Large data and Bayesian modeling—aging curves of NBA players

Vaci, Nemanja, Cocić, Dijana, Gula, Bartosz and Bilalić, Merim (2019) Large data and Bayesian modeling—aging curves of NBA players. Behavior Research Methods, 51 (4). pp. 1544-1564. ISSN 1554-3528

[img]
Preview
Text (Final published version)
Vaci2019_Article_LargeDataAndBayesianModelingAg.pdf - Published Version
Available under License Creative Commons Attribution 4.0.

Download (1MB) | Preview
[img]
Preview
Text (Advance online version)
Vaci2019_Article_LargeDataAndBayesianModelingAg.pdf - Published Version
Available under License Creative Commons Attribution 4.0.

Download (1MB) | Preview
Official URL: https://doi.org/10.3758/s13428-018-1183-8

Abstract

Researchers interested in changes that occur as people age are faced with a number of methodological problems, starting with the immense time scale they are trying to capture, which renders laboratory experiments useless and longitudinal studies rather rare. Fortunately, some people take part in particular activities and pastimes throughout their lives, and often these activities are systematically recorded. In this study, we use the wealth of data collected by the National Basketball Association to describe the aging curves of elite basketball players. We have developed a new approach rooted in the Bayesian tradition in order to understand the factors behind the development and deterioration of a complex motor skill. The new model uses Bayesian structural modeling to extract two latent factors, those of development and aging. The interaction of these factors provides insight into the rates of development and deterioration of skill over the course of a player’s life. We show, for example, that elite athletes have different levels of decline in the later stages of their career, which is dependent on their skill acquisition phase. The model goes beyond description of the aging function, in that it can accommodate the aging curves of subgroups (e.g., different positions played in the game), as well as other relevant factors (e.g., the number of minutes on court per game) that might play a role in skill changes. The flexibility and general nature of the new model make it a perfect candidate for use across different domains in lifespan psychology.

Item Type: Article
Uncontrolled Keywords: Bayesian modeling, Big dat,a Aging, Motor expertise, Skill development, Lifespan psychology, Basketball
Subjects: C800 Psychology
Department: Faculties > Health and Life Sciences > Psychology
Depositing User: Becky Skoyles
Date Deposited: 11 Jun 2019 10:01
Last Modified: 11 Oct 2019 10:28
URI: http://nrl.northumbria.ac.uk/id/eprint/39639

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics


Policies: NRL Policies | NRL University Deposit Policy | NRL Deposit Licence