Available at: https://digitalcommons.calpoly.edu/theses/3006
Date of Award
6-2025
Degree Name
MS in Biomedical Engineering
Department/Program
Biomedical Engineering
College
College of Engineering
Advisor
Scott Hazelwood
Advisor Department
Biomedical Engineering
Advisor College
College of Engineering
Abstract
Pitch tipping refers to unintentional variations in a pitcher’s motion that reveal the pitch type, often due to inconsistent arm speed when throwing off-speed pitches like changeups. This issue is common in youth players with underdeveloped mechanics and makes it easier for batters to predict pitches. Wearable sensors, such as Inertial Measurement Units (IMUs), provide a non-invasive way to quantify these subtle motion discrepancies. This study analyzed linear acceleration data from a glove arm IMU for 14 pitchers, evenly split between high school-aged (13-15 years) and college-aged (18-23 years) groups. The aim was to examine the linear accelerations of the tested participants recorded by the IMU to analyze differences in arm motion for each pitch type: fastballs, changeups, and other off-speed pitches. It was hypothesized that there will be differences between all pitch types when examining the associated pitch “pulse”, both when combining the entire dataset and for individual analysis of each participant. Further, it was hypothesized that variability, portrayed by both coefficient of variation and standard deviation, will be significantly greater for high school pitchers compared to college-age pitchers since high school pitchers have less-developed pitching mechanics.
Each pitcher completed a bullpen session, identifying pitch types before each throw. Data were processed and analyzed with a custom MATLAB algorithm, which identified two distinct acceleration peaks during the throw. Key parameters analyzed included peak intervals, area under the curve (AUC), and peak magnitudes. While no differences were observed in the dataset of all pitchers combined due to high variability in the measured parameters, results showed that discrepancies in at least one parameter were often observed between pitch types when individual pitchers were examined, particularly between fastballs and changeups. In addition, one participant exhibited differences across all measured parameters between pitch types. Statistically significant differences in coefficient of variation were found between high school and college groups, especially regarding peak magnitudes. The participant-specific results indicate that motion discrepancies were detected between pitch types for these respective pitchers, further suggesting that these pitch types were being “tipped”. Additionally, analysis comparing high school and collegiate age groups suggested greater motion inconsistency among younger pitchers, aligning with the hypothesis that less-developed mechanics lead to more pronounced tipping. Limitations included inconsistent pitch counts and skill levels across age groups, facilitating the need for more standardized data collection in future studies.