Baseball Research Review #3 – Does Pitch Speed and Release Point Consistency Predict Pitching Success?

In this edition of the Baseball Research Review, we’re going to dig into a recent study which looked at predictors of pitching success in Major League Baseball. On the surface this study looks pretty straightforward, but it actually would have involved a ton of work, and some serious MATLAB skills. The authors should definitely be commended for their work! Special shout-out to Dr. Mike Sonne for passing this one along!

Study: Ball Speed and Release Consistency Predict Pitching Success in Major League Baseball

Authors: Whiteside, D., Martini, D.N., Zernicke, R.F., & Goulet, G.C.

Background:

What makes a great pitcher? Traditional baseball wisdom suggests that the ability to locate your fastball, change speeds, and throw your mix from a consistent release point all contribute to success on the mound. Oh, and throwing cheddar never hurt anyone, either! While these all seem like very reasonable assumptions, no empirical research has directly implicated any of these factors in pitching success at the Major League level. However, with the introduction of the PITCHf/x system in Major League Baseball stadiums, it is now possible to elucidate factors that contribute to pitching success in the most ecologically valid way possible — by directly analyzing them in the competitive arena.

And, that is exactly what the authors of this study set out to do.

Purpose:

The purpose of this study was to examine how ball speed, ball movement, release location, and variation in the three aforementioned variables (i.e., speed, movement, and release location)—as measured during games—related to pitching success in Major League Baseball, as measured by FIP (“fielding independent pitching”).

Methods:

To do so, Whiteside et al. obtained PITCHf/x data from pitchers who pitched at least 100 innings in Major League Baseball between 2008 and 2014. The authors excluded pitchers from their analysis if they had not thrown at least 100 of each of the following pitches during the study period: fastballs, changeups, curveballs, and sliders. In total, data from 190 Major League Baseball pitchers were used in this study. Next, the authors randomly-selected 100 fastballs, changeups, curveballs, and sliders (400 pitches in total) thrown by each individual pitcher during the study period. Then, for each pitcher, the authors calculated the average speed, movement, and release location of his four pitches, as well as the variability in those three kinematic parameters using data obtained from PITCHf/x. Finally, they calculated the magnitude of the relationship between each of these parameters and a pitcher’s success as measured by FIP. (Did you manage to get all that?)

Results:

Ball speed, release point consistency, the ability to change speeds, and a lower (i.e., a more horizontal) arm slot were all significant predictors of pitching success. Together, these factors explained 24% of the variation in FIP. Interestingly, neither the magnitude, nor the variation in ball movement (i.e., “break”) was significantly correlated to FIP. (I will discuss why I think this is the case later.)

Discussion:

Of all the dependent variables analyzed in this study, pitch speed (most notably fastball velocity) was the most critical predictor for pitching success, explaining 10.4% of the variance in FIP. As noted by the authors:

“this makes logical sense considering that faster pitches afford batters less time to detect the ball, forecast its trajectory, and execute their swing.”

So, “velocity is king!”, right? Well, perhaps, yes. But, in my opinion, that would be a poor interpretation of this particular study. Yes, pitch speed was the greatest predictor of pitching success in this study, but recall that pitch speed only explained 10.3% of the variance in FIP. So, clearly, pitch speed is not the only factor, nor is it necessarily the most important. (Remember: 76% of the variance in FIP was unexplained!) For the record, I don’t think anyone is claiming that “velocity is the only thing that matters”, much in the same way I don’t think I’m claiming “velocity isn’t all that important”. I think it is pretty clear that elite ball velocity is an important prerequisite to compete at the Major League level.

In regards to developing velocity, the authors make a great point, and I think it is worth pointing out:

“[…] greater pitch speeds have been associated with greater upper extremity joint loads and implicated in catastrophic ulnar collateral ligament injury. Accordingly, increases in pitch speed may yield diminishing returns beyond a threshold, which future research should seek to quantify.”

This is not to say, “don’t develop velocity”. I just think that there is a lot more to consider about a pitcher before making velocity the focus of his/her development. (Again, just my opinion.) Coaches need to assess their pitchers on an individual basis in order to identify the areas of focus that are most appropriate for each individual pitcher, given his strengths, weaknesses, training/injury history, and so on. Moreover, the saying, “all roads lead to Rome”, is appropriate here in that there are many ways to improve velocity, although some are arguably more appropriate than others for certain individuals. For instance, interventions aimed at improving efficiency of motor output (rather than interventions focused on improving output alone) can set the foundation for both improved resiliency and performance. (In this regard, pitching coaches can gain valuable insight from the strength & conditioning coach and/or the physical therapist.)

The second most critical predictor of a pitcher’s effectiveness in this analysis was release point consistency — most notably the consistency between pitch types. This makes a lot of sense. As the authors suggest:

“[t]he most likely explanation for this association is that a consistent release location would be expected to limit the perceptual cues available to the batter, thereby impeding their anticipatory proficiency.”

Batters use visual cues from the pitcher (e.g., forearm/wrist kinematics, release point location, etc.) coupled with early ball flight characteristics to predict where the ball will be when it reaches the plate. The video below does a good job explaining what I‘m talking about:

Video courtesy of Business Insider.

Interestingly, previous research in NCAA Division I pitchers failed to show a relationship between release point consistency — as measured in a laboratory setting — was not related to in-season pitching success. Arguments about ecological validity aside, the discrepancy between findings makes some sense. I think it is reasonable to assume the average hitter at the Major League level is more attuned to visual cues in a pitcher’s delivery and throughout early ball flight, than your average hitter at the college level. Thus, a pitcher with a more consistent release point would better defend himself against the superior perceptual-cognitive abilities of Major League hitters.

The finding that more successful pitchers produce more consistent release points may not come as much of a surprise, but it raises an interesting question: How does a pitcher generate a consistent release point? For as long as I’ve played the game, I have been told that, to generate a consistent release point, I need to be able to “repeat” my delivery. However, this traditional view of pitching biomechanics is at odds with more contemporary theories of motor control and coordination, which suggest that variability is an inherent and functional feature of human movement which affords stable task performance (such as a consistent release point). We have previously discussed the issue with “repeatability” here. This isn’t to say that, to the naked eye, a pitcher’s delivery shouldn’t exhibit some degree of “repeatability” (or, perhaps more appropriately, “stability” or “metastability”), it just means that analytical approaches which seek to understand pitching biomechanics at the macro-level will most likely be insufficient to explain inter- and intra-individual differences in something like release point variability, or, taking a step further, the ability to locate one’s pitches. This is a topic I would like to explore further in a future article.

Interestingly, neither the magnitude, nor the variation in ball movement (i.e., “break”) was significantly correlated to FIP. I think this may be the result of failing to control for release point variability. Let me explain. Off-speed that is delivered from the same release point as your fastball (and travels along a similar early trajectory) is going to be harder to “pick-up” than off-speed that travels along a more distinct path. So, a thorough analysis of “break” probably needs to account for between-pitch variability in release point and early ball flight kinematics. This is not a knock on the study, I just think it may be an important consideration when designing future research in this area. I am also not 100% on this, so I am happy to be corrected!

Now for a clarification from the Methods section of the paper…

The author’s note (bold text is mine):

“As a first pass, a pitch count test was performed to determine whether the pitcher threw at least 100 of each of the following pitches: fastballs, changeups, curveballs, and sliders during the analyzed period. If he had not, he was removed from the cohort and the script proceeded to the next pitcher.”

This is one point that, quite frankly, I am not so clear on, and would like to clarify with the authors. To me, this means that a pitcher who didn’t have both a curveball and a slider in his repertoire was excluded from their analysis. That’s excluding a lot of pitchers! And, to me, this could explain some of the weak correlations we’re seeing in this study. I mean, if you have a larger repertoire, maybe these variables carry less “weight”? But if you only have a three-pitch mix, maybe the ability to change speeds, and deliver your mix from a consistent release point becomes even more important? Maybe we would start to see a correlation between something like ball movement (i.e., “break”) and FIP if we looked at pitchers with a smaller repertoire? (Then again, maybe not. Just a thought!)

Finally, it’s important to reiterate that only 24% of the variance in FIP was explained by the dependent variables (i.e., pitch speed, release point consistency, variation in pitch speed, and a horizontal release point). This means that 76% of the variance could not be explained by the above pitching metrics. This underscores the importance of evaluating other factors that may affect pitching performance, which might include things like: “command” (i.e., the ability to locate one’s pitches), “stuff” (i.e., the overall “nastiness” of a pitcher’s repertoire), and more tactical aspects such as effective pitch sequencing (i.e., a metric of pitching strategy; the pitching analog of “approach” in hitting).

Major Takeaways:

• The ability to generate elite ball speed is almost certainly a prerequisite for pitching success, and should be a focus — but not the only focus — in training.

• The ability to deliver an entire repertoire of pitches from a consistent release point seems to be another important factor for pitching success, and is probably a major factor in “deception”, most likely because it limits the perceptual cues a hitter can use to make a decision about what pitch is about to be thrown.

• Exactly how elite pitchers generate a consistent release point is less clear, but we introduced some preliminary thoughts in our webinar. We will be revealing more of the work we have been doing in this space in future blog posts.

And, as always, these are just my thoughts. There is plenty of room for discussion here. Questions, comments, concerns, etc. are most welcome!

If you like what you’ve read, and you’d like to receive updates about future instalments of the Baseball Performance Research Review, be sure to sign up for our newsletter using the form below.