt. He threw 18 curves in the seven starts that came after his stint on the DL. This September was the month in which Cashner showed the best whiff rate on his curveball in his career.

The second part of the plan was paired with the first, he admitted. That high fastball is "on the same plane" as the curveball. That makes all sorts of intuitive sense, considering the idea of a 94-mph high fastball coming in the same general area as a big, dropping slow curve. It's the kind of thing that seems to work for other pitchers.

September 15th against the Phillies, Cashner threw a good number of high four-seamers and low curves (though you can see he missed a few spots). Looks like the plan in action:

In Jon Roegele's great work on pitch sequencing, he found that that pitches that look the same at first but don't end up in the same place at the plate have good outcomes. And that the four-seamer is the pitch most often connected to the curve when it comes to these "€œin-band"€ pitches.

What's also interesting about these pitch tunnels is how relative they are. Average four-seam fastball height at the plate doesn't really have a statistical relationship with the shape, frequency or success of your curveball usage at least.

Four-seam fastball height is not at all correlated significantly with curveball vertical movement, usage rates, swinging strikes, called strikes or groundball rates within a decent sample of 113 regular curveball users. It doesn't get any better in the bigger sample of 250 who threw 50 or more yakkers last season, or even if you add more years of data.

Curveball users don't generally throw more high fastballs than the rest of the population, even if last season the curveball set had a slightly higher average fastball height than the league average (by .43 of an inch, in a world where pitchers might be missing their spots by more than a foot on average). You're not more likely to use your high fastball more if you use your curveball more.

So "high fastball" is maybe a term that's relative to each hitter and each pitcher and each umpire. And maybe there's just too much of a difference between curves and even high fastballs at 21 feet away. After all, curveballs have the lowest swing rate in baseball -- batters often see them coming and decline to offer.

Or maybe these things are constantly being adjusted on a small scale by each pitcher -- an extra curve per game over six weeks, Cashner might say -- and so it's just a little hard to be all that prescriptive about pitching.