A daily longitudinal analysis of reciprocal relationships between psychological states and performance in NHL players

This study aimed to explore the temporal relationships between psychological states and performance among players in the National Hockey League (NHL). We used an intensive longitudinal design, with 22 players reporting their psychological states daily between 7 February 2024, and 14 March 2024. Performance was measured using Game Score (GS), a composite metric based on weighted performance variables to provide an overall objective measure of individual performance, commonly used by teams in the NHL. Bayesian multilevel models were utilised to analyse the data, focusing on temporal effects between psychological states and performance. The results showed that none of the other psychological states predicted subsequent performance. In the opposite direction, performance predicted several different psychological states the following day, where better performance positively predicted positive states and negatively predicted negative states. In summary, psychological states appear to have a small effect on how NHL players perform, whereas how they perform seem to affect a wide range of psychological states the following day.