The ability to score from placekicks discriminates winning from losing Rugby Union teams. We aimed to identify which biomechanical variables related to successful placekicking in professional Rugby Union players, and use self-organising maps (SOM) to determine whether meaningful sub-groups existed. Three professional placekickers performed 10 kicks outdoors. Placekicks were categorised into best, worst, and typical performances based on outcomes and coach and player perceptions. Seven 3D biomechanical variables consistently and meaningfully (moderate Cohen's effect size) discriminated best from worst placekicks in all players. The three-cluster solution from SOM on these seven variables highlighted differences between players rather than best, worst, and typical attempts. Within-clusters, however, the best and worst placekicks tended to be represented in separate map regions. The seven variables identified using standardised effect sizes can be useful for group-level coaching of placekicking skills in absence of individual data, and translated in an applied setting using verbal and visual cues to promote overall placekicking performance. However, players' idiosyncrasies formed the main SOM boundaries, indicating that optimising placekicking success would benefit from an individualised approach and numerous effective movement templates may exist.
Keywords: 3D motion analysis; artificial neural network; biomechanics; self-organising map; sport performance.