Looking Back at Pitch Mix Variation in 2023

Awhile back I took a hack at introducing a catchall metric to describe how much pitchers vary their pitch offerings, i.e. how often they throw distinct pitch types. I called that metric Pitch Mix Variation (PMV).

Following that introduction, I covered a couple different options for how to measure PMV. As this post does not devote much time to the creation or construction of the metric itself, you can check out that initial introduction or updated look here and here, respectively, should you want more background.

Instead, the goal of this post is to provide a glimpse into those (mostly starting) pitchers with the highest and lowest PMV values in 2023. For these PMV values, I relied on the Gini Index to measure PMV. The Gini Index is probably most popularly applied by economists to measure the variation in income among citizens of countries; the greater the Gini coefficient, the greater the wealth disparity/inequality among inhabitants. Applied here, the higher the Gini Index, the greater the disparity in usage among all different pitch types. Thus, the more a single —or couple— pitches are relied upon disproportionately, the lower the PMV (greater disparity in pitch usage rates means lower variation in pitch offerings).

For illustrative purposes, below is the Gini Index equation as well as the resulting value —.625— of a hypothetical pitcher that has a three-pitch mix and who throws a single pitch 50% of the time and differs to his secondary offerings each 25% of the time.

The Gini Index equation is really simple, which makes it a great option for measuring PMV. Additionally in this post, after calculating PMV, all data has been min-max normalized so that the players with the highest PMV values approach 1 and those with the lowest PMV values approach 0.

Below are those qualifying pitchers in 2023 with the greatest PMV.

In terms of PMV, Yu Darvish led the way in 2023. Statcast classified Darvish’s offerings, remarkably, nine different ways. Of his 2,217 total pitches, he didn’t throw one single pitch more than 413 times. Paul Blackburn and Josiah Gray sit in a near tie in the second spot, and each threw at least six distinct pitches last year according to Statcast.

Just glancing across the table above highlights the diverse offerings of those players. Every single pitcher employed at least two fastball variants; every single pitcher also made use of two distinct breaking balls. This is not the case for those pitchers on the other end of the PMV spectrum.

Among low-PMV pitchers, “Fastball-Slider Guys” seem to dominate. Justin Steele leads the way, having thrown fastballs and sliders for all but several dozen (94 total) of his pitches in 2023. He even eclipses Spencer Strider, who I consider to be a pioneer of two-pitch starter-dom (fairly or not, I cannot say). Overall, the majority of the pitchers on this list lean one pitch in particular —most often a fastball variety— and has just 1 or 2 other secondary offerings.

More generally, I’ll note the PMV is intended to be descriptive, not predictive. When controlling for measures like Stuff+ and Location+, PMV was not statistically significantly predictive of metrics like FIP or xFIP among these players in 2023, nor is it intended to be. Put another way, higher PMV values do not correspond to better pitchers or results. However, as one might reasonably guess, there is a modest negative correlation between pitchers’ Stuff+ and PMV (-0.19), suggesting that player with better raw stuff need not mix their pitches so much (Strider is a perfect example of that case). Pitchers with more pedestrian (MLB-level, anyway) stuff seemingly sometimes defer to a greater quantity of offerings to make up for that slightly lesser quality.

For players like Strider, Steele, and Hunter Greene, who have great stuff, the concept of PMV might become more relevant as their offerings inevitably lose a tick. Until then, it is fun to see them thrive at the game’s highest level all the while mixing their offerings less than their peers.

Lots is made of pitch quality and command/control among pitchers, and rightfully so. However, in my eyes, the way that pitchers go about mixing their various offerings is less discussed, but still fascinating. PMV seems to be a potentially useful shorthand to contextualize how varied any given pitcher’s arsenal is, and with any luck is something that can start a conversation here or there.