Lachlan P James1, Sam Robertson2, G Gregory Haff3, Emma M Beckman4, Vincent G Kelly5. 1. School of Human Movement and Nutrition Sciences, University of Queensland, Australia. Electronic address: l.james1@uq.edu.au. 2. Institute of Sport, Exercise & Active Living (ISEAL), Victoria University, Australia. 3. Centre for Sport and Exercise Science Research, Edith Cowan University, Australia. 4. School of Human Movement and Nutrition Sciences, University of Queensland, Australia. 5. School of Human Movement and Nutrition Sciences, University of Queensland, Australia; Brisbane Broncos Rugby League Football Club, Australia.
Abstract
OBJECTIVES: To determine those performance indicators that have the greatest influence on classifying outcome at the elite level of mixed martial arts (MMA). A secondary objective was to establish the efficacy of decision tree analysis in explaining the characteristics of victory when compared to alternate statistical methods. DESIGN: Cross-sectional observational. METHODS: Eleven raw performance indicators from male Ultimate Fighting Championship bouts (n=234) from July 2014 to December 2014 were screened for analysis. Each raw performance indicator was also converted to a rate-dependent measure to be scaled to fight duration. Further, three additional performance indicators were calculated from the dataset and included in the analysis. Cohen's d effect sizes were employed to determine the magnitude of the differences between Wins and Losses, while decision tree (chi-square automatic interaction detector (CHAID)) and discriminant function analyses (DFA) were used to classify outcome (Win and Loss). RESULTS: Effect size comparisons revealed differences between Wins and Losses across a number of performance indicators. Decision tree (raw: 71.8%; rate-scaled: 76.3%) and DFA (raw: 71.4%; rate-scaled 71.2%) achieved similar classification accuracies. Grappling and accuracy performance indicators were the most influential in explaining outcome. The decision tree models also revealed multiple combinations of performance indicators leading to victory. CONCLUSIONS: The decision tree analyses suggest that grappling activity and technique accuracy are of particular importance in achieving victory in elite-level MMA competition. The DFA results supported the importance of these performance indicators. Decision tree induction represents an intuitive and slightly more accurate approach to explaining bout outcome in this sport when compared to DFA.
OBJECTIVES: To determine those performance indicators that have the greatest influence on classifying outcome at the elite level of mixed martial arts (MMA). A secondary objective was to establish the efficacy of decision tree analysis in explaining the characteristics of victory when compared to alternate statistical methods. DESIGN: Cross-sectional observational. METHODS: Eleven raw performance indicators from male Ultimate Fighting Championship bouts (n=234) from July 2014 to December 2014 were screened for analysis. Each raw performance indicator was also converted to a rate-dependent measure to be scaled to fight duration. Further, three additional performance indicators were calculated from the dataset and included in the analysis. Cohen's d effect sizes were employed to determine the magnitude of the differences between Wins and Losses, while decision tree (chi-square automatic interaction detector (CHAID)) and discriminant function analyses (DFA) were used to classify outcome (Win and Loss). RESULTS: Effect size comparisons revealed differences between Wins and Losses across a number of performance indicators. Decision tree (raw: 71.8%; rate-scaled: 76.3%) and DFA (raw: 71.4%; rate-scaled 71.2%) achieved similar classification accuracies. Grappling and accuracy performance indicators were the most influential in explaining outcome. The decision tree models also revealed multiple combinations of performance indicators leading to victory. CONCLUSIONS: The decision tree analyses suggest that grappling activity and technique accuracy are of particular importance in achieving victory in elite-level MMA competition. The DFA results supported the importance of these performance indicators. Decision tree induction represents an intuitive and slightly more accurate approach to explaining bout outcome in this sport when compared to DFA.
Authors: Sherrilyn Walters; Ben Hoffman; William MacAskill; Michael A Johnson; Graham R Sharpe; Dean E Mills Journal: Eur J Appl Physiol Date: 2021-08-25 Impact factor: 3.078
Authors: João C A Bueno; Heloiana Faro; Seth Lenetsky; Aleksandro F Gonçalves; Stefane B C D Dias; André L B Ribeiro; Bruno V C da Silva; Carlos A Cardoso Filho; Bruna M de Vasconcelos; Júlio C Serrão; Alexandro Andrade; Tácito P Souza-Junior; João G Claudino Journal: Sports (Basel) Date: 2022-05-24
Authors: Fernando C Loio Pinto; Henrique Neiva; Célia Nunes; Mário C Marques; António C Sousa; Daniel A Marinho; Luís Branquinho; Ricardo Ferraz Journal: Int J Environ Res Public Health Date: 2020-09-28 Impact factor: 3.390