Profiling the Physical Performance of Young Boxers with Unsupervised Machine Learning: A Cross-Sectional Study

Mexico City is the location with largest number of boxers in Mexico; fact, it first city country to open a Technological Baccalaureate Education and Sports Promotion pugilism orientation. This cross-sectional study aimed determine physical–functional profile applicants for admission baccalaureate sports. A total 227 young athletes (44F; 183M; 15.65 (1.79) years; 63.66 (14.98) kg; >3 years boxing experience) participated this study. Body mass (BM), maximal isometric handgrip (HG) strength, height countermovement jump (CMJ), velocity straight punches (PV), rear hand punch impact force (PIF) were measured. The profiled using unsupervised machine learning algorithms, probability superiority (ρ) was calculated as effect size differences. K-Medoids clustering resulted two sex-independent significantly different groups: Profile 1 (n = 118) 2 109). Except BM, statistically higher (p < 0.001) clear distinction terms on PIF (ρ 0.118), PIF-to-BM ratio 0.017), PIF-to-HG 0.079) PIF-to-BM+HG 0.008). In general, strength levels explained most data variation; therefore, reasonable recommend implementation tests at assessing applied gestures. identification these profiles might help differentiate training programs during sports specialization athletes.