Influence of Pitcher Fielding Plays and Pick off Attempts on Overall In-Game Workload during an NCAA Division 1 Season
2Central Michigan University, Mt. Pleasant, Michigan
3Driveline Baseball, Kent Washington
During the course of a game or practice, pitchers are asked to not only pitch, but occasionally control the running game via pick off throws to bases, or collect outs on their own via pitcher fielding plays. As a regular fan at home, one thing you can notice is that during either of these “non-pitching throws”, that little ticker in the corner of your screen does not go up – these throws don’t add to the pitch count. The purpose of this study was two-fold – to examine the differences in throwing mechanics and elbow stress during PFPS, POs, and flat ground throws, and to see how to account for these throws during games throughout a season.
Study 1: Relationship between kinematics and elbow torque during PFPs, POs and flat ground throws
In what I hope is the first of many data collection endeavors, we were able to collect data using radar guns and the MOTUSthrow device, at two different facilities – Driveline Baseball in Kent, Washington, and the Baseball Development Group in Toronto, Ontario. Both facilities followed the same protocol and had access to similar pitching populations.
Both facilities recruited 5 pitchers who were currently working out in their facilities. Height, Age, and Weight were recorded, and each pitcher’s account was created in the MOTUSthrow app. On average, the pitchers were 25.2 +/- 2.15 years old, 73.5 +/- 1.96 “ tall, and 202.0 +/- 11.29 lbs.
Pitchers completed 15 total throws – a maximum intensity flat ground throw, a simulated pickoff throw to first base, and a simulated pitcher fielding play with a throw to first base. All MOTUSthrow metrics were recorded for each throw, as well as the velocity of the throw via radar gun.
For data analysis, descriptive statistics were calculated for each of the dependent measures of throw velocity (mph), torque (Nm), arm slot (°), arm speed (RPM), and external rotation (°), and sorted by the independent variable of throw type (flat ground, pickoff, and pitcher fielding play). A one-way ANOVA was then performed for each dependent variable, with the independent factor being throw type. A Tukey’s HSD post hoc test was used to determine significance between throw types, with alpha set to 0.05.
The flat ground throw was significantly faster than the pick off, and the pfp throw, with an average velocity of 80.7 +/- 3.9 mph. This was compared to the pickoff throw (74.1 +/- 6.0 mph), and the fielding play (72.1 +/- 8.3 mph). Tukey’s HSD did not reveal any significant differences between the PFP throw and the pickoff throw.
However, there was no statistically significant difference between throw types for elbow torque, despite the flat ground throw having a higher average elbow torque than the pickoff and pfp throws (58.5 +/- 9.8 Nm, compared to 54.4 +/- 9.2 Nm, and 53.3 +/- 8.6 Nm, respectively).
There was no significant difference between throw types for external rotation, however, both metrics of arm speed and arm slot were significant. In the case of arm slot, the pfp was significantly higher (38.2 +/- 16.5°) than the pickoff (26.5 +/- 16.5°), and the pfp (27.0 +/- 17.9°). For arm speed, the flat ground was faster (893.1 +/- 79.8 RPM), compared to the pickoff (822.8 +/- 95.0 RPM), and pfp (822.82 +/- 131.1 RPM). In all cases, Tukey’s HSD revealed significance between the flatground throw and the pfp/pickoff. There was no significance between the pfp and pickoff throws.
So, while there was no statistical significance for elbow torque between the three throw types, changes in the kinematics of those throw types (a lower arm slot) leads to an increased level of stress relative to the velocity that the ball is thrown. With this in mind, pick off throws, and pitcher fielding plays should be accounted for when looking at throwing within a game.
Now, how often do these throws occur, and what would their impact be on overall workload within a season? To answer that question, Jason Castleman Williams, an athletic therapist at Central Michigan University (Fire up, Chips!), and co-author to this study, leant us his keen eye and abacus over the span of the Chippewa’s incredible 2019 season.
Study 2: Evaluating workload using Fatigue Units during an NCAA baseball season
Jason and his staff at CMU recorded the total in game pitch counts for all of their pitchers during the 2019 season – one where the Chips went to win the MAC conference championship, played in the NCAA tournament, and an incredible 46-12 overall record. All of these extra games do however mean, that pitchers are going to throw more.
For each pitching performance, Jason marked the number of pitches, the time that the pitcher started their appearance and ended their appearance, and their average pitch velocity during the season. This allowed us to calculate Fatigue Units for the entire season. Jason’s team also calculated the total number of pitching fielding plays, and number of pick off throws as well for his team’s pitchers. Given what we had learned in study 1 of this experiment, the fatigue units were calculated with and without the total throws involved in PFPs and pickoffs.
When you’re accounting for the additional throws, the worst case in this situation from this season was an increase in seasonal workload of 3.1% (for Pat Leatherman). In general, the average workload increased by 1.4% when accounting for the additional throws for all of the pitchers on the team. While this doesn’t seem like a lot, it could still be the difference between a spike in acute to chronic workload of greater than 1.3, or the level that is associated with increased risk of injury.
With that in mind, it’s also important to consider bullpen throws – a recent study by Dowling et al., (2019) that was presented at the ISB conference in Calgary, showed there was no significant difference in elbow torque between bullpen and in game throws.
However, that’s often much more difficult to track for those of us not at the game, or in the bullpen. What this does shed light on, is the fact pitching workloads assumed from in-game data, and only from pitches thrown to home plate, grossly underestimate workloads.
This post was written by BDG Research Director, Dr. Mike Sonne. All data from Study 1 can be found here.