In addressing the practical difficulty of surveying elite swimmers` race patterns within an individual event from multiple swimming championships across various comparable criteria at the same time, a system of cluster analysis is described that uses standard statistical algorithms to arrange elite swimmers according to similarity in their race patterns. The outputs are displayed graphically using heat maps and tree-like dendrograms, transmitting the information to our human brains by means of a high-bandwidth channel than a "number-reading" channel. Since the end product is a visual representation of elite swimmers` race patterns within an individual event from multiple swimming championships across various comparable criteria at the same time, it would meet with a great deal of enthusiasm if coaches could develop an integrated knowledge into elite swimmers` race patterns being studied from such images. As an illustration, a combined dataset relating to the finalists` race patterns for the men`s 200m freestyle finals from 2000 European Championships in Helsinki, Sydney 2000 Summer Olympics, 2002 European Championships in Berlin, and 2004 European Championships in Madrid are used to represent the hidden group structures in connection with their race results at any one time.
© Copyright 2007 International Journal of Sports Science and Engineering. World Academic Press. Alle Rechte vorbehalten.