Validation of a tactical analysis methodology for the study of pick and roll in basketball

• Published in: Revista de psicología del deporte 2013-01, Vol.22 (1), p.277
• Main Authors: Vaquera, Alejandro, Cubillo, Rodrigo, J. Vicente García-Tormo, Morante, Juan C
• Format: Article
• Language: eng ; spa
• Subjects: Basketball
• ISSN: 1132-239X; 1988-5636

This study examines the design and validation of a tactical analysis methodology proposal for the study of pick and roll in basketball, applying the principles of observational methodology and the recording of actions in their natural context. We have developed three phases: Phase 1) Identification and selection of analysis variables, in which behavioral variables have been identified and cataloged (pick and roll), contextual analysis variables (court area, time of possession and period of play) and evaluative analysis variables (result of the action); Phase 2) Ad hoc creation of recording instruments, in which a computer tool was adapted (VA-Sports Software: Basketball) to address the particularities and specific needs of the study of the pick and roll situations.; and Phase 3) Quality control of the data, in which selection criteria were set for the sample, the training process of the observers was carried out, and the level of internal agreement and intraobserver concordance were verified in order to assess the reliability of the methodological approach. A total of 192 pick and roll situations corresponding to 6 games of the Spanish ACB League were analyzed in this study. The results suggest a high level of reliability of the methodology designed; as a whole it earned a high average Kappa value (.994), leading to the conclusion that the proposed analysis variables are precise in their definition and categorization in order to study pick and roll situations. The analysis variables; time of possession, period of play and results of the action showed a maximum level of agreement (Kappa value 1.000), while the analysis variables; pick and roll and court area also show a high degree of objectivity (Kappa values of .990 and .981 respectively).