摘要：Authors deals with the problematics of group classification of athletics disciplines, which influence the sports performance in the men's decathlon. For the group identification, the indicators of the best world's performance in decathlon above the 8000 points according to the data from IAAF (1986-2019) were used. From the classification methods of clustering the hierarchical models as the Average linkage (Between & Within-group), Single Linkage - Nearest neighbor, Complete Linkage - Farthest neighbor, Centroid linkage, Median clustering, and Ward´s method were used. The substructure of the clusters is differentiated according to the relations of used methods. All seven clustering methods agreed in four groups of disciplines in 4 cluster [100 meters, 400 meters, Long jump, 110 meters hurdles, Pole vault] [High jump] [1500 meters]. Stability test with a substructure of decathlon of 4th. cluster is identified in 85.71% cases. Hierarchical models allow identifying groups of athletics disciplines that influence the sports performance in men's decathlon. Understanding the structure of sports performance contributes to the streamlining of the training process and determining the typology of top-class decathletes.
