Evolution of game-play characteristics within-season for the National Basketball Association

Abstract

The aim of this study was to explore the (dis)similarity of game-play characteristics throughout an in-season period within the National Basketball Association. Thirteen performance-related indicators of all 1230 games of the regular 2016–2017 season in the National Basketball Association were analysed. Non-metric multidimensional scaling was used to examine (dis)similarity of team profiles. The two-dimensional multivariate matrix showed that team profiles generally presented similarity, while the beginning and ending of the season (October and April) showed relative dissimilarity. Although each team presented unique paths throughout the in-season period, the dominant teams in the National Basketball Association presented similar game styles. In addition, the game-play of the teams evolved into effective interactions in terms of offence and defence as the competition progressed while presenting an increased trend in the number of three-point field-goals made (p < 0.000, small effect size (η2) = 0.011) throughout the in-season period. The analytics performed in this study could be used practically to evaluate temporal changes of game-play characteristics in basketball as well as inform strategic periodization plans within season periods.

The Performance Evolution of Match Play Styles in the Spanish Professional Basketball League

“…The aim of the current study was to model the performance of basketball teams' match play considering match-related statistics across eight seasons, and for the most successful or dominant teams during the period under study. As the first examination of the long-term evolution of several professional basketball team's performance [17], the current findings indicated that most match-related performance indicators failed to evolve over time, except for 3-point field goals and assists (2012-2017) which increased while 2-point field goals (2013-2016) decreased. In particular, these indicators reflected the growing impact of shot selection on ball possessions and match outcome [5,12].…”

“…A multivariate analytical method was used to uncover dynamical trends of the team performance indicators [17]. Multivariate methods were chosen as they enabled the mapping of the entire team match styles rather than individual indicators that determine inferences based upon sets of models [21].…”

“…Furthermore, a multivariate method allowed the capture of the temporal trend, simultaneously accounting for all variables in the dataset [17]. While univariate models (e.g., linear regression) can offer powerful insight into individual team performance indicator variability over time, the multivariate technique used allowed for simultaneous analysis and visualisation of the data.…”

“…To our knowledge, only one study has examined the evolution of basketball performances, albeit during one season [17]. This study examined all National Basketball Association teams during the 2016-2017 season and via non-metric multidimensional scaling (nMDS) identified that: (i) team playing profiles were similar at the beginning and end of the season; (ii) each team displayed a unique evolution of performance throughout the season; (iii) there was an increased number of 3-point field goals made by all teams as the season progressed.…”

“…This study examined all National Basketball Association teams during the 2016-2017 season and via non-metric multidimensional scaling (nMDS) identified that: (i) team playing profiles were similar at the beginning and end of the season; (ii) each team displayed a unique evolution of performance throughout the season; (iii) there was an increased number of 3-point field goals made by all teams as the season progressed. This unique analysis was limited to one season only with examination of longer-term trends and team's performances needed to elucidate the evolution of basketball match play and the importance of team's playing styles [16,17].…”