Heterogeneity in Community Size Effects: Exploring Variations in the Production of National Hockey League Draftees Between Canadian Cities

Abstract

Previous research has explored ‘community size effects’ in a multitude of sporting and regional contexts and has shown that athletes are more likely to originate from small-medium population size categories, and less likely to originate from very small or large ones. However, it is not clear whether the production of athletes is homogenous within population size categories. Place of birth data were collected for all Canadian born hockey players drafted into the National Hockey League (NHL) from 2000–2014 from British Columbia (N = 192), Alberta (N = 218), Saskatchewan and Manitoba (N = 216), Ontario (N = 561), Quebec (N = 241), and the Atlantic Provinces (N = 74). To explore variations in the production of draftees within population size categories, proportions of productive cities, population mean (μ), population standard deviation (σ), as well as minimum/maximum values of the number of draftees were calculated for the different categories (<2,500; 2,500–4,999; 5,000–9,999; 10,000–29,999; 30,000–99,999; 100,000–249,999; 250,000–499,999; 500,000–999,999; >1,000,000). In addition, the number of draftees produced per 1,000 residents (i.e., yield) was calculated for each city within all categories. Results showed substantial intra-categorical variability in NHL talent development; moreover, heterogeneity in draftee production existed in various degrees across provincial regions of Canada. Intra-categorical variability suggests that a single homogenous community size effect may not exist for Canadian NHL draftees, and that future research may benefit from exploring other environmental constraints on athlete development such as income, population density, and proximity to local sport clubs.

Female Youth Soccer Participation and Continued Engagement: Associations With Community Size, Community Density, and Relative Age

“…The use of wide population categories can potentially hide meaningful variation within the categories themselves (Wattie et al, 2018). Indeed, this was the case when engagement rates for individual communities were examined and has also been observed for National Hockey League draftees across Canada (Farah et al, 2019). Underlying reasons for these findings are likely multifactorial and variable between regions.…”

“…Within-category variation and associations between CD and CS, and CD and the RAE were assessed to further expand current knowledge in this area and direct future research. In consideration of previous research findings, it was hypothesized that mediumsized communities with low-density populations would exhibit a greater likelihood of participation and continued engagement across the longitudinal period; but within-category variation would be observed similar to findings in other Canadian contexts (Wattie et al, 2018;Farah et al, 2019).…”

“…League draftees in all provincial regions of Canada; while also reporting significant heterogeneity when the origin of the same population of athletes (n = 1,502) was analyzed by CS (Farah et al, 2019). It has yet to be determined whether inclusion of the population density variable helps to explain the inconsistencies in CS research; however, it is evident that both overall size and CD should be included in future studies.…”

“…However, a trend toward larger, more densely populated cities was found for elite (>30,000 inhabitants; >250 people/km 2 ) and national (>50,000 inhabitants; ≥1,000 people/km 2 ) football players, while mid-sized communities appeared to be best for the development of elite (between 30,000 and 100,000 inhabitants; 250–1,000 people/km 2 ) and national (between 30,000 and 50,000 inhabitants; no optimal population density identified) handball athletes. Similarly, Farah et al ( 2018 ) found a positive relationship between population density and National Hockey League draftees in all provincial regions of Canada; while also reporting significant heterogeneity when the origin of the same population of athletes ( n = 1,502) was analyzed by CS (Farah et al, 2019 ). It has yet to be determined whether inclusion of the population density variable helps to explain the inconsistencies in CS research; however, it is evident that both overall size and CD should be included in future studies.…”

“…Similar observations were reported for elite ice hockey players in North America (Côté et al, 2006 ) as well as other sports at the professional level, including basketball (Côté et al, 2006 ), baseball (Côté et al, 2006 ), football (MacDonald et al, 2009a ), and golf (Côté et al, 2006 ; MacDonald et al, 2009b ). However, the consistency of these estimates may be somewhat misleading as significant heterogeneity within similar-sized communities has been reported among Canadian National Hockey League draftees from different regions (Wattie et al, 2018 ; Farah et al, 2019 ).…”