Release Points: We’ve Been Getting Cutters All Wrong

Tim Hudson throws a cutter. Kenley Jansen throws a cutter. But are we really talking about the same pitch?

It wasn’t that long ago that features popped up across the baseball world heralding the rise of the cutter and the profound effect it was supposedly having on the hitting-pitching balance. Here’s a Sports Illustrated piece from 2011:

There is a mysterious and magical pitch that is changing baseball. The pitch is saving careers, perhaps even extending them, turning journeymen into shutdown relievers and restoring the dominance of aging All-Stars. It's the secret reason why the game's power balance has shifted from the hitter to the pitcher. The pitch screams toward the hitter with the speed and the spin of a fastball and on a plane as flat as a vinyl LP and then, just as it begins to cross the plate, the ball darts like a badminton birdie. “Your brain is telling you fastball,” says Angels rightfielder Torii Hunter. “Then the ball breaks, and you're done.” …

The pitch is not only why the Yankees' Mariano Rivera is the greatest closer ever, but also why the reigning NL Cy Young Award winner, Roy Halladay, is having one of his most dominant seasons, at age 34. The pitch is why virtual unknowns such as Cleveland's Josh Tomlin, St. Louis's Jaime Garcia and Tampa Bay's James Shields are blooming into All-Stars—and All-Stars such as Haren, Philadelphia's Cole Hamels and Boston's Josh Beckett are as good as, or better than, ever.

Hold on a second. Dan Haren and Roy Halladay did not throw the same pitch, and you wouldn’t confuse an Andy Pettitte cutter for a Mariano Rivera one any day of the week. What exactly are we calling a cutter? Here’s the SI piece again:

This is the cutter, though to the untrained eye it may look like a slider—the difference is often subtle, though they are two clearly distinct pitches. A slider breaks horizontally but also has downward movement. A cutter moves mostly laterally. "The difference is that the cutter is thrown harder, it doesn't have as much depth and the break is much shorter," Patterson says. "If someone's throwing [his fastball] at 90, their slider should be at 82. Their cutter should be at 86, about a four-mile-per-hour difference."

A four mph difference might be okay for a rule of thumb, but what do you say to guys whose cutters are the fastest pitch they throw, like Jansen? What about someone like Pettitte whose cutter was six mph off his fastball and had much more depth? Are sliders and cutters really “two clearly distinct pitches”?

The term “cutter” is kind of a catch-all term for a pitch somewhere between a slider and fastball that is not considered one or the other. But according to one point of view, the “cut fastballs” of Rivera and company can basically be considered fastballs, and what is said to be a “cutter” is really a hard breaking ball. Dan Duquette implicitly acknowledged his adherence to this view in a controversial 2012 interview where he said the Orioles discourage their farmhands from throwing cutters. He said that major-league pitchers almost never find success with the pitch, and had a ready response for the retort that Mariano was a prime counterexample: “That’s a fastball. That’s a fastball. That’s his only pitch, he’s a one-pitch wonder. It’s his fastball.” Got that? Mariano Rivera did not throw a cutter.

Cutters tend to be more nebulously defined because they are picked up by pitchers usually after they already have the core elements of a repertoire: a primary fastball (two-seam or four-seam), a curveball or slider as their breaking ball, and a changeup. That’s all, at a minimum, you’re expected to have by the time you’re called up to the show.

By contrast, a cutter is usually a later-career addition. Pitchers often start throwing cutters because they think there’s a hole in their repertoire, or one of their current offerings—either their fastball or their slider—isn’t up to par. Dan Haren famously added the cutter because he was losing fastball velocity and wanted a new pitch to keep hitters off balance. Rick Porcello began messing around with a cutter in 2012 because his slider was awful. Brandon McCarthy used it to balance out the sinker he began throwing in 2011 as part of his transformation into a groundball pitcher. There’s also the group of pitchers—populated notably by Rivera, David Robertson, Jansen, and recently Michael Pineda—for whom cutters are just their default fastball. The quirk of the cutting action on their heater is part of their signature.

In other words, cutters play a more versatile role than most other pitch types. They are often pitches by design, individually tailored for a need and purpose.

Luckily, an opportunity presented itself to test whether there are major differences in cutter type. In the course of tagging pitches from the PITCHf/x system, Harry Pavlidis and I learned that some teams don’t really treat cutters as a distinct pitch type when they do data analysis and scouting reports. They like to think of a pitch as intended to be a breaking ball or a fastball. Piqued by this information, we proceeded to group all of the pitchers who have thrown a cutter in the PITCHf/x era in either a “fastball” or “slider” category judging by the spin and speed characteristics of each pitcher’s cutter. A few of them were close calls, but most were pretty easy to sort.

The point of the following exercise isn’t necessarily to try to find out whether there are neat, distinct sub-types of cutters. (As it turns out, our data suggests there’s actually more of a spectrum than two clearly defined clusters.) The idea is more to explore whether teams that buy into a fastball/slider mentality have some justification in splitting up the cutter pitch category. If there’s not much difference between these two populations we created, it would suggest those teams are making a questionable decision in their pitching and data processing philosophies.

As mentioned, Harry and I manually sorted cutters into two groups—those we thought were intended to act like fastballs, and those that were supposed to be like sliders. In originally dividing up these cutters, we had no explicit intention of making data comparisons, no access to outcomes and batted ball data, and we have not altered our selections since. Whatever selection bias existed in the process, it was not for the purpose of making this column more compelling. To keep the exercise simple, I’m including only right-handed pitchers in this case. Here are our results:*

This chart gives us a look at the physical characteristics of what we might call “hard” and “soft” cutters—and suggests there are, indeed, some real distinctions between the groups. As intuition would imply, the hard cutters are faster (+3 mph), have less horizontal break (-.077 in) and have more backspin (+2.06 in). The spin and speed combined give the soft cutters about five inches of additional “drop” in the actual path to the plate.

There’s also a subtle difference in the pitch result breakdown of the hard and soft cutters.

The last two variables were significant at .05 alpha, and they also support the notion of a more fastball-like cutter population (more pop-ups, fewer whiffs) and a more slider-like one. Remember, these significant differences exist within one pitch type.

All of this data bolsters the idea that there is more than one valid way to analyze cutters. At Brooks Baseball, we call cutters a distinct pitch, and personally, that’s what I’m used to. For most pitchers with a cutter, they have a discrete offering meant to complement their repertoire. But then there are also guys like Drew Storen: He says he throws a cutter, but in reality it appears to be some subtle variation on his fastball that is undetectable in PITCHf/x. Pitchers always bend rules and definitions, so it makes sense to analyze their work with a flexible mindset.

* For those curious whether one group throws harder on average, thus skewing the results, I checked what the average fastball speed was for each group (among those who threw both a cutter and a fastball)—and they were nearly identical: about 92.1 mph. Thus, it’s fair to compare the speed of each type of cutter against the same baseline.