Considering Pitch Movement Across Types of Pitches

The creation and availability of pitch tracking technology has naturally beget all types of interest and analysis, and for good reason. Pitch characteristic data is rich in information and offers a path to better understand the relative success of any given pitcher’s offerings. What’s more, it is simply enjoyable to consider how pitchers manipulate the baseball, refine their craft, and seek success.

While research into pitch movement is already rich and continuously growing, it seems to predominantly quantify various pitch values in isolation, or to aggregate a pitcher’s offerings in one concise number summarizing their “stuff” as a whole. Taking a pitch-specific perspective, it’s generally accepted that additional pitch movement is good, and maximizing that movement in turn maximizes positive outcomes for a pitcher. Squeezing more horizontal movement out of a slider and calling it a sweeper has gained popularity for a reason, after all.

Alternatively, here I look to consider pitch movement across two pitches at a time, through the lens of stereotypically complementary pitches. Put another way, I look to consider the movement of that slider while simultaneously considering a complementary sinker that darts, in the opposite direction, to a pitcher’s arm-side.

To do so, I am drawing on a couple of the most classic examples of pitches that are considered to be complementary and their characteristic movement. First, four-seam fastballs and curveballs and, second, sinkers and sliders. Pitchers who throw either of these combinations leverage the fact that the those pitches generally move, or break, in opposite directions. In the case of four-seam fastballs and curveballs, I consider vertical movement, while in the case of sinkers and slider, horizontal movement is the focus.

Data was pulled from FanGraphs leaderboards for the 2024 season and limited to those pitchers who threw at least 100 innings (arbitrary, I know) and either combination of the aforementioned pitches. In the case of four-seam/curveball pitchers, 79 pitchers qualified. For sinker/slider types, there were 90 qualifiers.

To offer a quick view of things, below is a scatterplot covering the movement on all fastballs and curveballs.

When looking at the data above, it should be noted that “four-seam” fastballs here come from the “FA-X” value in FanGraphs (provided by Statcast), which in fact rolls up four-seam fastballs as well as unclassified fastballs. So, just know that there might be some muddying of the water there from a classification perspective.

Still, this chart illustrates movement I would have broadly anticipated: four-seam fastballs have positive vertical movement while curveballs generally bend downward. This gap between pitches is what is of interest here; the goal being not to consider a fastball’s vertical movement in isolation, but rather the absolute difference in vertical movement between the two pitches for any given pitcher. But more on that in a bit.

As for the second group of pitchers, the scatterplot below depicts qualifying pitchers’ sinker and slider movement.

Here, I admit to being a bit taken aback by the amount of horizontal movement of sinkers on the whole outstripping that of sliders. Still, the clusters of pitches is pretty clear. It is also worth noting here that pitcher handedness makes for clustered data on either side of the vertical axis depending on that handedness.

So, how is pitch movement measured across these types of pitches? To keep things as simple as possible, the vertical distance —in the case of four-seams and curveballs— between average pitch movement was calculated. In the case of sinkers and sliders, the distance between each pitch’s average horizontal movement was calculated. Alternatively, I might have calculated the Euclidean (“as the crow flies”) distance between the average movement coordinates of those pitches, but in this case the emphasis was on the single direction of movement for pitches —either “east west” for sinker sliders or “north south” for four-seam curveballs— as opposed to incorporating movement on the other axis.

With that, we can jump into the data. First, a view of fastballs that might “play up” given their movement when also considering that pitcher’s curveball’s opposite movement.

The chart above might be a lot to digest at first glance. Using Max Fried as an example might clarify things.

Fried’s four-seam fastball, on average, doesn’t have all that much vertical movement —6.51 vertical inches, or ~9th percentile in vertical movement among pitchers— but his curveball has excellent vertical movement —negative 9.76 vertical inches, or ~99th percentile— which means that the “combined vertical movement percentile” among these two offerings is actually quite high, ~82nd percentile, when taken together.

As such, when simply subtracting the fastball movement from the combined movement profile (not terribly scientific, so it goes) Fried’s fastball leads this list in fastballs that might “play up” given the presence of the high-movement curveball. Put another way, the thinking might be that Fried’s curveball changes a batter’s eye level so much so that his fastball can have middling movement in the opposite direction but still make for a big difference in movement between the two. While 6.51in of vertical movement isn’t all that much, the 16.27in of range between those pitches vertically ranks quite high among all qualifying pitchers.

Next up are curveballs that might play up on the back of high-movement four-seams.

Here, Tobias Myers throws a fastball with such strong vertical movement that his curveball, which doesn’t have exceptional drop, might still play up; together, those pitches have vertical movement that differs 15.11 inches on average, which is in the 64th percentile of combined four-seam/curveball vertical movement difference. His case is opposite that of Max Fried. Those other players on this list also have relatively strong “combined vertical movement percentiles” relative to their curveball vertical movement in isolation.

The next two tables cover, in much the same manner, sinkers and sliders, respectively. Of note though is that some pitchers in this group had sinkers and sliders that, unintuitively to this blogger, moved in the same horizontal direction. As such, just note that the “combined horizontal movement” field here captures the difference between the two pitches’ movements, regardless of whether they are breaking in the same direction.

While some of these pitchers appearing on this list pass the eye test, there are others I am much more skeptical of. Whatever the case, an appearance here in my mind is less a referendum on any particular pitch (after all, horizontal or vertical movement alone are just one ingredient to an effective offering) than something to note when considering a pitcher’s arsenal more generally. As such, these last four tables may well be misleadingly labeled. Whether or not the aggregate movement of two pitches meaningfully influences the value of a single pitch remains to be seen, making “sliders playing up” an unfounded assertion for a table title.

Delving into that question would require considerably more time and would likely meet obstacles in the way of statistical rigor. To begin to get at such a question, comparing the raw stuff for any given pitch to the actual value/outcomes from throwing that pitch would likely be a first step. Do Max Fried’s fastball outcomes outperform that fastball’s “stuff,” for instance? Well, the presence of his curveball might be an omitted variable, depending on the model used for stuff. Whatever the case, there may be some value in considering pitches this way, if only to be thinking a bit differently.