NHL aging curves using functional principal component analysis

Cavan Elijah, Cao Jiguo and Swartz Tim B. ()
Additional contact information
Cavan Elijah: Swish Analytics, Burnaby, Canada
Cao Jiguo: Department of Statistics and Actuarial Science, Simon Fraser University, 8888 University Drive, Burnaby, BC V5A1S6, Canada
Swartz Tim B.: Department of Statistics and Actuarial Science, Simon Fraser University, 8888 University Drive, Burnaby, BC V5A1S6, Canada

Journal of Quantitative Analysis in Sports, 2025, vol. 21, issue 3, 177-189

Abstract: When considering future performance in sport, age is an important feature for prediction models. On average, players tend to improve from their rookie (earliest) season, plateau, and then decline in performance until they retire from the league. In this paper we apply Functional Principal Component Analysis to the careers of players from the National Hockey League in order to construct individual aging curves. The approach is nonparametric in the sense that a parametric structure is not imposed on the aging curves. A main aspect of our work is the consideration of selection bias whereby players who have long careers are not randomly sampled but tend to be exceptional players. Whereas the literature constructs aging curves that represent the average player, we produce aging curves for individual players; this is particularly useful in roster construction.

Keywords: aging curves; functional data analysis; National Hockey League; principal component analysis; sports analytics (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://doi.org/10.1515/jqas-2024-0083 (text/html)
For access to full text, subscription to the journal or payment for the individual article is required.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:bpj:jqsprt:v:21:y:2025:i:3:p:177-189:n:1002

Ordering information: This journal article can be ordered from
https://www.degruyte ... ournal/key/jqas/html

DOI: 10.1515/jqas-2024-0083

Access Statistics for this article

Journal of Quantitative Analysis in Sports is currently edited by Mark Glickman

More articles in Journal of Quantitative Analysis in Sports from De Gruyter
Bibliographic data for series maintained by Peter Golla ().