Background Artificial neural networks (ANN) are proving to be a useful tool to assist professionals in multiple fields of study. However, the use of ANNs to match sport initiation (SI) standards and to assist during the selection and mentoring of young sports talent has not yet been tested. Objective To use artificial multilayer neural networks (MLPs) to perform a combination of the morphological and neuromuscular patterns of SI youth with those of young athletes. Methods 75 young men (13.3 ± 1.65-years), 87% of whom were athletes from different sports (volleyball, rowing, soccer, tennis, Brazilian-jiu-jitsu (BJJ), swimming) and 13% were SI practitioners were included. Their morphology was verified by anthropometry and dual-energy x-ray absorptiometry. Neuromuscular performance was verified by neuromotor tests (handgrip, vertical jump, countermovement, and medicine-ball throw). MLPs were programmed to verify the percentage of similarity between the morphological and neuromuscular patterns of youngsters in SI with those of young athletes of different sports. Results SI indicated similarity with the morphological patterns of 90% with tennis, 87% with soccer, 80% with swimming and 79% with BJJ. SI indicated similarity with neuromuscular patterns of 87% with soccer, 81% with swimming and 75% with BJJ. When combining the morphological and neuromuscular patterns SI showed similarity of 88% with soccer, 79% with swimming, 77% with BJJ and 70% with tennis. For rowing, there were no significant similarities. Conclusion It was possible to conclude that using MLPs is a strategy that helps direct young people from SI to a specific sport.
© Copyright 2023 Sport Sciences for Health. Springer. All rights reserved.