The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1352-7592.htm

TPM
15,3/4 Organizational structure and
home team performance
William M. Foster
158 University of Alberta, Edmonton, Canada, and
Marvin Washington
Van Vliet Centre, University of Alberta, Edmonton, Canada

Abstract
Purpose – The purpose of the paper is to demonstrate that organizational task interdependence has
an impact on the performance of home teams in sport.
Design/methodology/approach – The paper uses a cross-sectional research design. It tests the
authors’ hypothesis using a probit analysis of nine years of data from Major League Baseball and eight
years of data from the National Hockey League.
Findings – The paper determines that the underlying task interdependence of a sport has a
significant impact on the performance of a sport team.
Originality/value – The paper argues that sport managers need to consider organizational structure
when accounting for team performance. Moreover, the structure of the sport(s) needs to be considered
when making adjustments to the league(s) that might affect the competitive balance.
Keywords Team performance, Organizational structures, Sports
Paper type Research paper

1. Introduction and background to research

There is a body of research that has sought to explain team performance (Bloom, 1999;
Gittell, 2000, 2001), sport team performance (e.g. Avrutin and Sommers, 2007; Berman
et al., 2002; Frick and Simmons, 2008; Gee and Wen-Jhan, 2008; Montanari et al., 2008;
Wolfgang and Felicitas, 2008) and, more specifically, why in certain sports the home
team has a significant advantage in terms of their likelihood of winning the game or
contest (e.g. Baumeister, 1995; Baumeister and Steinhilber, 1984; Courneya and Carron,
1990; Gomez et al., 2008; Pace and Carron, 1992; Page and Page, 2007; Ruano et al.,
2007; Sampaio et al., 2008; Schlenker and Phillips, 1995; Smith et al., 2000). Courneya
and Carron (1992) argue that there are a number of variables that can affect home and
visiting team performance (game location, game location factors, psychological states
and behavioural states). Moreover, they demonstrate that the magnitude of the home
advantage varies from sport to sport. While there have been numerous extensions of
this article, research has yet to definitively explain why there is a home advantage and
why it is different between sports.
We argue that despite the comprehensiveness of the Courneya and Carron (1992)
model there is a significant component that has not been included in any discussion of
home team performance: the underlying structure of the sport. In the management and
sport management literature, structure has been cited as a factor that influences and
Team Performance Management affects organizational action and performance (e.g. Chandler, 1962; Kikulis et al., 1995;
Vol. 15 No. 3/4, 2009
pp. 158-171 Miles et al., 1978). With this in mind we argue that the structure of a sport can have a
q Emerald Group Publishing Limited significant effect on the home advantage. More specifically, we will argue that the
1352-7592
DOI 10.1108/13527590910964937 inherent structure of a sport moderates the effect of the home advantage. The way that
the play flows between players to score or prevent runs/goals will significantly affect Home team
how much of an advantage the home team possesses. performance
In the next section we will review the literature that discusses and critiques the
Courneya and Carron model. Next, we will argue that the structure of the sport
moderates the effect of the home advantage, thus producing differences in the
magnitude of the home advantage between sports. We then test our assertion that a
sport’s structure moderates the effect of the home advantage. To test our hypotheses 159
we examine the game-by-game performance of professional baseball and hockey
teams. We examine nine years of data from Major League Baseball and eight years of
data for the National Hockey League. We use probit regression analysis to examine the
data. Finally we discuss our results and the implication of our study.
This paper makes three significant contributions to the literature on home advantage.
The first is that we identify practical implications that our findings have for sport
managers. These implications are of particular importance when the sport is compared
to other industries. Plunkett Research Ltd estimated that that the entire US sport
industry was worth $441.1 billion in 2008. This makes the sport industry in the USA
larger than both the US auto industry and the US film industry (Plunkett Research,
2008). Nevertheless, over the past ten years sport leagues have made significant changes
to their games, many without considering how these changes might affect the sport and
its competitive balance. We argue that an important responsibility of leagues and
governing sport bodies is to understand how changes to a sport can significantly affect
the sport’s competitive balance because of the way it is structured.
Our second contribution is a discussion of the implication our findings have for
team and organizational performance. More specifically, we argue that there are many
situations where teams and organizations, outside a sporting context, have a home
field advantage. This home advantage is a valuable resource and a source of a
competitive advantage if properly identified and leveraged (Barney, 1991). On the other
hand, a poorly managed home advantage can, possibly, lead to a home team
disadvantage (e.g. Loignon et al., 2007). Our third contribution is that we make a
significant addition to the Cournyea and Carron model for explaining why there is a
difference in the magnitude of the home advantage across sports.

2. The home advantage

The research on home advantage has two main tenets that are almost unassailable.
The first is that the home team in sport has an advantage over the visiting team. The
second is that the home advantage is fundamentally different in different sports
(Courneya and Carron, 1992). These fundamental tenets have been the starting points
for all research on the home advantage.
Courneya and Carron (1992) conducted one of the first reviews of the home
advantage. Their main contribution was the creation of a model that identifies the
factors that “moderate the degree of home advantage” (Courneya and Carron, 1992,
p. 18) and explains the reasons “why the home advantage exists” (Courneya and
Carron, 1992, p. 22). The factors that moderate the home advantage are the home
crowd, learning effects (e.g. the home team’s familiarity with the venue), the effects of
travel on a visiting team, and rules that give the home team certain advantages (e.g.
last change in ice hockey, last bat in baseball). The reasons that were identified as
explaining why the home advantage exists are the critical psychological and
TPM behavioural states of the officials, coaches and players involved in the contest.
15,3/4 Although there is an extensive literature on the psychological, behavioural and
physiological reasons for the home advantage, we will focus our attention on the
reasons for the difference in degree in the home advantage. Other scholars have
elaborated on the home advantage literature by examining the impact of travel
distance (Clarke and Norman, 1995), batting order (Bray et al., 2005), and crowd size
160 (Nevill and Holder, 1999). Still some scholars have even argued that there is a
significant home field disadvantage in sport (Baumeister and Steinhilber, 1984).
There is also conflicting evidence about the impact that familiarity with a particular
location has on the home advantage. Pollard (2002) demonstrates that when NHL and
NBA teams move to a new building during the season there is a significant drop in the
home advantage. This indicates that the location where a team plays can have the effect
of reducing the home advantage. Nevertheless, there are other studies that have shown
that the home advantage is not location-specific. Moore and Brylinsky (1995) studied
Western Michigan University men’s and women’s basketball teams while they had to
play home games in a variety of different locations. They found that the team still had a
home advantage despite having to play home games in five different locations
throughout the season. Moreover, both teams had a stronger home advantage than their
counterparts in the conference. This led Moore and Brylinsky (1995) to conclude that
location is a factor that does not significantly contribute to a home advantage.
The Courneya and Carron model and its variations have painted a picture of the
home advantage that has been unchanged for over a decade. What has not been asked
is whether there are other factors that could be responsible for the home advantage. In
the next section we make a case for the inclusion of another variable into the model:
organizational structure.

3. Organizational structure
There are a number of ways that organizational structure has been discussed in the
management and sport management literature (e.g. Chandler, 1962; Donaldson, 1996).
One aspect of structure that is important for our research is task interdependence
(Thompson, 1967). The definition of task interdependence to be used throughout is as
follows: task interdependence refers to the manner that work flows between an
organization’s work units (Van de Ven and Ferry, 1980, p. 166). Work flow between
work units is conceptualized as the transfer of work activities (organizational inputs
and outputs) between individuals, groups, teams, and/or organizations in order to
achieve a common organizational goal. This definition best reflects how the structure
of an organization impacts an organization’s actions.
The four most common types of task interdependence are pooled, sequential,
reciprocal (Thompson, 1967) and team (Van de Ven et al., 1976). As organizations
become more complex so does the task interdependence of the organization. The types
differ in the manner in which work flows to different work units in the organization. As
an organization’s workflow becomes more complex, the coordination of organizational
work units becomes more difficult (Wageman, 2001). Moreover, evidence indicates that
the way that work flows between different organizational work units affects a firm’s
actions and behaviours (Thompson, 1967). As an organization’s workflow becomes
increasingly interconnected, the mechanisms needed for coordination and control also
increase in complexity (Thompson, 1967; Van de Ven et al., 1976).
 It has been argued that sport is an excellent site to observe the different forms of Home team
task interdependence (Wolfe et al., 2005). For example, batting in Major League performance
Baseball (MLB) is an excellent example of pooled task interdependence. When a
baseball team is on offense the team’s structure is akin to a simple organization that is
organized using pooled interdependence (e.g. Keidel, 1984, 1987). When a player comes
to bat he[1] is independent of his teammates. He alone faces the pitcher. At bat the
player is a work unit. The batter, as an organizational work unit, acts independently to 161
produce a common output, i.e. runs scored. The batter’s output is not contingent on the
other players on the team for any component of the work. Communication and
coordination between teammates is unnecessary because each batter is independent.
Two sports that are reciprocally and team interdependent are basketball and hockey.
Although the sport of hockey is not commonly referred to as an example of reciprocal or
team interdependence, this does not mean that it is not a good example of these types of
task interdependence. Hockey and basketball are similar in the way the work flows
between work units. The players on each team are work units. In both sports the players
are dependent on their teammates’ actions for their inputs so they can produce outputs.
The flow of the work between the players on the ice and on the court is such that the
outputs of one player become the inputs of another. This occurs either immediately, with
team interdependence, or after a period of time, with reciprocal interdependence. Keidel
(1984, 1987) argues that basketball is an example of reciprocal interdependence and Van
de Ven et al. (1976) argue that basketball is an example of team interdependence.
Basketball and hockey are similar in that players in both hockey and basketball play
offense and defense and there are few stoppages of play.
We argue that the structure of a sport will have a significant effect on the home
advantage. For example, baseball players require only limited communication and
coordination to produce outputs (runs or out). As a result, we anticipate that the home
advantage will be less pronounced because the poor performance of one player is not likely
to significantly impact the performance of the team. In contrast, to produce outputs hockey
players need to constantly communicate and coordinate with their teammates. As a result,
we would anticipate that the poor performance of one player will have more of an effect on
the performance of the entire team. Therefore, the research question we are testing is
whether the home advantage will be weaker in baseball (a pooled and sequentially
interdependent sport) than it is in hockey (a reciprocal and team interdependent sport).

4. Methodology
A cross-sectional research design (De Vaus, 2001) is employed to examine the effects of
structure on the home advantage. There are a number of reasons why this approach is
most appropriate for this investigation. First, cross-sectional designs are particularly
appropriate for measuring differences between groups. The research question that is
asked concerns the differences between two different sports. Second, cross-sectional
designs are used to examine existing differences that exist in the sample.
A weakness of a cross-sectional research design is that it is difficult to determine if
there is a causal link between different variables. The goal in a cross-sectional design is
to explain the relations between the dependent variable and one or more independent
variables. This is often difficult to do because of the nature of the research design.
We employ multivariate probit[2] regression models to test the association between
the variables in the models. The use of multivariate statistical models, with statistical
TPM control variables, helps “remove the confounding effects of a set of variables at once
15,3/4 and focus on the ‘pure’ effects of an independent variable” (De Vaus, 2001, p. 211).

4.1 Sample
4.1.1 Major League Baseball. Major League Baseball has a large amount of statistical
information about the league, its teams, and its players. The baseball data for this
162 investigation are collected from “Baseball Oracle 2005”, an online baseball database. The
database comprises statistics for all players and teams starting in 1871 and ending in
2004. The reliability and accuracy of the data are confirmed with other data sources,
specifically the Baseball Almanac (Schlossberg, 2002). Other data include player salaries.
The data for player salaries are also from the “Baseball Oracle 2005” database. This
information is verified using Rodney Fort’s online Sports Economics Data and
Bibliography. The MLB study period begins with the individual game-by-game results
and payroll information from the 1996 season. The study period ends with the individual
game-by-game results and payroll information from the 2004 season.
4.1.2 The National Hockey League. The NHL also has large quantity of statistical
information that provides insight into the operation and performance of the league, its
teams, and its players. Most of the hockey data are collected from the Hockey Research
Association. This organization operates a web site (see www.hockeyresearch.com/
stats/index.phtml) that has a comprehensive archive of 11 years of team and player
data starting with the 1988-1989 season and ending with the 1998-1999 season. Data
are also collected from The Hockey Summary Project (see www.shrpsports.com/hsp/),
the NHL’s web site (see www.nhl.com), and Hockeydb.com (see www.hockeydb.com).
The accuracy of the data is verified by consulting Total NHL (Diamond, 2003). The
necessary team payroll information is from The Hockey News. The Hockey News
publishes an annual list of team payrolls and player salaries. The NHL study period
begins with the individual game-by-game results and team payroll information from
the 1996/1997 season. The study period ends with the individual game-by-game results
and team payroll information from the 2003/2004 season.

4.2 Variables
4.2.1 Dependent variable. Game-by-game wins or losses is the dependent variable of this
study. This measure is an unambiguous way of determining if one team has an
advantage over the other. Seasonal measures of performance such as total seasonal wins
or team winning percentage have been traditionally used as the independent variable in
sport and management studies (e.g. Courneya and Carron, 1992; Smart and Wolfe, 2003).
Game-by-game wins or losses is used as the dependent variable of this study because it
is, arguably, the best way to measure the effects of task interdependence on team
performance. Each player works intra-game to produce either wins or losses. Because of
this, aggregate measures of team performance such as win percentage or total wins do
not capture the intra-game, moderating effect of task interdependence. A home team is
assigned a 1 for a win and a 0 for a loss. In hockey, tied games are removed from the
sample to eliminate possible confusion as to the outcome of each game. To operationalize
the dependent variable we use home team wins and losses.
4.2.2 Independent variable. A variable that affects a team’s performance is the
location of the game. A number of studies indicate that the home team has an
advantage (e.g. Courneya and Carron, 1992; Schwartz and Barsky, 1977; Smith et al.,
2000) or a disadvantage (e.g. Baumeister and Steinhilber, 1984; Wright and Voyer, Home team
1995). The home team advantage or disadvantage factor is controlled for through the performance
design of the study because only home team results are used as the measure of team
performance. This independent variable measure is operationalized as the intercept of
the regression model for both hockey and baseball.
4.2.3 Control variables – player quality. The moderating effect of team quality on the
home advantage has been tested. It was determined that stronger teams have more of a 163
home advantage than weaker teams (Madrigal and James, 1999). One way to test the
strength of a team is to measure the quality of the players on the team. This is because
some athletes, especially highly skilled ones, have the ability to produce team wins
without the help of other players. There are instances where one player has had a
significant effect on the outcome of a game.
When these athletes are under contract to a team, they cannot work for another team.
Each MLB and NHL team has a measure of the amount of money spent on player
contracts. The amount of a player’s salary is an indication of his past and future
productivity. Players who are, or have been, more productive are paid higher salaries than
players who are, or have been, less productive. The team that signs more high-quality
players will have a higher team payroll than a team with fewer high-calibre players.
There are some limitations to this measure. The first is that intra-season roster
changes make it difficult to accurately aggregate team payroll on a game-by-game
basis. If a player is added to a team’s roster through a trade or free agency, it is difficult
to know how the new player’s contract should be tabulated in the team’s payroll. It is
for this reason that payroll is measured at the start of each season.
Another limitation of this measure is in the data. The payroll data for both the NHL
and MLB do not include player bonuses. Some players do not have bonuses in their
contracts. Others do not reach bonus targets. Nevertheless, there are times when a
player’s base salary is underrepresented. Despite these limitations we believe salary is
a strong proxy for player quality.
A strength of the player salary measure is that it places a value on player abilities that
are not easily quantifiable. For example, player productivity is usually measured by their
offensive or defensive contributions (Smart and Wolfe, 2003). Yet the productivity of a
player is not always reflected by his statistical contributions. Often players are paid for
the “intangible” qualities that they bring to the team. These intangible qualities are best
measured in the way a team values the quality of its players. To operationalize player
quality we use a team’s payroll in US dollars at the start of the season.
Another variable common to both sports is the number of all-star players on the
team’s roster at the start of the current season. All-stars are the players who performed
at the highest level during the first half of the season. The fans of the NHL and MLB
vote for the starting line-ups of the all-star games[3]. The league then selects the
remainder of the team on the basis of player performance. Regardless of fan practices,
all-star players are, for the most part, those individuals whose play has been better
than their peers for the first half of the season. The number of all-stars a team employs
can affect a team’s performance. First, all-stars are players who have performed at a
high level the previous season. There is the expectation that past performance will be
duplicated in future seasons. In addition, the influence of all-stars on other players
cannot be underestimated. All-star players can provide their teammates with
invaluable experience and knowledge. To operationalize this measure we calculated
TPM the percentage of a team’s players who, at the start of the season, had played in the
15,3/4 previous year’s all-star game.
4.2.4 Control variable – learning effects. A sport team with less year-to-year turnover is
likely to have developed more complementarities between the players. Over time, players
identify the strengths and weaknesses of their teammates. As players work together they
also recognize that each player has specific tendencies. When the players on the team
164 identify these patterns and routines they can adjust their actions to complement their
teammate’s actions. A team with more players who complement each other will likely win
more games than its competitors. To operationalize this control variable we measure the
percentage of new players on a team’s roster at the start of the season.
There are some limitations to the turnover measure, however. The first limitation is
that there is an assumption that complementarities will develop because teams have
little turnover. A measure of turnover does not take into consideration which players
are leaving the team. A team with a large amount of turnover could merely be losing
fringe players. The loss of these players is not likely to significantly affect team
performance. On the other hand, a team that loses players that are at the core of the
team that has been together for a substantial period of time could see a significant
decline in team performance.
The chemistry that develops between players is a learning effect. This learning
effect could have an effect on team performance. A team that has greater synergy
amongst its core players will produce more wins. A team with less synergy amongst its
core players will produce fewer wins. To operationalize this variable we measure the
percentage of players whose tenure with the team is three consecutive years or less at
the start of the season.
A three-year cut off point is used to distinguish between a team’s core players and
the team’s fringe contributors. A core contributor is a player who the coach and GM
identify as a significant part of the team. A player who is not a core contributor to the
team will have a shorter tenure than a player who is a substantial contributor to the
team’s performance.
In the NHL, three years is used as the cut-off point for a number of reasons. The first
is that the length of a NHL rookie contract is three years. As of 1994, a player who
enters the NHL signs a rookie contract. This contract covers the first three seasons of a
player’s career[4]. Although there are some exceptions to this rule, the majority of
players in the NHL signed a three-year rookie contract. A player who is not a core
contributor to a NHL team is not re-signed after his first three seasons with the team.
A second reason for the three-year cut off is that three years is the approximate
average tenure of a NHL player. The average length of a player’s career in the NHL
ranges from three to seven years. Any player who is on a team for more than three
years is a significant team asset.
In MLB the three-year cut-off is used because a player is eligible for arbitration rights
after three seasons of service. Although a player does not have unrestricted free-agent
rights after this period of time, a team will walk away from any salary arbitration
decision that is unreasonable compared to the player’s performance. A player who is not
with a team after three years is not an asset that the team values highly.
Another reason for the three-year cut-off is that the average career length in MLB is
5.6 years for batters and 4.8 years for pitchers (Schall and Smith, 2000). The mid-point
of these career lengths is 2.6 and 2.4 years, respectively. Because it is difficult to
measure half seasons, the average of the two midpoints is rounded up to three years. Home team
This provides an indication that the average player will play close to three years with a performance
team. If a player plays longer than three years with one team he is an above average
player and, thus, a core contributor to the team.

5. Data and analysis

We use two separate models to test the effects of the home advantage because of the 165
strong correlations between the player quality measures (team payroll and all stars) and
the learning effects measures (turnover and chemistry). The home advantage is
represented by the intercept of both models. A hockey dummy variable and a baseball
dummy variable are included in the models to test the differences between the two sports.
A probit analysis is used to test the differential effects of the home advantage in
baseball and hockey. The sample size is 29,447 data points. The data set is comprised
of 21,510 data points from baseball and 7,937 data points from hockey. The payroll
data points in the analysis are the relative comparison of the home team’s payroll
divided by the away team’s payroll for each game. The turnover, all stars and
chemistry data points are absolute comparisons. For example, the data points used for
the turnover measure is the turnover on the away team from the previous year
subtracted from the turnover of the home team from the previous year.

6. Results
Tables I and II present the descriptive statistics and correlations for hockey and
baseball, respectively. The correlation matrices provide evidence that player quality
variables (payroll and turnover) and learning effects variables (turnover and
chemistry) are both correlated with wins in both hockey and baseball[5]. These results
do not provide support to our research question; however, these results demonstrate
that these measures are good variables to include as controls in both models.

Variables Mean SD 1 2 3 4

1. Wins 35.52 7.91

2. Turnover 0.39 0.15 2 0.40 *
3. Payroll 3.40E þ 07 1.28E þ 07 0.36 * 2 0.17 *
4. All stars 1.56 1.16 0.47 * 2 0.42 * 0.42 *
5. Chemistry 0.74 0.13 2 0.41 * 0.61 * 2 0.27 * 2 0.43 * Table I.
Hockey descriptive
Notes: *p , 0:01; p , 0:05 (two-tailed tests) statistics and correlations

Variables Mean SD 1 2 3 4

1. Wins 80.93 12.23

2. Payroll 5.53E þ 07 2.57E þ 07 0.39 *
3. Turnover 0.45 0.12 20.31 * 20.04
4. All stars 2.26 0.10 0.61 * 0.43 * 20.14 * *
5. Chemistry 0.78 1.57 20.31 * 20.18 * 0.55 * 2 0.20 * Table II.
Baseball descriptive
Notes: *p , 0:01; * *p , 0:05 (two-tailed tests) statistics and correlations
TPM Table III presents evidence that there is a significant difference between the home
15,3/4 advantage in hockey and the home advantage in baseball. In Model 1 the results from
the base model are displayed. These results confirm that there is a home advantage in
sport (b ¼ 0:11, p , 0:001). As well, the results show that player quality (b ¼ 0:20,
p , 0:001) and learning effects (b ¼ 20:57, p , 0:001) have an effect on winning in
baseball and hockey. The results in Model 2 indicate that there is a significant home
166 advantage for Major League Baseball teams (b ¼ 0:09, p , 0:001) and that player
quality (b ¼ 0:18, p , 0:001) and learning effects (b ¼ 20:48, p , 0:001) impact team
performance. Similarly, the results in Model 3 demonstrate that wins in the National
Hockey League are affected by the home advantage (b ¼ 0:16, p , 0:001), player
quality (b ¼ 0:29, p , 0:001) and learning effects (b ¼ 20:66, p , 0:001).
What is interesting, however, are the results when the home advantage of the two
sports is compared. Not only is the home advantage in baseball less than the base model, it
is also significantly less than the home advantage in hockey (b ¼ 0:07, p , 0:001). This
result shows that when baseball and hockey are compared a home team in hockey has
more of a home advantage than a home team in baseball. From these results we can
conclude that teams in sports that are pooled or sequentially interdependent have a smaller
home advantage than teams in sports that are reciprocally or team interdependent.
Table IV presents the results of the alternate model. The results once again show
that there is a home advantage in both sports (b ¼ 0:11, p , 0:001). Moreover, the
results also show that the difference in the home advantage between the two sports

Model 1: Model 2: Model 3:

Variable MLB and NHL MLB NHL Difference

Home advantage 0.11 * * 0.01 0.09 * * 0.01 0.16 * * 0.01 0.07 * * 0.02
Team payroll 0.20 * * 0.01 0.18 * * 0.02 0.29 * * 0.04 0.11 * * 0.04
Turnover 20.57 * * 0.04 20.48 * * 0.06 20.66 * * 0.08 20.17 * 0.10
Observations 29,447 21,510 7,937
Table III. x2 464.92 263.70 217.53
Probit models testing the Log-likelihood 220,066.44 214,721.42 25,327.92
difference between the
home advantage in Major Notes: The dependent variable in all cases is home team game-by-game wins or losses.
League Baseball and the Unstandardized regression coefficients are shown with standard errors to the right. *p , 0:1;
National Hockey League * *p , 0:001

Model 1: Model 2: Model 3:

Variable MLB and NHL MLB NHL Difference

Home advantage 0.11 * 0.01 0.09 * 0.01 0.16 * 0.01 0.07 * 0.02
All stars 0.06 * 0.00 0.04 * 0.00 0.10 * 0.01 0.06 * 0.01
Table IV. Chemistry 2 0.03 0.05 2 0.44 * 0.07 0.38 * 0.08 0.82 * 0.10
Alternate probit models Observations 29,447 21,510 7,937
testing the difference x2 293.17 215.32 182.73
between the home Log-likelihood 2 20,125.32 214,745.61 25,345.33
advantage in Major
League Baseball and the Notes: The dependent variable in all cases is home team game-by-game wins or losses.
National Hockey League Unstandardized regression coefficients are shown with standard errors to the right. * p , 0:001
remains the same when other player quality and learning effect variables are included Home team
as controls in the model. These results demonstrate that hockey teams have a performance
significant home advantage when compared to baseball teams.
From our results we can conclude that the home advantage is significant in both
simple and complex sports. There is also a significant difference in the magnitude of
the home advantage in simple and complex sports. Finally, the results show that
factors other than the home advantage, such as player quality and the development of 167
player complementarities, play a role in the production of a team’s wins or losses.

7. Discussion
The results of our study demonstrate that the inherent structure of a sport can have a
significant effect on home team performance. This finding is important for two
reasons. The first is that practically, these results suggest that the structure of a sport
may have an effect on the competitive balance of different leagues. For example, pooled
or sequentially interdependent sports may want to schedule longer road trips for teams
because travel effects tend to have less impact on team performance. Similarly,
reciprocally and team-interdependent sports may want to look at more back-to-back
games in one location such as what the NHL implemented in the 2005/2006 season. The
effect of these changes might be to increase the competitiveness of the visiting team.
This would increase the uncertainty that surrounds the outcome of the game. Ideally,
increasing the indeterminacy of the game outcome would increase interest in the game
and increase attendance and, in certain leagues, increase team revenues.
Another practical implication of our study is that changes in the rules of the game
could have significant impacts on the home advantage because of the structure of the
sport. The most obvious example is the change that the NHL made to decide games by
shootouts if the game was tied after a five-minute overtime period. When the NHL
implemented shootouts they were effectively changing the structure of the game. A
sport that is reciprocally and team-interdependent has been altered to include a pooled
interdependent facet to the game. Because of this change we would expect that the home
advantage in the National Hockey League will drop to a magnitude that more closely
resembles that of baseball. Ideally this change will increase the outcome indeterminacy
of the games, which will increase the competitive balance across the league.
Another practical consideration for sport teams is the effect the home advantage has
on playoff series. As we demonstrated, the home advantage is less for simple
organizations. Nevertheless there is still a significant advantage. In MLB this advantage
has been decided by the outcome of the All-Star game. We would argue that the
importance of the home advantage is large enough that the outcome of the World Series
can be decided because of this advantage. To advantage the teams of one league over
another because of an exhibition game is an example of a league decision that could have
significant negative long-term performance effects. Similarly, there has been discussion
of MLB playing the World Series at a neutral location. This decision could help to reduce
the home advantage and thus, increase the competitive balance of the series.
Our study demonstrates that the task interdependence of a sport moderates the home
advantage. Sports that are team or reciprocally interdependent will have larger home
advantages than sports that are pooled or sequentially interdependent. These results
suggest that the underlying structure of a sport has a significant impact on how a game
is played and the effect it can have on the competitive balance within a league.
TPM Consequently, sport bodies and league governors have to be judicious when making
15,3/4 changes because of the potential to disrupt the competitive balance in the sport or the
league.
Another practical implication is that there is possibly a home advantage for teams
and organizations outside the sport context. The way that a team’s work flows could
significantly impact different team decisions especially when firms relocate or expand.
168 For example, when a firm expands its operations staffing and hiring decision should be
treated differently in simple and complex organizations. A team that is pooled or
sequentially interdependent may find that staffing will have less impact on performance
because new team members are less reliant on each other and their skills are relatively
fungible. In contrast, complex organizations that expand will need to consider the
specialized and interdependent nature of their work flow. In complex organization this
will mean that hiring decisions need to be considered carefully because of the highly
interdependent nature of the tasks and the inimitability of team members.
Another insight for teams outside the sporting context applies to the local
knowledge needed in local contexts. In simple organizations the need for local
knowledge is limited because of the nature of the work. In teams that are pooled or
sequentially interdependent tasks are simple and the knowledge of the local
environment usually does not impact performance. For a complex team, however, local
knowledge could be a significant advantage because the work flow of the work units is
interdependent. Thus, if a complex team is operating in a way that is not consistent
with the local modus operandi this can lead to significant negative performance.
Our findings also demonstrate that the accepted model of the home advantage is
incomplete. The results from our study demonstrate that there is a significant
difference in the home advantage between sports that are pooled or sequentially
interdependent and sports that are team and reciprocally interdependent. In sports
where the players act as independent or semi-autonomous work units the home
advantage is less significant. This result can also be seen in other studies that have
investigated the home advantage. For instance, Carron et al. (2005) demonstrate that
there is a difference in the magnitude between different sports with home teams in
baseball having the smallest home advantage while soccer teams have the highest.
Courneya and Carron (1992) also reported findings that demonstrate that baseball
teams have the smallest home advantage while soccer teams have the largest. Pollard
(2006) also reports similar results.
We argue that these findings indicate that there are at least two types of home
advantage. There is a home advantage for organizations that have pooled or sequential
task interdependence, which is relatively small, and a home advantage for
organizations that have reciprocal or team interdependence, which is significantly
larger. The implication of this finding is that more research should follow the lead of
Pollard (2006) and Nevill and Holder (1999) by investigating the differences in the home
advantage within the same sporting context. In so doing, the reasons for the home
advantage within a sport may become better defined.

8. Conclusion
In this paper we demonstrate that the performance of sport teams at home is
moderated by the inherent structure of the sport. In so doing we have shown that
performance in sport is greatly affected by the task interdependence of the players on
the field or on the ice. The implication is that teams and leagues have to understand Home team
that factors other than the quality of the players, coaching and leadership can performance
significantly impact the performance of a team and, thus, the competitive balance of
the league.

Notes
169
1. Gender-specific pronouns will be used when discussing MLB and the NHL because only men
participate in these leagues.
2. Although probit models were used in the investigation, logit models were also tested. It was
found that the differences in the results were negligible.
3. This can lead to some bias as to which players are selected to the team because of ballot box
stuffing.
4. This rule applies to players who are between 18 years old and 25 years old. Players who are
older than this when they sign their first contracts are not subject to this rule.
5. Wins were used to calculate the correlation coefficients with the other variables because the
independent variable (game-by-game wins) is not conducive to correlation calculations.

References
Avrutin, B.M. and Sommers, P.M. (2007), “Work incentives and salary distributions in Major
League Baseball”, Atlantic Economic Journal, Vol. 35 No. 4, pp. 509-10.
Barney, J.B. (1991), “Firm resources and sustained competitive advantage”, Journal of
Management, Vol. 17 No. 1, pp. 99-120.
Baumeister, R.F. (1995), “Disputing the effects of championship pressures and home audiences”,
Journal of Personality and Social Psychology, Vol. 68 No. 4, pp. 644-8.
Baumeister, R.F. and Steinhilber, A. (1984), “Paradoxical effects of supportive audiences on
performance under pressure: the home field disadvantages in sports championships”,
Journal of Personality and Social Psychology, Vol. 47 No. 1, pp. 85-93.
Berman, S.L., Down, J. and Hill, C.W.L. (2002), “Tacit knowledge as a source of competitive
advantage in the National Basketball Association”, Academy of Management Journal,
Vol. 45 No. 1, pp. 13-32.
Bloom, M. (1999), “The performance effects of pay dispersion on individuals and organizations”,
Academy of Management Journal, Vol. 42 No. 1, pp. 25-40.
Bray, S.R., Obara, J. and Kwan, M. (2005), “Batting last as a home advantage factor in men’s
NCAA tournament baseball”, Journal of Sport Sciences, Vol. 23 No. 7, pp. 681-6.
Carron, A.V., Loughhead, T.M. and Bray, S.R. (2005), “The home advantage in sport
competitions: Courneya and Carron’s (1992) conceptual framework a decade later”, Journal
of Sports Sciences, Vol. 23 No. 4, pp. 395-7.
Chandler, A.D. (1962), Strategy and Structure: Chapters in the History of the Industrial Enterprise,
MIT Press, Cambridge, MA.
Clarke, S.R. and Norman, J.M. (1995), “Home ground advantage of individual clubs in English
soccer”, The Statistician, Vol. 44, pp. 509-21.
Courneya, K.S. and Carron, A.V. (1990), “Batting first versus last: implications for the home
advantage”, Journal of Sport and Exercise Psychology, Vol. 12, pp. 312-6.
Courneya, K.S. and Carron, A.V. (1992), “The home advantage in sport competitions: a literature
review”, Journal of Sport and Exercise Psychology, Vol. 14, pp. 13-27.
TPM De Vaus, D.A. (2001), Research Design in Social Research, Sage Publications, Thousand Oaks,
CA.
15,3/4
Diamond, D. (Ed.) (2003), Total NHL, Dan Diamond and Associates, Toronto.
Donaldson, L. (1996), “The normal science of structural contingency theory”, in Clegg, S.R.,
Hardy, C. and Nord, W.R. (Eds), Handbook of Organization Studies, Sage Publications,
London, pp. 57-76.
170 Frick, B. and Simmons, R. (2008), “The impact of managerial quality on organizational
performance: evidence from German soccer”, Managerial & Decision Economics, Vol. 29
No. 7, pp. 593-600.
Gee, S. and Wen-Jhan, J. (2008), “Wage dispersion and team performance: evidence from the small
size professional baseball league in Taiwan”, Applied Economics Letters, Vol. 15 No. 11,
pp. 883-6.
Gittell, J.H. (2000), “Organizing work to support relational coordination”, The International
Journal of Human Resource Management, Vol. 11 No. 3, pp. 517-39.
Gittell, J.H. (2001), “Supervisory span, relational coordination and flight departure performance: a
reassessment of postbureaucracy theory”, Organization Science, Vol. 12 No. 4, pp. 468-83.
Gomez, M.A., Lorenzo, A., Barakat, R., Ortega, E. and Palao, J.M. (2008), “Differences in
game-related statistics of basketball performance by game location for men’s winning and
losing teams”, Perceptual and Motor Skills, Vol. 106 No. 1, pp. 43-50.
Keidel, R.W. (1984), “Baseball, football, and basketball: models for business”, Organizational
Dynamics, Vol. 12 No. 3, pp. 5-18.
Keidel, R.W. (1987), “Team sports model as a generic organizational framework”, Human
Relations, Vol. 40 No. 9, pp. 591-612.
Kikulis, L.M., Slack, T. and Hinings, C.R. (1995), “Sector-specific patterns of organizational
design change”, Journal of Management Studies, Vol. 32 No. 1, pp. 67-100.
Loignon, A., Gayton, W.F., Brown, M., Steinroeder, W. and Johnson, C. (2007), “Home
disadvantage in professional ice hockey”, Perceptual and Motor Skills, Vol. 104 No. 3,
pp. 1262-4.
Madrigal, R. and James, J. (1999), “Team quality and the home advantage”, Journal of Sport
Behavior, Vol. 22 No. 3, pp. 381-98.
Miles, R.E., Snow, C.C., Meyer, A.D. and Coleman, H.J. Jr (1978), “Organizational strategy,
structure and process”, Academy of Management Review, Vol. 2, pp. 546-62.
Montanari, F., Silvestri, G. and Gallo, E. (2008), “Team performance between change and
stability: the case of the Italian ‘Serie A’”, Journal of Sport Management, Vol. 22 No. 6,
pp. 701-16.
Moore, J.C. and Brylinsky, J. (1995), “Facility familiarity and the home advantage”, Journal of
Sport Behavior, Vol. 18 No. 4, pp. 302-11.
Nevill, A.M. and Holder, R.L. (1999), “Home advantage in sport”, Sports Medicine, Vol. 28 No. 4,
pp. 221-36.
Pace, A. and Carron, A.V. (1992), “Travel and the home advantage”, Canadian Journal of Sport
Sciences, Vol. 17 No. 1, pp. 60-4.
Page, L. and Page, K. (2007), “The second leg home advantage: evidence from European football
cup competitions”, Journal of Sports Sciences, Vol. 25 No. 14, pp. 1547-56.
Plunkett Research (2008), “Sports industry overview”, available at: www.plunkettresearch.com/
Industries/Sports/SportsStatistics/tabid/273/Default.aspx (accessed December 11, 2008).
Pollard, R. (2002), “Evidence of a reduced home advantage when a team moves to a new Home team
stadium”, Journal of Sports Sciences, Vol. 20, pp. 969-73.
Pollard, R. (2006), “Home advantage in soccer: variations in its magnitude and a literature review
performance
of the inter-related factors associated with its existence”, Journal of Sport Behavior, Vol. 29
No. 2, pp. 169-89.
Ruano, M.A.G., Calvo, A.L., Toro, E.O. and Zafra, A.O. (2007), “Differences in the performance
indicators of winning and losing women’s basketball teams during home/away games”, 171
Revista De Psicologia Del Deporte, Vol. 16 No. 1, pp. 41-54.
Sampaio, J., Ibanez, S.J., Gomez, M.A., Lorenzo, A. and Ortega, E. (2008), “Game location
influences basketball players’ performance across playing positions”, International
Journal of Sport Psychology, Vol. 39 No. 3, pp. 205-16.
Schall, T. and Smith, G. (2000), “Career trajectories in baseball”, Chance, Fall, pp. 35-8.
Schlenker, B.R. and Phillips, S.T. (1995), “Championing pressures: choking of triumphing in one’s
own territory?”, Journal of Personality and Social Psychology, Vol. 68 No. 4, pp. 632-43.
Schlossberg, D. (2002), The Baseball Almanac, Triumph Books, Chicago, IL.
Schwartz, B. and Barsky, S.F. (1977), “The home advantage”, Social Forces, Vol. 55 No. 3,
pp. 641-61.
Smart, D.L. and Wolfe, R.A. (2003), “The contribution of leadership and human resources to
organizational success: an empirical assessment of performance in Major League
Baseball”, European Sport Management Quarterly, Vol. 3 No. 3, pp. 165-88.
Smith, D.R., Ciacciarelli, A., Serizan, J. and Lamber, D. (2000), “Travel and the home advantage”,
Sociology of Sport Journal, Vol. 17, pp. 364-85.
Thompson, J.D. (1967), Organizations in Action: Social Science Bases of Administrative Theory,
McGraw-Hill, New York, NY.
Van de Ven, A.H., Delbecq, A.L. and Koenig, R. Jr (1976), “Determinants of coordination modes
within organizations”, American Sociological Review, Vol. 41, pp. 322-38.
Van de Ven, A.H. and Ferry, D.L. (1980), Measuring and Assessing Organizations, Wiley, New
York, NY.
Wageman, R. (2001), “The meaning of interdependence”, in Turner, M.E. (Ed.), Groups at Work:
Theory and Research, Lawrence Erlbaum Associates, Mahwah, NJ, pp. 197-217.
Wolfe, R.A., Weick, K.E., Usher, J.M., Terborg, J.R., Poppo, L., Murrell, A.J., Dukerich, J.M.,
Core, D.C., Dickson, K.E. and Jourdan, J.S. (2005), “Sport and organizational studies:
exploring synergy”, Journal of Management Inquiry, Vol. 14 No. 2, pp. 182-210.
Wolfgang, J. and Felicitas, M. (2008), “Navigating toward team success”, Team Performance
Management, Vol. 14 Nos 1/2, pp. 102-8.
Wright, E.F. and Voyer, D. (1995), “Supporting audiences and performance under pressure: the
home-ice disadvantage in hockey championships”, Journal of Sport Behavior, Vol. 18 No. 1,
pp. 21-8.

Corresponding author
William Foster can be contacted at: [email protected]

To purchase reprints of this article please e-mail: [email protected]

Or visit our web site for further details: www.emeraldinsight.com/reprints