Adaptações no tamanho do campo de futebol para crianças e adolescentes a partir do desempenho físico e variáveis antropométricas: utilização da ciência de dados
Arquivos
Data
2024-08-09
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal de Viçosa
Resumo
A prática de atividade física é essencial para o desenvolvimento fisiológico, psíquico e social de crianças e adolescentes, incentivar esse hábito aumenta a probabilidade de uma vida adulta saudável. Praticar futebol é uma escolha comum nessa fase, mas adaptar o esporte para os jovens favorece a prática saudável do esporte. Este estudo propõe o uso da aprendizagem de máquina, com o objetivo de propor um método de redimensionar campos de futebol nas categorias de base conforme a aptidão física e variáveis antropométricas dos participantes. Participaram do estudo 268 indivíduos do sexo masculino, com 6 a 19 anos de idade, matriculados em escolinhas de futebol nas cidades de Florestal, Pará de Minas e Itaúna, no estado de Minas Gerais. O projeto foi aprovado pelo Comitê de Ética em Pesquisa da UFV. Os procedimentos envolveram testes antropométricos (estatura, massa corporal e envergadura) e de aptidão física (capacidade cardiorrespiratória, força de membros inferiores e velocidade), propostos pelo PROESP-BR. Os dados foram tabulados em Excel® e analisados no software R® de linguagem de programação, onde foram tratados e a partir do algoritmo k-means os participantes foram agrupados de acordo com suas similaridades. Comparando com os padrões da OMS, observou-se que uma normalidade dos dados com relação a massa corporal dos jovens, já a estatura, apesar de os indivíduos se encaixarem nos padrões de normalidade, parece haver uma subestimação dos padrões de crescimento, sugerindo a necessidade de tabelas baseadas na população nacional. Nos testes de aptidão física os resultados foram comparados com a tabela de expectativa de desempenho fornecida pelo PROESP-BR, onde foi identificado para o teste de capacidade cardiorrespiratória um desempenho “bom” (10, 11, 12, 13, 14, 15 e 16 anos de idade), desempenho “razoável” (7, 9 e 17 anos de idade) e “fraco” (6 e 8 anos de idade). Para o teste de velocidade todos os grupos por idade foram classificados com um desempenho “fraco” e no teste de força de membros inferiores os indivíduos foram classificados com um desempenho “muito bom” (16 anos de idade), desempenho “bom” (6, 7, 9 e 15 anos de idade), e “razoável” (8, 10, 11, 12, 13, 14 e 17 anos de idade). Ao comparar os resultados dos testes de aptidão física com a literatura, acredita-se que é possível que os testes realizados em campos de futebol, podem ter resultados subestimados em comparação com quadras poliesportivas, sugerindo a importância de futuras comparações. De acordo com os testes estatísticos desenvolvidos o grupo estudado pôde ser dividido em apenas 2 grupos, com características físicas e antropométricas significativamente diferentes entre sí. O algoritmo de aprendizado de máquina desenvolvido mostrou-se eficaz para agrupar indivíduos e identificar outliers, com potencial aplicação em contextos de alto rendimento e saúde. Os achados sugerem que apenas duas medidas de campo podem atender às necessidades das categorias de base, com uma transição acontecendo entre 11 e 12 anos. Estudos futuros devem incluir dados de adultos para validar e ajustar as dimensões propostas para campos de futebol. Palavras-chave: Futebol. Aptidão Física. Ciência de Dados.
Physical activity is essential for the physiological, psychological, and social development of children and adolescents. Encouraging this habit increases the likelihood of a healthy adult life. Playing soccer is a common choice at this stage, but adapting the sport for young people promotes healthy sports practice. This study proposes the use of machine learning to propose a method for resizing soccer fields in the youth categories according to the physical fitness and anthropometric variables of the participants. The study included 268 male individuals, aged 6 to 19, enrolled in soccer schools in the cities of Florestal, Pará de Minas, and Itaúna, in the state of Minas Gerais. The project was approved by the Research Ethics Committee of UFV. The procedures involved anthropometric tests (height, body mass, and wingspan) and physical fitness tests (cardiorespiratory capacity, lower limb strength, and speed), proposed by PROESP-BR. The data were tabulated in Excel® and analyzed in the R® programming language software, where they were processed and, using the k-means algorithm, the participants were grouped according to their similarities. Comparing with the WHO standards, it was observed that the data were normal in relation to the body mass of the young people, while in relation to height, although the individuals fit the normal standards, there seems to be an underestimation of the growth patterns, suggesting the need for tables based on the national population. In the physical fitness tests, the results were compared with the performance expectation table provided by PROESP-BR, where a “good” performance (10, 11, 12, 13, 14, 15 and 16 years of age), “reasonable” performance (7, 9 and 17 years of age) and “poor” performance (6 and 8 years of age) were identified for the cardiorespiratory capacity test. For the speed test, all age groups were classified as having a “poor” performance, and in the lower limb strength test, individuals were classified as having a “very good” performance (16 years old), “good” performance (6, 7, 9 and 15 years old), and “reasonable” performance (8, 10, 11, 12, 13, 14 and 17 years old). When comparing the results of the physical fitness tests with the literature, it is believed that it is possible that the tests carried out on soccer fields may have underestimated results compared to multi-sports courts, suggesting the importance of future comparisons. According to the statistical tests developed, the studied group could be divided into only 2 groups, with significantly different physical and anthropometric characteristics. The machine learning algorithm developed proved to be effective in grouping individuals and identifying outliers, with potential application in high-performance and health contexts. The findings suggest that only two pitch sizes can meet the needs of youth teams, with a transition occurring between ages 11 and 12. Future studies should include data from adults to validate and adjust the proposed dimensions for football pitches. Keywords: Soccer. Physical Fitness. Data Science.
Physical activity is essential for the physiological, psychological, and social development of children and adolescents. Encouraging this habit increases the likelihood of a healthy adult life. Playing soccer is a common choice at this stage, but adapting the sport for young people promotes healthy sports practice. This study proposes the use of machine learning to propose a method for resizing soccer fields in the youth categories according to the physical fitness and anthropometric variables of the participants. The study included 268 male individuals, aged 6 to 19, enrolled in soccer schools in the cities of Florestal, Pará de Minas, and Itaúna, in the state of Minas Gerais. The project was approved by the Research Ethics Committee of UFV. The procedures involved anthropometric tests (height, body mass, and wingspan) and physical fitness tests (cardiorespiratory capacity, lower limb strength, and speed), proposed by PROESP-BR. The data were tabulated in Excel® and analyzed in the R® programming language software, where they were processed and, using the k-means algorithm, the participants were grouped according to their similarities. Comparing with the WHO standards, it was observed that the data were normal in relation to the body mass of the young people, while in relation to height, although the individuals fit the normal standards, there seems to be an underestimation of the growth patterns, suggesting the need for tables based on the national population. In the physical fitness tests, the results were compared with the performance expectation table provided by PROESP-BR, where a “good” performance (10, 11, 12, 13, 14, 15 and 16 years of age), “reasonable” performance (7, 9 and 17 years of age) and “poor” performance (6 and 8 years of age) were identified for the cardiorespiratory capacity test. For the speed test, all age groups were classified as having a “poor” performance, and in the lower limb strength test, individuals were classified as having a “very good” performance (16 years old), “good” performance (6, 7, 9 and 15 years old), and “reasonable” performance (8, 10, 11, 12, 13, 14 and 17 years old). When comparing the results of the physical fitness tests with the literature, it is believed that it is possible that the tests carried out on soccer fields may have underestimated results compared to multi-sports courts, suggesting the importance of future comparisons. According to the statistical tests developed, the studied group could be divided into only 2 groups, with significantly different physical and anthropometric characteristics. The machine learning algorithm developed proved to be effective in grouping individuals and identifying outliers, with potential application in high-performance and health contexts. The findings suggest that only two pitch sizes can meet the needs of youth teams, with a transition occurring between ages 11 and 12. Future studies should include data from adults to validate and adjust the proposed dimensions for football pitches. Keywords: Soccer. Physical Fitness. Data Science.
Descrição
Palavras-chave
Futebol - Métodos estatísticos, Aptidão física - Testes, Aprendizado do computador
Citação
ROCHA, Julio Cesar Piedade de Medeiros dos Santos. Adaptações no tamanho do campo de futebol para crianças e adolescentes a partir do desempenho físico e variáveis antropométricas: utilização da ciência de dados. 2024. 72 f. Dissertação (Mestrado em Educação Física) - Universidade Federal de Viçosa, Viçosa. 2024.