Many procedures have been developed to analyse games using simple hand notation or more complex computerized analysis. In order to construct a valid notational analysis system, it is necessary to determine the intra- and inter-observer reliability of the whole notational procedure. Therefore, the purpose of this study was to validate a notational analysis system for soccer. Two Finnish national level U-17 junior teams participated in this study. They played one official match with FIFA rules. The match was video recorded and analysed afterwards by three trained observers. They coded the whole match twice using SAGE Game ManagerTM for Soccer software. All the observers were experienced players and coaches; two observers in soccer and one in volleyball. All the match events were defined in written form and entered for the computer using playback with video. The output of the analysis gave frequency distributions, percentages and mean values of the following variables: running playing time (RPT), location of the player (LP), success of the event (SE), player identification (PI), event identification (EI) and total amount of all events (TAE). The intra- and inter-observer reliability of all variables were calculated using Pearson´s correlation coefficients. The total match included about 2000 events. The inter-observer agreement level varied in TAE from 89 to 94%. The inter- and intra correlation coefficients ranges were in RPT; r = 0.99-1.00, in LP; r = 0.95-.99, in SE; r = 0.74-.94, in PI; r = 0.89-.98 and in EI; r = 0.91-.99. The overall results indicated that the analysis system used was reliable and valid for evaluating running playing time (RPT), player´s location on the ground (LP), identification of the player (PI), event identification (EI) and the total amount of all events (TAE). But it was less reliable when success of the events (SE) was concerned. In conclusion, this computer based notation system offers a reliable and valid method to assess game performance in soccer.
© Copyright 2001 3rd International Symposium on Computer Science in Sport - 5th World Congress of Performance Analysis of Sport. Alle Rechte vorbehalten.