It’s cold and there’s something to do: The changing geography of Canadian National Hockey League players’ hometowns

Abstract

Set within the framework of the birthplace effect literature and the seminal work of Curtis and Birch, this paper draws information from the publicly available database www.hockeydb.com and from the Census to examine the hometowns of Canadian National Hockey League (NHL) players from 1970 to 2015. It found that from a regional perspective, the distribution of players’ hometowns remained fairly stable over the 46-year period with Ontario and the three Prairie provinces being prominent. Players from small centres have been well represented in the NHL. While larger urban areas have historically produced the most players, there has been a marked increase in ‘big city’ players while the odds of making it are low. However, when the analysis is adjusted according to the population aged 10-19, boys growing up in small and mid-sized centres were at advantage in reaching the NHL until 2009. Finally, we discuss whether the growing presence of big city players in the NHL will affect the image of hockey as a national sport, as for many, small-town hockey remains at the heart of Canadian sporting culture.

Keywords

birthplace effect Canada geography hometown ice hockey national sport

Introduction

As a contextual factor, a youth’s community of residence has far-reaching impacts on their life chances, including participation in elite-level sports. Growing up in small communities is known to benefit young people in developing and nurturing athletic skills and encouraging success in competitive sports (Dubé et al., 2007). Research has shown that small-town youth often receive individualized support and attention from families, neighbours, coaches, local fans and media, which help them to develop confidence and athletic skills (Markey et al., 2015).

In Canada, ice hockey is recognized as Canada’s national winter sport, being played across the country including in small remote communities (Gruneau and Whitson, 1993). It is well known that legends of the game, such as Bobbie Orr and Wayne Gretzky, grew up in smaller communities (Parry Sound and Brantford, Ontario, respectively) to advance to the National Hockey League (NHL), giving the residents of these places great community pride and inspiring local youth (Elcombe, 2010; Holman, 2007; Watson, 2017).¹ A number of studies characterized as birthplace effect research have confirmed that smaller and medium-sized communities have higher odds of producing elite hockey players than large metropolitan areas (Baker and Logan, 2007; Côté et al., 2006; Schorer et al., 2010).

Despite the nostalgic image of hockey as a small-town sport, often reinforced by the media and in the national discourse, there is growing awareness that hockey in Canada is emerging as a suburbanized high-cost sport (Ogden and Edwards, 2016). It has been suggested that the game is becoming an exclusive activity in which only children from wealthy suburban families can participate and, in some cases, reach the top level (Pecoskie, 2016c).

It is such a contrasting image of community and hockey in Canada, the long-lasting notion of the small-town advantage or the big city reality of a high-cost sport, that motivates our study. We will illustrate and contextualize the changing relationship between community and participation in elite-level hockey in Canada over the past four decades. We aim to put the birthplace effect research into broad social context; we argue that the birthplace effect research focuses on the optimal community size from a sports science point of view but does not address the changing hometown geography of elite Canadian hockey players. Our research questions are twofold. First, how has the geographic distribution of Canadian-raised NHL players changed since 1970 in terms of province, community size, and distance to large cities? Second, how has the association between hometown community size and participation in elite hockey changed over time? To answer these questions, we examine the geographic distribution of the hometown of Canadian-raised NHL players from 1970 to 2015 using publicly available information on NHL players.

This study will make three contributions. First, it complements existing qualitative case studies that provide rich description of sports in small communities by quantifying the changing representation of hockey players at regional and national levels. Second, it will make a methodological contribution to the study of contextual factors in participation in elite-level sports by identifying NHL players’ hometowns, rather than birthplaces, as a more accurate measure of athletes’ place of talent development by way of a thorough ‘hometown verification’ (detailed in the Data and Methods section). In this paper, we define hometown as the place where a boy spent at least a portion of his childhood and developed his skills in minor hockey before moving to play at a higher level such as major junior or US college hockey, with this move typically happening between the age of 15 and 19. We regard hometown to be the place where the boy lived, for at least a period of time between the ages of 10 to 19, before leaving to play elite-level hockey. Finally, our study aims to spark a public debate about the changing social dynamics of youth hockey in Canada, given that hockey has a strong connection to Canada’s national identity (Allain, 2011; Watson, 2017).

Literature review

Arguably, participation in elite-level team sports, such as becoming a professional hockey player, is a unique socioeconomic outcome, attained only by select individuals with outstanding athletic skills. Numerous studies have examined factors contributing to this exceptional outcome, including: genetic; psychological; training; socio-cultural factors; and family support (Baker and Horton, 2004; Baker and Logan, 2007; Hancock et al., 2018; Low et al., 2015). One contextual factor that has attracted sociologists’ attention is the size of hometown, the community where an elite athlete spent developmental years. The type of community support can be diverse, including volunteer coaches, financial support for sports facilities and events from local businesses, and emotional support from the local media and fans. As such, research has examined the association between community size and participation in elite-level team sports, considering community size as a broad-based indicator of community support for early exposure to the development of sport talent (Côté et al., 2006).

In the case of participation in elite-level hockey in Canada, it is widely believed that male hockey players are disproportionately coming from small farming towns in the Prairie (Alberta, Saskatchewan, and Manitoba) and single resource industry towns in the North (Northern Ontario and Quebec) and West (British Columbia). These rural and small-town communities are often considered to have played a role in helping their local boys to make it to elite-level hockey (Gruneau and Whitson, 1993). These two types of small communities fit with Davis’ concept of hinterland, the remote place that is economically and politically dominated and exploited by an urban metropolis (Davis, 1971). And anecdotal evidence abounds. For example, the mining towns of northern Ontario, such as Renfrew, Cobalt, and Haileybury, formed the Temiskaming Hockey League in the early 20th century with the financial backing of local mining companies (Gruneau and Whitson, 1993). Moreover, in the 1940s–1960s, with the development of infrastructures (e.g. rural roads and arenas), some remote mining and mill towns (e.g. Rouyn (Quebec), Kirkland Lake (Quebec), Timmins (Ontario)) thrived as a feeder of elite hockey players. In these company towns, the local communities were virtually formed around minor-hockey organizations, involving Lions Clubs, church groups, businesses, volunteer coaches, forming a community hockey culture (Gruneau and Whitson, 1993).

In Canada, such a rural small-town hockey narrative had been established when Curtis and Birch (1987) quantitatively examined the association between community size and participation in elite-level hockey. Their analysis of the rosters of players in the three most prestigious North American hockey leagues (the NHL, the American Hockey League, and the Central Hockey League) during the 1982–1983 season and the USA and Canadian national teams in the Olympics in 1968–1984 found players from rural areas (population under 1000) and the largest communities (500,000 or more) are noticeably underrepresented among elite hockey players, while those from smaller- and middle-size communities (1000–499,999) are overrepresented. Curtis and Birch (1987) speculated that the underrepresentation of rural players might be related to the lack of opportunity structures; rural areas tend to have fewer sports facilities and less access to competitive organized sports and quality coaching. They also argued that the underrepresentation of players from the largest cities may be due to the greater availability of other alternatives for leisure, less visibility to scouts, and limited all-star teams per capita.

Curtis and Birch’s (1987) study helped to demystify the rural small-town hockey narrative which implied the linear relationship between community size and recruitment of elite hockey players and the smallest rural community’s advantage. Instead, growing up in the smallest rural community (population of less than 1000) in fact disadvantages youth’s prospects of making it to elite-level hockey. Although Curtis and Birth (1987) did not delve into the advantage of smaller- and middle-size communities, their interpretation suggests that in these communities, hockey plays a more central role in community events than in the largest urban areas and that a sufficient population base of male youths exists to maintain quality team and star players (compared to rural areas).²

Since the publication of Curtis and Birch (1987), numerous studies have examined the community size and participation in various elite-level sports (e.g. baseball, basketball, cricket, volleyball, soccer, handball) in different countries (e.g. Canada, USA, UK, Israel, Sweden, Germany, Finland, Ireland, Portugal), having developed into the birthplace effect research (Baker et al., 2009, 2011; Lidor et al., 2014; Rossing et al., 2016). Although findings vary depending on the sport and countries studied, there is consensus in the Canadian research that those born in smaller communities (except for rural areas) are more likely to succeed in elite hockey. As the field developed in sports science, the birthplace effect research focuses on identifying the optimal community size for participation in elite-level sport. In the case of hockey in Canada, studies agree that the optimal community size is approximately 100,000–500,000 (Baker et al., 2014; Baker and Logan, 2007; Côté et al., 2006).

While the present study builds on the birthplace effect literature by examining the size of NHL players’ hometowns, we aim to put the community size–NHL entry nexus in broad context, going beyond the cross-sectional, national-level analysis that is commonly performed in the existing research. We extend our analytical scope both spatially and temporally; we examine the geographic distribution of the hometowns of Canadian-raised NHL players who played their first NHL game between 1970 and 2015 and consider multiple levels of geography, province, community size, and proximity to large urban areas. As discussed in Curtis and Birch (1987), distributions of rural versus urban areas vary across provinces, in that rural areas are overrepresented in Western Canada. Our province-specific analysis thus helps to delve into the roles of socio-cultural contexts that may explain the small-community advantage (Rooney, 1969). In addition, long-distance travelling is an integral part of youth hockey talent development in smaller communities, yet previous research has rarely considered such a distance factor (exceptions: Atherley, 2006; Tonts and Atherley, 2005).

Moreover, this study contributes to the timely debate concerning the changing nature of youth hockey in Canada that is emerging in journalistic and academic works: the rising financial cost of playing the game and the suburbanization of competitive youth hockey (Decosse, 2015; Gruneau, 2010; Solutions Research Group Consultants Inc., 2014). For instance, Teri Pecoskie’s (2016a, 2016b, 2016c) three-part series ‘Pay to Play’ in the Hamilton Spectator newspaper points to the fact that the majority of elite junior players in the Ontario Hockey League (an important feeder league to the NHL) grow up in wealthier suburbs of the largest cities. And the suburbanization of the paths to elite hockey may be related to the privatization of elite hockey enterprise in which privately-owned ice rinks are emerging in wealthy suburbs and urban areas.

Meanwhile, remote hinterland communities that were once thriving as single resource industry towns are losing community hockey resources as a result of de-industrialization, economic restructuring, and the departure of main employers. Qualitative evidence from Flin Flon (Manitoba), Cape Breton (Nova Scotia), and Trail (British Columbia) shows that the closure of the local mills and mines has had broad spillover effects on the local community hockey culture (Decosse, 2015; Ferguson, 2014; Lawson, 1998; Oncescu, 2015). The consequential demographic changes (e.g. permanent out-migration of families, long-distance commuting of middle-aged men, who go to work in another province) and economic decline (e.g. disappearance of well-paid ‘good’ jobs) are considered to be contributing to the decline of volunteer hockey coaches, families’ growing challenge in financially supporting child hockey training, and the community’s inability to maintain quality minor hockey teams.

Our examination of the hometowns of the NHL players over 46 years helps to shed light on such divergent pictures of community and hockey – small-town community sport versus middle-class suburban sports – by contextualizing the relationship between the size of NHL players’ hometown of origin and entry in the NHL in changing geographic distributions of the players over 46 years.

Data and methods

We use the Hockey Database (www.hockeydb.com) to create a data file consisting of every person who played at least one game in the NHL between 1970 and 2015. The information includes birthdate, birthplace, number of seasons, position and statistics including games played, goals, assists and penalty minutes. Our file lists 5637 players over the 46-year study period.

Hometown verification

We aim to determine the impact of geographic location on the likelihood of a person making it to the NHL. For this reason, we are primarily interested in the hometown of a player, rather than solely their birthplace. For each player listed as being born in Canada and the USA as well as select European and other players (n = approximately 4500), we conducted an exhaustive search from multiple sources to ascertain the player’s hometown. Undeniably, any person who becomes a professional athlete is an elite participant in their sport, and their development as such is established in adolescence. For professional hockey players, the formative years of their development are perhaps between the ages of 10 and 19 when the influence of parents, coaches, volunteers and community facilities (e.g. arenas) is crucial (Dubé et al., 2007).

While the pathways to the NHL are varied, in most cases in North America, these elite players are recruited or drafted from Bantam, Midget or High School leagues to Major Junior leagues or US colleges and are often required to leave their hometown as teenagers. As stated earlier, this paper considers hometown to be the place where a boy played minor hockey for a period of time between the ages of 10 and 19 (typically at the AA or AAA levels) before leaving home (most likely between the ages of 15 and 19). The age of departure can vary. For instance, a Bantam or Midget player drafted into the Ontario Hockey League is typically between the ages of 15 and 17 and in the large majority of cases is required to leave their hometown to be billeted with a local family in a new town or city. A Midget or junior player recruited to a National Collegiate Athletic Association (NCAA) Division 1 team is likely to be 17 to 19 years old and in virtually all cases will live in college housing. While the circumstances vary, we regard hometown as the place where the boy played minor hockey (and most likely lived with his parents) before leaving home to play at a higher level.

We consulted several sources in addition to the Hockey Database to verify a player’s hometown.

Hockey Draft Central (www.hockeydraftcentral.com)

The Society for International Hockey Research (SIHR) database

Official site of the Hockey Hall of Fame (HHOF) NHL Player Search (www.hhof.com/html/search.shtml)

Wikipedia entries

NHL team media guides

These sources contain brief profiles of players, including their birthplace, hometown and professional and personal accomplishments. For each player, our approach was to check him in the first three sources (Draft Central, SIHR and HOOF), and if we identified a consistent mention of hometown in two or more, this was recorded. If we detected an inconsistency or a lack of information, we further checked the player by way of Wikipedia and a media guide (if available online). Our search resulted in a total of 603 edits (11% of all player entries) to the birthplace to record the hometown of each player. We concentrated on Canadian-born players but also verified others born outside Canada. The large majority of edits were Canadian-born players moving within the country (568, 16% of all Canadian-born players). Our sample also includes a small number of USA-born players who moved to Canada during childhood (n = 7) and several European-born and other region-born players (n = 22) who grew up playing minor hockey in Canada. Meanwhile, we exclude some Canadian-born players who moved to the USA (n = 20) during childhood. After the hometown verification, we obtained the information about 3469 NHL players whose hometowns are located within Canada. This is the main sample of our analysis of Canadian-raised players.

Time periods

To capture trends over time, we divided the 46-year study period into five time-blocks: 1970 to 1979; 1980 to 1989; 1990 to 1999; 2000 to 2009; and one six-year period (2010 to 2015). To adequately account for Canada’s changing population, we used data from Statistics Canada’s Censuses covering this period (1976, 1986, 1996, 2006 and 2011) to adjust the geographic classification of places (provinces, towns and cities). We obtained the Census information from the University of Toronto’s Canadian Census Analyser.

Geographic classification

In addition to the ten provinces and three territories, we considered geography based on Statistics Canada’s longstanding definition of places.

Class 1: Largest urban (500,000 + residents)

Class 2: Large urban (100,000–499,999 residents)

Class 3: Medium centre (30,000–99,999 residents)

Class 4: Small centre (1000–29,999 residents)

Class 5: Rural/remote (less than 1000 residents)

For each of the 3469 Canadian-raised NHL players, we classified their hometown into one of these five types according to the five time periods based on population data from the Canadian Censuses. We used the Canadian Census Analyser to verify the population of each hometown over the 46-year period and to adjust its geographic classification (Class 1 to Class 5) if population change (primarily growth) warranted it. Certain hometowns (e.g. Toronto, Montreal, Calgary, Edmonton) are classified as largest urban (Class 1) throughout the study period. Due to population growth, other hometowns, including Milton (Ontario), Brantford (Ontario), Charlottetown (Prince Edward Island), Kelowna (British Columbia) and Lethbridge (Alberta), moved up in the classification to become larger places over time (e.g. Class 4 to Class 3, Class 3 to Class 2).

Distance to a big city

Another important dimension in amateur athletics is travel distance, particularly with respect to smaller, remote communities. Like all aspiring athletes, youth hockey players benefit from a certain critical mass of competition and access to resources available in more populated places. More players, teams, arenas, ice time, coaches and volunteers all lead, in theory, to the positive development of young athletes.

With reference to small centres (Class 4) and rural/remote places (Class 5), we measured proximity to a large urban area (Classes 1 and 2) by distance. If the community is within 100 km (or about an hour drive), it was in proximity (coded 1) and not in proximity (coded 0) otherwise. For each of the 1161 players who grew up in small centres or rural/remote areas (Classes 4 and 5) in our dataset, we used the travel distance function in Google Maps to calculate proximity. Admittedly, a limitation in this measurement is that the size and quality of many highways and major arterial routes changed over the 46-year study period. For instance, a small county road in Ontario in the early 1970s may have developed into a four-lane highway in the late 1990s, significantly improving travel times. We were unable to account for these changes and used only the current information provided in Google Maps.

Population-based proportions

Given that Canada has witnessed considerable demographic change in the past five decades, it is important that population-based studies, such as this one, utilize a relative or proportional measure of an outcome. Using data from five Censuses (1976, 1986, 1996, 2006 and 2011), we calculated provincial and territorial rates of NHL hockey players per 1000 males aged 10–19 over the study period. These rates were produced according to Statistics Canada’s classification of places permitting us to not only examine total numbers but also the relative impact of each province in producing NHL players based on its community size. To render the rates per 1000 comparable over the five time periods, we estimated 10-year totals for 2010–2015. We did this by multiplying the total number of NHL players over the six years in the provinces and territories by 1.67 (=10/6). For comparing population proportions, this renders the final time-block (2010–2015) arithmetically comparable to the others.

Calculating odds ratios (ORs)

Following previous research, we also calculated the ORs by dividing the odds of playing in the NHL by the odds of growing up in a community of specific size (Classes 1–5) (Côté et al., 2006; Hancock et al., 2018; Low et al., 2015; Turnnidge et al., 2014). An OR greater than one indicates players growing up in a community of a specific size are more likely to make it to the NHL than those growing up elsewhere. Conversely, an OR smaller than one means male youth growing up in a community of a specific size are less likely to play in the NHL. In addition, the 95% confidence interval (CI) is calculated for each OR. When the CI includes the value of one (the null value of OR), the OR is not significantly different from one (p > 0.05).

Results

The following findings reflect the entire study period (1970–2015) as well as five sub-periods (1970–1979, 1980–1989, 1990–1999, 2000–2009 and 2010–2015). For the five sub-periods, it is important to note that the regional breakdowns refer to those who played their first game in the NHL during that period and not the total number of Canadian-raised players who suited up for a team in those years.

NHL players by region of the world

Our database indicates that 5637 players appeared in at least one NHL game in 1970–2015. Table 1 shows the distribution of these players according to region of the world and time period. Over the 46-year period, players with Canadian hometowns (n = 3469) comprised 62% of the total. In the 1970s, the NHL was dominated by Canadian-raised players, making up 84% of all players. Since then, the proportion has dropped steadily, falling to below half (47%) in 2010–2015. Meanwhile, the proportion of US- and European-raised players has been steadily rising, particularly since the 1990s. The rapid rise in European-raised players (who now make up more than one-quarter of all players) can be partially attributed to the dissolution of the Soviet Union in 1991 which resulted in dozens of elite hockey players from former Soviet republics (including Russia) being made available to NHL teams (Baker et al., 2014). Parallel to this is the ascendency of US hockey that has produced numerous elite athletes through the high school, private academy, USA Junior Hockey Leagues and NCAA programmes. With minor and junior Canadian hockey leagues remaining strong, these trends have certainly resulted in a more skilled and competitive NHL that grew from 14 teams in 1970 to 30 in 2015.

Table 1.

National Hockey League (NHL) player’s hometowns by region of the world and time period.

Region	1970–2015	1970–1979	1980–1989	1990–1999	2000–2009	2010–2015
Total	5637 (100%)	984 (100%)	1175 (100%)	1243 (100%)	1367 (100%)	868 (100%)
Canada	3469 (61.5%)	830 (84.3%)	853 (72.6%)	692 (55.7%)	690 (50.5%)	404 (46.5%)
USA	1036 (18.4%)	117 (11.9%)	191 (16.3%)	211 (17.0%)	282 (20.6%)	235 (27.1%)
Europe	1021 (19.9%)	37 (3.8%)	131 (11.1%)	337 (27.1%)	387 (28.3%)	229 (26.4%)
Other	11 (0.2%)	0 (0%)	0 (0%)	3 (0.2%)	8 (0.6%)	0 (0%)

Note: NHL players refer to those who played at least one game in the NHL. The time period is based on the year the individual played his first NHL game.

Canadian-raised NHL players’ hometowns by region and community size

From now on, our analysis focuses only on Canadian-raised NHL players (n = 3469). Table 2 suggests that while the percentage of Canadian-raised players fell between 1970 and 2015, the regional distribution of the players’ hometown within Canada remained fairly stable. As Canada’s most populous province, Ontario supplied the largest share of NHL players, 42% for the entire period and 46% for 2010 to 2015. Interestingly, the three Prairie provinces (Manitoba, Saskatchewan and Alberta) consistently produced about twice as many players as Quebec despite having a combined population significantly lower than their eastern counterpart.

Table 2.

Canadian-raised National Hockey League (NHL) players’ hometowns by region and time period.

Region	1970–2015	1970–1979	1980–1989	1990–1999	2000–2009	2010–2015
Total	3469 (100%)	830 (100%)	853 (100%)	692 (100%)	690 (100%)	404 (100%)
Atlantic	134 (3.9%)	35 (4.2%)	26 (3%)	23 (3.3%)	34 (4.9%)	16 (4.0%)
Québec	536 (15.5%)	135 (16.3%)	126 (14.8%)	136 (19.7%)	92 (13.3%)	47 (11.6%)
Ontario	1455 (41.9%)	386 (46.5%)	374 (43.8%)	260 (37.6%)	250 (36.2%)	185 (45.8%)
Prairies	971 (28.0%)	213 (25.7%)	232 (27.2%)	195 (28.2%)	227 (32.9%)	104 (25.7%)
British Columbia/Territories	373 (10.8%)	61 (7.3%)	95 (11.1%)	78 (11.3%)	87 (12.6%)	52 (12.9%)

Note: Canadian-raised NHL players refer to the individual who grew up in Canada and played at least one NHL game. The time period is based on the year the individual played his first NHL game.

Sports journalists and researchers have commented on the rising costs to parents of having their children participate in hockey and the concentration of the most competitive teams and leagues in the wealthier suburbs of big cities (Decosse, 2015; Frisby et al., 2005; Pecoskie, 2016a, 2016b, 2016c). Table 3 appears to confirm this notion. While larger urban areas (with a population of 100,000 or more) have historically produced the most NHL players, players from ‘big city’ areas noticeably increased from 44% in 1970–1979 to 58% in 2010–2015. In fact, by 2010–2015, the three largest metropolitan areas, the regions of Greater Toronto, Montreal and Vancouver (including suburbs – peripheral municipalities that surround the Toronto, Montreal and Vancouver municipalities, such as Markham, Oakville and North Vancouver) represented the hometowns of more than one-third (35%) of all Canadian-raised NHL players who started playing in that period (Turcotte, 2008; Vézina and Houle, 2018). When the five other most populous cities, Winnipeg, Edmonton, Calgary, Ottawa–Gatineau and the Hamilton–Niagara region, are added, this proportion increases to 52%.

Table 3.

Canadian-raised National Hockey League (NHL) player’s hometowns by centre size and time period.

Classification	1970–2015	1970–1979	1980–1989	1990–1999	2000–2009	2010–2015
Total	3469 (100%)	830 (100%)	853 (100%)	692 (100%)	690 (100%)	404 (100%)
Total large urban	1702 (49.1%)	364 (43.8%)	415 (48.7%)	339 (49.0%)	350 (50.7%)	234 (57.9%)
Largest urban (500,000+)	940 (27.1%)	202 (24.3%)	214 (25.1%)	204 (29.5%)	187 (27.1%)	133 (32.9%)
Large urban (100,000–499,999)	762 (22.0%)	162 (19.5%)	201 (23.6%)	135 (19.5%)	163 (23.6%)	101 (25.0%)
Medium centre (30,000–99,999)	606 (17.5%)	157 (18.9%)	163 (19.1%)	101 (14.6%)	102 (14.8%)	83 (20.5%)
Small centre (1000–29,999)	987 (28.5%)	257 (31.0%)	235 (27.5%)	223 (32.2%)	195 (28.3%)	77(19.1%)
Rural/remote area (less than 1000)	174 (5.0%)	52 (6.3%)	40 (4.7%)	29 (4.2%)	43 (6.2%)	10 (2.5%)

Note: Canadian-raised NHL players refer to the individual who grew up in Canada and played at least one NHL game. The time period is based on the year the individual played his first NHL game.

Table 3 also shows that players from small centres and rural and remote communities are well represented in the NHL, comprising one-third (34%) of the total between 1970 and 2015. This proportion remained fairly stable until 2010–2015, when it suddenly dropped to less than one-quarter (22%), with just 10 players from rural or remote hometowns starting in the NHL over the six years.

Small-town hockey: the pull of the big city

Travelling has significant implications for young minor hockey players in smaller communities as it requires substantial time and cost for their skill development in competitive hockey. Interestingly, Table 4 reveals an overall ‘distance advantage’ for NHL players from small or rural/remote hometowns with a population less than 30,000. Over the 46-year study period, about 60% of these players lived in a place located more than 100 km (and likely more than a one-hour drive) from a large urban centre. Of course, owing to their unique geography, the distance factor is particularly noticeable in the three Prairie provinces (results not shown here), which as indicated, have produced a relatively large number of NHL players. We can speculate that an extra level of commitment, dedication and organization is present among athletes, parents, coaches and volunteers in smaller, more distant places, creating a positive social environment from which young hockey players can benefit (Dubé et al., 2007).

Table 4.

Canadian-raised National Hockey League (NHL) player’s hometown, for rural/remote areas and small centres by proximity to a large urban centre and time period.

	Distance to large urban area	1970–2015	1970–1979	1980–1989	1990–1999	2000–2009	2010-2015
Total rural/remote and small centres (30,000+)	Total	1161 (100%)	309 (100%)	275 (100%)	252 (100%)	238 (100%)	87 (100%)
	> 100 km	696 (59.9%)	195 (63.1)	174 (63.3%)	155 (61.5%)	137 (57.6%)	35 (40.2%)
	< 100 km	465 (40.1%)	114 (36.9%)	101 (36.7%)	97 (38.5%)	101 (42.4%)	52 (59.8%)
Small centres (1000-29,999)	Total	987 (100%)	257 (100%)	235 (100%)	223 (100%)	195 (100%)	77 (100%)
	> 100 km	558 (56.5%)	155 (60.3%)	141 (60.0%)	131 (58.7%)	103 (52.8%)	28 (36.4%)
	< 100 km	429 (43.5%)	102 (39.7%)	94 (40.0%)	92 (41.3%)	92 (47.2%)	49 (63.6%)
Rural/remote (< 1000)	Total	174 (100%)	52 (100%)	40 (100%)	29 (100%)	43 (100%)	10 (100%)
	> 100 km	138 (79.3%)	40 (76.9%)	33 (82.5%)	24 (82.5%)	34 (79.1%)	7 (70.0%)
	< 100 km	36 (20.7%)	12 (23.1%)	7 (17.5%)	5 (17.2%)	9 (20.9%)	3 (30.0%)

Note: Canadian-raised NHL players refer to the individual who grew up in Canada and played at least one NHL game. The time period is based on the year the individual played his first NHL game.

However, Table 4 also indicates the distance advantage from the early 1970s until 2009. In fact, the proportions in 2010–2015 are essentially the reverse of the preceding period (2000–2009) when smaller centres within 100 km of a large urban area made up 60% of players starting to play in the NHL. Not surprisingly, this development coincides with the growing urbanization of hockey likely the direct result of demographics (e.g. the population shift to big cities, particularly the suburbs) and socioeconomic trends (e.g. changes in income distribution and rising costs to play hockey).

Provincial impact factors

To account for Canada’s considerable demographic changes since the 1970s, we calculate the rates of NHL players per 1000 males aged 10–19 adjusted for the five periods by way of corresponding Census information (Table 5). The question is: all things equal, is there a provincial or hometown size advantage which enables a boy to play in the NHL? It is noteworthy that when the analysis is reframed according to an adjusted population, Table 5 indicates that the Ontario large urban ‘advantage’ presented above is not so apparent. Due to the province’s large population (e.g. over 833,000 males aged 10–19 in 2011) and the exclusivity of being an NHL player, the rates for large urban Ontario hometowns between 2000 and 2015 are not noticeably higher than other provinces (for emphasis, rates above 0.70 are highlighted in Table 5). In other words, while increasing numbers of professional players are coming from select big cities in Canada, the mass of minor and junior hockey players registered in leagues in these places means that the likelihood of making it to the NHL is not higher.

Table 5.

National Hockey League (NHL) players by size of hometown (rates per 1000 males aged 10–19).^a

Region	Class	1970–1979	1980–1989	1990–1999	2000–2009	2010–2015^b
Atlantic	Large urban	0.03	0.19	0.20	0.45	0.27
	Medium centre	0.50	0.27	0.26	0.35	0.48
	Small centre	0.29	0.24	0.33	0.44	0.29
	Rural/remote	0.01	0.03	0	0.01	0.05
Quebec	Large urban	0.20	0.26	0.29	0.16	0.12
	Medium centre	0.27	0.71	0.47	0.48	0.67
	Small centre	0.38	0.53	0.64	0.48	0.32
	Rural/remote	0.07	0.04	0	0	0
Ontario	Large urban	0.43	0.54	0.33	0.28	0.36
	Medium centre	1.34	1.43	0.79	0.56	0.83
	Small centre	0.82	0.89	0.75	0.63	0.68
	Rural/remote	0.02	0.04	0.01	0.01	0
Manitoba	Large urban	0.42	0.44	0.39	0.37	0.58
	Medium centre	2.26	2.17	0.38	2.56	2.53
	Small centre	1.13	0.98	0.98	1.20	0.93
	Rural/remote	0.23	0.18	0.14	0.24	0.18
Saskatchewan	Large urban	0.94	1.14	0.4	1.04	0.58
	Medium centre	2.06	1.59	0.93	0.19	0.35
	Small centre	1.71	2.14	1.51	2.11	0.60
	Rural/remote	0.38	0.37	0.38	0.80	0.32
Alberta	Large urban	0.33	0.68	0.62	0.41	0.46
	Medium centre	1.04	1.24	0.64	0.54	0.76
	Small centre	0.88	0.92	0.87	0.85	0.47
	Rural/remote	0.18	0.17	0.18	0.17	0.04
British Columbia	Large urban	0.16	0.30	0.17	0.24	0.28
	Medium centre	0.71	0.71	0.28	0.50	0.87
	Small centre	0.53	1.31	1.12	0.85	0.42
	Rural/remote	0.02	0.02	0.02	0.02	0
Territories	Large urban	0	0	0	0	0
	Medium centre	0	0	0	0	0
	Small centre	0.80	0	0.26	0.21	0
	Rural/remote	0	0	0.22	0	0

Note: ^ayouth population estimates derived from the 1976, 1986, 1996, 2006 and 2011 Censuses; ^bfor 2010–2015, totals and rates are based on 10-year estimate; values greater than 0.70 are in bold.

Moreover, we find several instances where smaller community size did present an advantage prior to 2010. Consistent with the results above, these are visible in the three Prairie provinces among medium and small centres, particularly during the 1970s and 1980s. Up to 2010, Saskatchewan had impressive success in producing NHL players from its small centres (places with a population between 1000 and 29,999) even in the 2000–2009 period (with a rate of 2.11) when the NHL was undergoing a transition in its player’s base. What is even more remarkable is that these communities have relatively small numbers of males aged 10–19 compared to the rest of Saskatchewan and other provinces. Above we referred to the strong community commitment that is almost certainly a driving force in these places and others across Canada. In the context of small population sizes (and hence fewer minor hockey registrations) and scarcer resources, we can further advance that dedicated young athletes receive more individualized attention and encouragement from coaches and volunteers that greatly assists in their development (Hansen and Musynzki, 1990). While big city hockey programmes certainly have their strong points, perhaps many children become lost in the sheer size and complexity of these leagues.

The odds of making it to the NHL

Table 6 shows the ORs of making it to the NHL for a male youth growing up in a community of a specific size nationally, as well as by the province/territory from 1970 to 2015 in 10-year intervals. At the national level, we identify three notable trends. First, rural and remote areas were consistently underrepresented throughout the study period. For a player who grew up in rural and remote areas, the ORs of making it to the NHL remained 0.1–0.2 from 1970 to present, significantly lower than elsewhere (p < 0.05). Second, small and medium centres were continuously overrepresented as the hometowns of the NHL players. Interestingly, the advantage of growing up in small centres has been on the rise since the 1990s (except for the 2010–2015 period), whereas the advantage of coming from medium centres has been declining. Third, there was neither significant advantage nor disadvantage in growing up in large urban centres; the ORs were not statistically different from 1 (p > 0.05). These findings are consistent with the existing birthplace effect research on Canada’s elite hockey players: the advantage of coming from smaller cities (but not rural/remote communities) (Baker et al., 2014; Baker and Logan, 2007; Côté et al., 2006; Curtis and Birch, 1987; Turnnidge et al., 2014). Meanwhile, a novel finding is the changing magnitude of the advantage of growing up in small and medium centres over time, which requires further investigation.

Table 6.

Odds ratios (ORs) of playing in the National Hockey League (NHL) by region and geographic classification, 1970–2019.

Region	Geographic classification	1970–1979		1980–1989		1990–1999		2000–2009		2010–2019^a
Atlantic Canada	Large urban	0.2		1.6		1.6		2.5		1.6
	Medium centre	4.7		2.3		2.2		1.7		2.9
	Small centre	3.5		2.7		4.4		2.6		1.7
	Rural/remote	0.0		0.1		NA		0.0		0.2
Quebec	Large urban	1.0		0.9		1.1		0.7		0.5
	Medium centre	1.3		3.0	*	1.8		3.0		5.5	*
	Small centre	2.2		2.3		2.9	*	3.2	*	2.2
	Rural/remote	0.3		0.1		NA		NA		NA
Ontario	Large urban	0.8		0.9		0.8		0.9		0.9
	Medium centre	3.4	*	3.0	*	2.5	*	2.1		2.5	*
	Small centre	2.0	*	1.7		2.5	*	2.4	*	2.0	*
	Rural/remote	0.0	*	0.1	*	0.0		0.0		0.0
Manitoba	Large urban	0.7		0.9		1.1		0.6		1.1
	Medium centre	5.1		5.2		1.0		6.0		5.2
	Small centre	2.8		2.4		3.3		3.0		1.9
	Rural/remote	0.4		0.3		0.3		0.4		0.2
Saskatchewan	Large urban	1.1		1.2		0.6		0.9		1.4
	Medium centre	2.5		1.6		1.6		0.2		0.7
	Small centre	2.6		2.8		3.5		2.5		1.3
	Rural/remote	0.3	*	0.2	*	0.5		0.6		0.6
Alberta	Large urban	0.7		1.1		1.3		0.8		1.4
	Medium centre	2.9		2.1		1.2		1.2		2.1
	Small centre	2.7		1.6		1.8		2.4		1.2
	Rural/remote	0.4		0.2		0.3		0.3		0.1
British Columbia	Large urban	0.5		0.5		0.4		0.6		0.7
	Medium centre	3.4		1.6		0.9		1.7		3.5	*
	Small centre	3.2		4.6	*	7.2	*	3.6	*	1.3
	Rural/remote	0.0		0.0		0.1		0.1		0.0
Territories	Large urban	NA		NA		NA		NA		NA
	Medium centre	NA		NA		NA		NA		NA
	Small centre	NA		NA		1.2		NA		NA
	Rural/remote	NA		NA		0.8		NA		NA
Canada Total	Large urban	0.9		1.0		0.9		0.8		1.0
	Medium centre	2.7	*	2.6	*	1.7	*	1.8	*	2.8	*
	Small centre	2.2	*	2.1	*	2.9	*	2.8	*	1.6	*
	Rural/remote	0.2	*	0.1	*	0.1	*	0.2	*	0.1	*

Notes: *OR significantly different from 1 at p = 0.05 level, based on 95% confidence intervals (not shown here); NA: ORs not available because no one played in the NHL from these regions; ^a results for 2010–2019 are estimated using the 2010–2015 results.

When we calculated the ORs for each province/territory, most of the ORs lost statistical significance. None of the ORs for Atlantic Canada, Manitoba or Alberta were statistically different from 1 (p > 0.05), likely due to smaller numbers of cases under study. However, we found overrepresentation of players from small and medium centres in Quebec, Ontario and British Columbia. Overall, the ORs for small centres in these provinces are gradually declining over the past four decades, perhaps reflective of the declining representation of players from resource-based towns. At the same time, the ORs for medium centres in these provinces have surged in the most recent period (2010–2015). Also, the underrepresentation of players from rural/remote areas was identified in Ontario and Saskatchewan in the 1970s–1980s, yet this pattern lost statistical significance in the more recent decades.

Conclusion and discussion

This study has examined the changes in geographic distribution of the Canadian-raised NHL players’ hometown from 1970 to 2015. Using multiple sources of data that are publicly available online, we have identified an NHL player’s hometown as the place where he grew up and played minor hockey prior to leaving home to play at a more competitive level. Given that more than 10% of the NHL players grew up in a community that was different from their birthplace, our ‘hometown verification’ makes a methodological contribution to the birthplace effect research, which uses the player’s birthplace as a proxy of hometown.

Overall, our study supports both the journalistic and scholarly accounts of the geography of participation in elite-level hockey. While the share of Canadian-raised players has been steadily declining since 1970s, their regional distribution within Canada remains stable, with Ontario and the Prairie provinces (Alberta, Saskatchewan and Manitoba) having notable representation. Also, smaller communities (small and medium centres) have maintained a decent share of the NHL players’ hometowns until recently. Boys growing up in small and medium centres (especially those from Saskatchewan) were continuously overrepresented and at advantage in reaching the NHL, consistent with the findings from the existing birthplace effect research.

However, since 2010, a new trend has emerged. The most recent period (2010–2015) can be characterized by the rise of ‘big city’ players in terms of the overall share of players from large urban centres. Nevertheless, the odds of making it to the NHL from large urban areas are notably low. This is consistent with an emerging image of today’s elite male hockey in Canada: the suburbanization of hockey as a high-cost sport (Decosse, 2015; Pecoskie, 2016a, 2016b, 2016c). Our hometown database indeed indicates that suburban municipalities of the largest cities, such as Newmarket (Connor McDavid’s hometown) and Brampton (Sean Monahan’s hometown), have gained a notable share as the player hometowns.

These patterns can be summarized as a polarization of the paths to elite hockey. While large urban areas have become the dominant source of NHL players, the environment surrounding urban youth players may not be so supportive for their skill development. The results may indicate the emergence of a select number of highly competitive youth who possess exceptional talent or whose parents can afford to pay for expensive, personalized training and have the proper resources and connections to help their sons to navigate through a complex youth hockey system in suburban areas. By contrast, youth from smaller communities continue to have an advantage in making it to the NHL, which may be reflective of the stronger social and emotional support from coaches, families, friends, local fans and academic teachers (Smailes, 1997; Tonts and Atherley, 2005). Considering that a number of rural and smaller communities across Canada have been affected by economic restructuring since the 1970s, the fact that smaller communities remained as a stronghold of the NHL players’ hometowns until 2009 may be indicative of the resilience of smaller communities (Hansen and Musynzki, 1990; Oncescu, 2015). Nevertheless, in reality, players from smaller communities have been overtaken by savvy, competitive urban youth players in recent years.

Like many studies of birthplace effects, our research has a data limitation. Although we use the local male youth population in our analysis of the odds of making it to the NHL, more ideal data would be the pool of minor hockey players, who have the potential to reach the NHL. While minor hockey registration data are potentially available, it would be difficult to incorporate these into a large national level study with a long time period (46 years) and rapidly changing geographic classification such as this one. However, an appropriate follow-up study is one that employs registration data over a number of seasons for selected regions such as the Greater Toronto Hockey League with its 52 teams at the A, AA, AAA levels or the British Columbia Amateur Hockey Association in its Pee Wee, Bantam and Midget divisions (ages 11–17).

Moreover, to better understand the mechanism of the relationship between community size and hockey expertise, we need smaller geographies of players’ hometown communities, such as neighbourhoods or postal code areas. Publicly available data such as www.hockeydb.com do not contain such information and therefore we were unable to further explore the socioeconomic environment of a player during his development.

The findings from this paper can lead to a public discussion on the future of hockey as Canada’s national sport. In this light, it is perhaps fitting to reference the recent tragedy involving the Humboldt Broncos junior hockey team that generated an overwhelming public reaction. The Broncos are a junior ‘A’ hockey team in the Saskatchewan Junior Hockey League (SJHL) based in Humboldt, a small town of 5900 (Statistics Canada, 2017). On 6 April 2018, on their way to a SJHL semi-final game against the Nipawin Hawks, the Bronco’s bus collided with a transport truck at the intersection of rural highways in Armley, Saskatchewan. A total of 16 people on the bus, including 10 male teen players, were killed in the collision. This tragedy immediately led to a national outpouring of support for the team and the deceased family members. The GoFundMe page set up for the families collected CDN$15 million in just 12 days, the largest single such crowdfunding campaign in Canada (The Canadian Press, 2018). April 12 was declared ‘Jersey Day’ and tens of thousands of Canadians wore their favourite hockey jerseys to work and school as a gesture of support for the Broncos (Ghonaim, 2018; Warren, 2018).

It is possible that the tragedy of the Humboldt Broncos drew such overwhelming nation-wide empathy because the team, based in small town Saskatchewan, epitomizes the image of hockey that many Canadians, including those in the media, appear to cherish. However, as our study shows, small town players such as those from Humboldt are increasingly losing their foothold in elite-level hockey, most notably the NHL. When faced with such empirical reality, how would the general public react?

In closing, we raise three questions in the hope of sparking a public dialogue about the future of Canadian hockey as a national sport, First, will hockey’s status as Canada’s sport weaken as the game becomes more ‘big-city’, suburban and perhaps financially out-of-reach for the average family with aspirations of making it to the NHL? Second, will it be necessary to reconstruct the image of hockey as a primarily urban sport representing a departure from the nostalgic notion of the game as an anchor for small-town sporting culture? Third, should youth hockey in small towns be given additional support, through public and private initiatives, as part of a wider effort aimed at the revitalization of the rural economy and infrastructure?

Footnotes

Acknowledgements

An earlier version of this paper was presented at the annual meeting of the Society for the International Hockey Research in Parry Sound, Ontario, 26 May 2018. The authors thank three anonymous reviewers and the journal Editors for helpful comments on previous drafts of this manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

References

Allain

(2011) Kid Crosby or Golden Boy: Sidney Crosby, Canadian national identity, and the policing of hockey masculinity. International Review for the Sociology of Sport 46(1): 3–22.

Atherley

(2006) Sport, localism and social capital in rural Western Australia. Geographical Research 44(4): 348–360.

Baker

Horton

(2004) A review of primary and secondary influences on sport expertise. High Ability Studies 15(2): 211–228.

Baker

Logan

(2007) Developmental contexts and sporting success: Birth date and birthplace effects in National Hockey League draftees 2000–2005. British Journal of Sports Medicine 41(8): 515–517.

Baker

Schorer

Cobley

, et al. (2009) Circumstantial development and athletic excellence: The role of date of birth and birthplace. European Journal of Sport Science 9(6): 329–339.

Baker

Shuiskiy

Schorer

(2014) Does size of one’s community affect likelihood of being drafted into the NHL? Analysis of 25 years of data. Journal of Sports Sciences 32(16): 1570–1575.

Bruner

Macdonald

Pickett

, et al. (2011) Examination of birthplace and birthdate in world junior ice hockey players. Journal of Sports Sciences 29(12): 1337–1344.

Côté

Macdonald

Baker

, et al. (2006) When “where” is more important than “when”: Birthplace and birthdate effects on the achievement of sporting expertise. Journal of Sports Sciences 24(10): 1065–1073.

Curtis

Birch

(1987) Size of community of origin and recruitment to professional and Olympic hockey in North America. Sociology of Sport Journal 4(3): 229–244.

10.

Davis

(1971) Canadian society and history as hinterland versus metropolis. In: Ossenberg

(ed.) Canadian Society: Pluralism, Change and Conflict. Scarborough, ON, Canada: Prentice Hall, pp. 6–32.

11.

Decosse

(2015) New geographies of elite hockey player production in the neoliberal age. MA Thesis, York University, Toronto, ON, Canada.

12.

Dubé

Schinke

Hancock

, et al. (2007) The social support experiences of major junior ice hockey players in a physically removed region of Canada. Journal of Sports Science and Medicine 6(4): 393–400.

13.

Elcombe

(2010) The moral equivalent of “Don Cherry”. Journal of Canadian Studies 44(2): 194–218.

14.

Ferguson

(2014) From coal pits to tar sands: Labour migration between an Atlantic Canadian region and the Athabasca oil sands. Just Labour 17: 18106–18118.

15.

Frisby

Alexander

Taylor

, et al. (2005) Bridging the recreation divide: Listening to youth and parents from low income families across Canada. Report, Canadian Parks and Recreation Association, Ottawa, ON, Canada, 26 September.

16.

Ghonaim

(2018) Londoners sport jerseys to honour the Humboldt Broncos. CBC News, 12 April. Available at: www.cbc.ca/news/canada/london/london-jersey-day-1.4616471 (accessed 15 June 2018).

17.

Gruneau

(2010) Trends in community sport participation and community sport organizations since the 1990s: Implications for West Vancouver. Centre for Sport Policy Studies Working Paper Series No. 3. University of Toronto, Toronto, ON, Canada, February. Available at: https://kpe.utoronto.ca/sites/default/files/CSPS_Working_Paper_3_-_Trends_in_Community_Sport_Participation_and_Community_Sport_Organizations_Since_the_1990s.pdf (accessed 15 June 2018).

18.

Gruneau

Whitson

(1993) Hockey Night in Canada: Sport, Identities and Cultural Politics. Toronto, ON, Canada: Garamond Press.

19.

Hancock

Coutinho

Côté

, et al. (2018) Influences of population size and density on birthplace effects. Journal of Sports Sciences 36(1): 33–38.

20.

Hansen

Muszynski

(1990) Crisis in rural life and crisis in thinking: Directions for critical research. Canadian Review of Sociology and Anthropology 27(1): 1–22.

21.

Holman

(2007) The Canadian hockey player problem: Cultural reckoning and national identities in American collegiate sport, 1947–80. Canadian Historical Review 88(3): 439–468.

22.

Lawson

(1998) Hockey nights: The tough skate through junior-league life. Harper’s Magazine, pp. 65–80, January.

23.

Lidor

Arnon

Maayan

, et al. (2014) Relative age effect and birthplace effect in division 1 female ballgame players – The relevance of sport-specific factors. International Journal of Sport and Exercise Psychology 12(1): 19–33.

24.

Low

Ford

Williams

(2015) The effects of birth date and place on the development of expertise in professional cricket batters. International Journal of Sport Psychology 46(6): 676–688.

25.

Markey

Ryser

Halseth

(2015) “We’re in this all together”: Community impacts of long-distance labour commuting. Rural Society 24(2): 131–153.

26.

Ogden

Edwards

(2016) Are Canadian stakeholders resting on their laurels? A comparative study of the athlete pathway through the Swedish and Canadian male ice hockey systems. Journal of Sport Management 30(3): 312–328.

27.

Oncescu

(2015) Rural restructuring and its impact on community recreation opportunities. Annals of Leisure Research 18(1): 83–104.

28.

Pecoskie

(2016a) Pay to play: All the money goes to the top. The Hamilton Spectator, 29 October. Available at: https://www.thespec.com/sports-story/6936037-pay-to-play-all-the-money-goes-to-the-top-/ (accessed 8 July 2018).

29.

Pecoskie

(2016b) Pay to play: Odds stacked against many young hockey players. The Hamilton Spectator, 28 October. Available at: https://www.thespec.com/news-story/6933956-pay-to-play-odds-stacked-against-many-young-hockey-players/ (accessed 8 July 2018).

30.

Pecoskie

(2016c) ’Pretty Good Rich Kids’: Reaching the OHL takes more than talent. The Hamilton Spectator, 27 October. Available at: https://www.thespec.com/sports-story/6931904–pretty-good-rich-kids-reaching-the-ohl-takes-more-than-talent/ (accessed 8 July 2018).

31.

Rooney

(1969) Up from the mines and out from the prairies: Some geographical implications of football in the united states. Geographical Review 59(4): 471–492.

32.

Rossing

Nielsen

Elbe

, et al. (2016) The role of community in the development of elite handball and football players in Denmark. European Journal of Sport Science 16(2): 237–245.

33.

Schorer

Baker

Lotz

, et al. (2010) Influence of early environmental constraints on achievement motivation in talented young handball players. International Journal of Sport Psychology 41(1): 42–58.

34.

Smailes

(1997) Socio-economic change and rural morale in South Australia, 1982–1993. Journal of Rural Studies 13(1): 19–42.

35.

Solutions Research Group Consultants Inc. (2014) Massive competition in pursuit of the $5.7 billion Canadian youth sports market. Available at: www.srgnet.com/2014/06/10/massive-competition-in-pursuit-of-the-5-7-billion-canadian-youth-sports-market/ (accessed 22 February 2018).

36.

Statistics Canada (2017) Humboldt, CY [Census subdivision], Saskatchewan and Division No. 15, CDR [Census division], Saskatchewan (table). Census profile, 2016 census (Catalogue no. 98-316-X2016001), 29 November. Ottawa, ON, Canada: Statistics Canada. Available at: www12.statcan.gc.ca/census-recensement/2016/dp-pd/prof/index.cfm?Lang=E (accessed 14 June 2018).

37.

The Canadian Press (2018) Humboldt Broncos GoFundMe closes, raises more than $15M. CBC News, 19 April. Available at: www.cbc.ca/news/canada/saskatchewan/humboldt-broncos-gofundme-1.4626413 (accessed 15 June 2018).

38.

Tonts

Atherley

(2005) Rural restructuring and the changing geography of competitive sport. Australian Geographer 36(2): 125–144.

39.

Turnnidge

Hancock

Côté

(2014) The influence of birth date and place of development on youth sport participation. Scandinavian Journal of Medicine & Science in Sports 24(2): 461–468.

40.

Turcotte

(2008) The city/suburb contrast: How can we measure it. Canadian Social Trends 85: 2–19.

41.

Vézina

Houle

(2017) Settlement Patterns and Social Integration of the Population with an Immigrant Background in the Montréal, Toronto and Vancouver Metropolitan Areas: Ethnicity, Language and Immigration Thematic Series (Catalogue no. 89-657-X2016002). Ottawa, ON, Canada: Statistics Canada.

42.

Watson

(2017) Everyday nationalism and international hockey: Contesting Canadian national identity. Nations and Nationalism 23(2): 289–308.

43.

Warren

(2018) ‘A sea of jerseys’: Edmonton shows support for Humboldt Broncos with Jersey Day. The Star, 12 April. Available at: www.thestar.com/edmonton/2018/04/12/a-sea-of-jerseys-edmonton-shows-support-for-humboldt-broncos-with-jersey-day.html (accessed 15 June 2018).