A machine learning approach for the classification of sports based on a coaches' perspective of environmental, individual and task requirements: A sports profile analysis.

Well-designed talent programmes in sports with a focus on talent identification, orientation, development, and transfer support the engagement of young individuals and the pursuit of elite performance. To facilitate these processes, an analysis of task, environmental and individual characteristics per sport is much needed. The aims of this study were to 1) analyse whether unique profiles per sport could be established by generic characteristics and 2) to discuss similarities and differences for the potential application in talent development and transfer. By means of a validated survey, 1247 coaches from 34 sports ranked 18 characteristics on importance to their sports (0 = not important - 10 = very important). To discriminate the responses per sport a Discriminant Analysis (DA) was carried out. To refine the DA-classification, Uniform Manifold Approximation and Projection (UMAP) with CatBoost classifier was performed. To test the performance of the CatBoost classifier-algorithm, a confusion-matrix was generated. The cross-validated DA showed that 70.2% of the coaches were correctly classified to their sport. The UMAP/CatBoost technique revealed 75.1% accuracy with correctly predicted responses per sport ranging from 18.2% (sailing) to 98.2% (soccer). With varying precision, the algorithm was able to differentiate sports by importance of its characteristics indicating similarities and differences per sport.

Similarities and differences between sports subserving systematic talent transfer and development: The case of paddle sports.

OBJECTIVES: The aim of this study was to investigate similarities and differences for 18 sports toward canoe/kayak in order to identify donorsport and/or multisports, based upon a systematic analysis of the task constraints per sport that are assumed to be either crucial or less important from the coaches' viewpoint. DESIGN: Descriptive survey analysis. METHODS: 891 certified coaches from 19 sports valued (0-10; not important at all-very important) 15 characteristics by a questionnaire (Flemish Sports Compass) within their sport. Unique sport-profiles (discriminant analysis - DA) were constructed for 19 sports based on these characteristics. Similarities and differences between canoe/kayak and the other 18 sports were analyzed by means of MANOVAs on anthropometric, physical and motor coordination characteristics. RESULTS: Cross validated DA (rcan=0.660, Wilks' Lambda=0.564, p<0.001) showed that 72.1% of the canoe/kayak coaches were correctly assigned to their sport. For canoe/kayak seven characteristics were valued crucial; dynamic balance (8.51±1.69), core stability (8.45±2.27), pulling power (8.12±1.68), speed (7.54±2.07), endurance (7.27±2.03), stature (6.43±1.41) and rhythm (6.01±3.01). Least important characteristics were: flexibility (6.16±1.75), agility (4.27±3.10), catching (3.90±3.22), climbing (2.45±3.05), jumping (1.81±2.11), throwing (1.60±2.24), hitting (.94±1.77) and kicking (.61±1.04). CONCLUSIONS: This novel approach to determine important characteristics per sport makes identifying similarities and differences between sports possible. Similarities might enlarge talent-pools for possible talent transfers. Differences can help identify sports based on complementary characteristics for the construction of broad motor development programs. From this viewpoint gymnastics can serve as potential donorsport (similarities) for canoe/kayak, while handball and tennis can subserve broad development for young canoe/kayak athletes.

Accuracy of maturity prediction equations in individual elite male football players.

BACKGROUND: Equations predicting age at peak height velocity (APHV) are often used to assess somatic maturity and to adjust training load accordingly. However, information on the intra-individual accuracy of APHV in youth athletes is not available. AIM: The purpose of this study is to assess the accuracy of predication equations for the estimation of APHV in individual youth male football players. SUBJECTS AND METHODS: Body dimensions were measured at least every three months in 17 elite youth male football players (11.9 ± 0.8 years at baseline) from the 2008-2009 through the 2011-2012 seasons. APHV was predicted at each observation with four suggested equations. Predicted APHV was compared to the player's observed APHV using one-sample-t-tests and equivalence-tests. Longitudinal stability was assessed by comparing the linear coefficient of the deviation to zero. RESULTS: Predicted APHV was equivalent to the observed APHV in none of the players. A difference with a large effect size (Cohen's d > 0.8) was noted in 87% of the predictions. Moreover, predictions were not stable over time in 71% of the cases. CONCLUSIONS: None of the evaluated prediction equations is accurate for estimating APHV in individual players nor are predictions stable over time, which limits their utility for adjusting training programmes.