Open AccessA Discriminant Model For Skill Oriented Prediction of Female Cricketers Depending Upon Selected Performance Parameters
Author(s) -
Sapna Mandoli,
Deepak Sharma,
H. C. Joshi
Publication year - 2021
Publication title -
teorìâ ta metodika fìzičnogo vihovannâ
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.271
H-Index - 5
eISSN - 1993-7997
pISSN - 1993-7989
DOI - 10.17309/tmfv.2021.4.01
Subject(s) - discriminant function analysis , linear discriminant analysis , anthropometry , statistics , mathematics , multivariate statistics , discriminant , cricket , medicine , artificial intelligence , computer science , biology , ecology
Research Purpose. The study aimed to develop a discriminant model for cricketers on the basis of physiological & anthropometric variables.
Material and Methods. The study included sixty female seniors BCCI board players representing five different states with mean age 23.4 ± 2.03, mean height 152.1 ± 3.44, and mean weight 52.4 ± 4.21. A multivariate technique was used to predict the cricket performance by classifying the players into batsmen and pace bowlers on the basis of selected physiological & anthropometrical variables.
Results. All the assumptions were positively full-filled (Shapiro-Wilk test p > 0.05 and F = 8.121, p = 0.264 for Box’sM test) for applying discriminant analysis to develop the model. A total of eleven performance variables were initially selected for the study and after applying the stepwise statistical technique of discriminant analysis, the model selected certain variables, namely Muscle Mass (1.311), Fat (-0.182) & Shoulder Diameter (0.292) and showed its effectiveness as the Eigenvalue for the fit model was 0.848.
Conclusion. A discriminant function F1 = -29.531 + (1.311 × Muscle Mass) + (-0.182 × Fat) + (0.292 × Shoulder Diameter) was developed. The percentage of total variation explained by the model was 71.9%. A total of 93.2% of the observations were correctly classified using the proposed discriminant model.