Those of you familiar with my work know that quantifying and collecting data is a technique I find to be especially valuable. This last summer, I began experimenting with swing trackers from Diamond Kinetics. I devised an assessment technique designed to expose hitters swing flaws; if you have not read my first iteration in this series, I recommend reading https://gritperformance.home.blog/2019/07/30/data-driven-hitters-chapter-1/ first.
Assessing hitters was a prominent goal of mine. However, exporting the data, editing the file and then doing the calculations in R was a lengthy task. Being an undergraduate in Data Analytics, I felt that there must be a better way to approach this task. I felt that player development needed a program that could be easily scalable. A program that could perform assessments for a team or a program in minutes, not hours would let this technology benefit more athletes/coaches.
For this program, I utilized both the tidyverse and ggrepel packages. Both of these tools allowed me to perform calculations, visualize and more easily communicate the data.
Regardless of the datasets name, I chose to have dataset be the standard name for all incoming files. This could easily be changed by using R’s search function/replace all, but as I am only assessing one hitter at a time, I felt that merely having a common name in R would suffice. Next, the operator declares the number of swings included in the data set to allow for proper calculations of the averages and standard deviations.
The next step was writing functions that could accurately calculate and visualize the data. I chose to graph the average utilizing a horizontal tangent line and label values outside out the standard deviation using ggrepel. I felt that what this allowed for is hitters being able to see where they sit and how consistent they are.
Metrics displayed on the Y – Axis usually depict the responding variable; however, what exists on the X-axis changes. Sometimes I list the swing number showing trends by time. Other times, however, I list different variables on the X-Axis that are associated with superior performance. For metrics such as hand cast distance and trigger to impact, hitters vary and likely have an optimum range as opposed to a desired number.
Overall, the data science included in the program is nothing groundbreaking. There is no monumental discovery or statistical breakthrough. This program does allow for a quicker and broader implementation of Diamond Kinetics technology into a player development program. I would love to see this used with youth performance athletes throughout their minor careers. It would encourage hard work as hitters could see their progress before them, and coaching staffs could make more non-partial decisions as to who deserves playing time. If you have any more questions, please email me @grit.performance1@gmail.com
~Chris
Grit Performance 