RACE ANALYSIS OF THE 400M FREESTYLE AT THE 1996 ATLANTA PARALYMPIC GAMES

Abstract

INTRODUCTION: To achieve fair competition disability sport is organized according to a functional classification system. In swimming, athletes from a variety of impairment groups are placed together in one of 10 classes according to scores on muscle testing, coordination, range of motion or level of amputation. As might be expected, a great deal of discussion has taken place concerning the validity and credibility of this system. Swimming, as one of the most popular Paralympic sports, has been the subject of performance analyses using end race result (ERR) as the basis for discriminating between the classes. The distinction between the classes on other race and stroking variables, however, has had limited studied. The purpose of this investigation, therefore, was threefold: 1) to conduct a race analysis of the 400m freestyle, 2) to determine which race variables contribute to ERR and how these may differ among the four (most functional) classes, and 3) to determine the implications of these findings for the classification system. METHODS: Performances of the 400m freestyle were recorded for all men (N=51) and women (N=34) using static video surveillance cameras placed perpendicular to the swimming direction at 7.5m, 10m, 25m and 42.5m from the start. Camera data was fed to a video recorder via a central control panel and recordings (30fps) were embedded with a time code from the official timing system. The following variables were measured: clean swimming speed (CSS) per 25m segment, stroke rate and length (SR, SL) per lap, all turn times (TT = 7.5m in + 7.5m out), start and finish times (ST = first 10m; FT = final 7.5m). ANOVA, ANCOVA and Spearman correlations were calculated. RESULTS AND CONCLUSIONS: No distinction was seen between the men in classes S10 and S9 for any of the race or stroking variables. The difference in ERR between S8 and S7 was completely accounted for by a difference in TT. In women little difference was found between classes S8 and S7 in the race and stroking variables. Class S10 had a significantly faster ERR, but similar CSS. The difference in ERR was accounted for by faster TT, FT and TT. Almost no differences were seen for SR or SL in either men or women. Correlations confirmed that with all classes combined, TT was the most highly determinant variable for ERR (men = 0.95, women = 0.97), while FT closely followed, and the start being least important (men = 0.71, women = 0.77). SR correlated significantly with CSS (men = 0.63, women = 0.57). Due to the small N and differences in findings among men and women any systematic unfairness in classification could not be identified. Acknowledgments: Dr. DJ Smith, Human Performance Lab, Univ of Calgary, Canada; Alberta Paraplegic Foundation, Canada; Rick Hansen Centre, Alberta Canada; Prof. Dr. U. Persyn, K.U.Leuven, Belgium