Author(s): ŠARABON, N., KOZINC, Ž., SMAJLA, D., Institution: UNIVERSITY OF PRIMORSKA, Country: SLOVENIA, Abstract-ID: 2414

INTRODUCTION:
Vertical jumps are among most represented sport performance tests. Ground reaction force derived metrics (e.g., force, power, and force impulse) are commonly analysed. Prior research has investigated how characteristics of force-time profiles (e.g., timing of peak force [1], number of force peaks [2]) influence performance. Although a connection between the characteristics of the force-time curve and jump effectiveness is indicated, the challenge of identifying the optimal force-time profile necessitates further investigation. The aim of this paper is to address this knowledge gap by applying clustering approach based on ground reaction force variables of squat jump (SJ) and counter-movement jump (CMJ) force-time signals.
METHODS:
High-level athletes (basketball, soccer, tennis, long-distance running, dance, and martial arts; n=310)) participated. Three repetitions of SJ and CMJ were performed on Kistler force plate. K-means clustering method, that aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest mean, was applied to SJ and CMJ parameters. K-means clustering method was used to create 10 clusters, of which the 3 clusters with the highest representation were taken for further processing. Relative distances between cluster centres were calculated for each squat and countermovement jump parameter.
RESULTS:
809 SJs were taken into analysis. Three clusters with highest representation of jumps were: cluster 1 with 375 jumps, cluster 2 with 241 jumps and cluster 3 with 177 jumps. The four SJ parameters with the highest relative distance between cluster centres among the three clusters were: force impulse between maximal force and take off, start interval relative power, force impulse in the 2nd half of push off and start interval velocity.
841 CMJs were taken into analysis. Three clusters with highest representation of jumps were: cluster 1 with 290 jumps, cluster 2 with 277 jumps and cluster 3 with 249 jumps. The four CMJ parameters with the highest relative distance between cluster centres among the three clusters were: force impulse in the 2nd half of push off, push off force impulse, force impulse in the 1st half of push off, positive force impulse.

CONCLUSION:
Results of this study demonstrate the grouping of vertical jump actions into force-time curve types that are distinct from one as reflected in more or less standard outcome metrics. This study is an entry step into quality of the SJ and CMJ ground reaction force curves with the ambition to upgrade applied value of this common athletic performance tests.
1. McHugh MP, Hickok M, Cohen JA, Virgile A, Connolly DAJ. Is there a biomechanically efficient vertical ground reaction force profile for countermovement jumps? Transl Sport Med. 2021;4(1):138–46.
2. Peng H Te, Song CY, Chen ZR, Wang IL, Gu CY, Wang LI. Differences between Bimodal and Unimodal Force-time Curves during Countermovement Jump. Int J Sports Med. 2019;40(10):663–9.