Analyzing the aiming process in biathlon shooting using self-organizing maps

The aiming strategy in biathlon shooting is a crucial factor for success. Because of the preceding high exertions of the athletes a well controlled motion of the barrel just before shooting is essential (Zatsiorsky and Aktov, 1990). Methods are required for analyzing the stability of the aiming process with a special focus on exertion. The aim of this study was to investigate the applicability of special self-organizing maps to identify and compare patterns in the aiming motion in standing shooting. Methods A video based system (Baca and Kornfeind, 2006) was used to track the motion of the muzzle of the barrel in two dimensions (left-right, up-down) automatically. Six parameters were calculated describing the motion in ten time intervals of 0.2 s length before the shot. Four athletes (I, II, III, IV) (I, III, IV: Austrian national “B” team, II: “C” team) participated in the study. Each athlete performed four series of five shots before and after exertion making 160 shots altogether. Based on these data a special self organizing map (Perl et al., 2006) consisting of 400 neurons was trained and data sets were generated on the neurons. The attribute values of those data sets represent the six components describing a motion of a muzzle in a 0.2 s time interval. Similar neurons were combined to clusters. The ten successive data-sets describing each shot were then mapped to the corresponding neurons of the net. The sequence of the related clusters in the respective succession was then used as 1-dimensional representation (a pattern) of the complex aiming motion. Results Regarding intra-individual stability, some peculiarities were found. The number of a shot within a series of five shots clearly influenced the pattern observed. This was particularly the case after exertion. Moreover, in this condition less stable shot types were found. Subjects were able to maintain their pattern before the exertion to a different degree. Subject II, who showed the largest deviations, scored worst. Although inter-individual variability was difficult to assess, some similarities (e.g. in timing) could be identified. Discussion The method is promising to analyze inter- and inter-individual similarities and differences. One shortfall might be that only a restricted set of parameter values originating from a 2-dimensional recording of the muzzle has been considered. Time series data describing the 3-dimensional motion of the barrel would probably increase the explanatory power. In addition, alternatives not based of a fixed time interval for analysis (2 seconds before the shot) are conceivable. References Baca A, Kornfeind P (2007). IEEE Perv. Comp., 5, 70-76. Perl J, Memmert D, Bischof J, Gerharz C. (2006). Int J Comp Sci Sport, 5, 33–37. Zatsiorsky VM, Aktov AV (1990). J Biomech., 23 (Suppl. 1), 35-41.
© Copyright 2009 14th annual Congress of the European College of Sport Science, Oslo/Norway, June 24-27, 2009, Book of Abstracts. Published by The Norwegian School of Sport Sciences. All rights reserved.

Subjects: elite sport biathlon shooting technique analysis biomechanics
Notations: endurance sports training science technical and natural sciences
Published in: 14th annual Congress of the European College of Sport Science, Oslo/Norway, June 24-27, 2009, Book of Abstracts
Editors: S. Loland, K. Boe, K. Fasting, J. Hallen, Y. Ommundsen, G. Roberts, E. Tsolakidis
Published: Oslo The Norwegian School of Sport Sciences 2009
Pages: 126
Document types: congress proceedings
Language: English
Level: advanced