Data modelling reveals inter-individual variability of front crawl swimming

Abstract

ObjectivesIn accordance with dynamical systems theory, which assumes that motor behaviour emerges from interacting constraints (task, organismic, environmental), this study explored the functional role of inter-individual variability in inter-limb coordination. Design63 front crawl swimmers with a range of characteristics (gender, performance level, specialty) performed seven intermittent graded speed bouts of 25m in front crawl. MethodsEach bout was video-taped with a side-view camera from which speed, stroke rate, stroke length and index of arm coordination (IdC) were analysed for three cycles. Cluster analysis was used to classify the swimmers through speed and IdC values. ResultsCluster analysis and validation showed four profiles of IdC management expressing the swimmers’ characteristics as cluster 1: mainly national distance male swimmers, cluster 2: mainly international male sprinters, cluster 3: distinguished by female characteristics, and cluster 4: swimmers with the lowest level of performance. ConclusionsThese profiles generated different IdC-speed regression models, which (i) showed how the swimmers adapted their motor behaviour to overcome task constraints and (ii) supported the key idea that there is not a single ideal expert model to be imitated, but rather adapted behaviour emerging from individually encountered constraints.