Content uploaded by Miguel Ángel Gómez Ruano
Author content
All content in this area was uploaded by Miguel Ángel Gómez Ruano on Feb 25, 2014
Content may be subject to copyright.
1
Components of home advantage in 157 national football leagues worldwide
RICHARD POLLARD1, MIGUEL A. GÓMEZ2
1Statistics Department, California Polytechnic State University, San Luis Obispo, USA.
2Faculty of Physical Activity and Sport Sciences, Polytechnic University of Madrid, Spain
Running title: Home advantage in national football leagues worldwide
Key words: home advantage, competitive balance, crowd size, travel distance, altitude
2
Abstract
Comparison was made of home advantage in 157 national domestic football leagues
throughout the world for six seasons between 2006 and 2012, a total of 169,752 games.
Quantified as the number of competition points won by the home team expressed as a
percentage of all points gained in the league, the advantage was present in all continents, but
showed considerable variation between countries. It was by far the highest in Nigeria
(86.82%), followed by Bosnia-Herzegovina, Guatemala, Indonesia, Algeria, Bolivia and
Ghana, all between 70% and 80%. Regionally, there were pockets of high home advantage in
the Andes, the Balkans, West Africa and Central America, while in the Baltic republics and
in most of the Arabian Peninsula it was low. A multivariate model was developed using
proxy variables for the main explanations of home advantage and its worldwide variation.
After controlling for the competitive balance in each league, significant predictors of home
advantage were: FIFA ranking (a proxy for crowd support), the maximum geographical
distance between teams, the majority of teams coming from a single city, at least two teams
playing at high altitude, the recent occurrence of a civil war and the corruption perception
index. The model accounted for 43% of the variability in home advantage, the rest of which
needs to be investigated after developing more precise measures for territoriality, the ethnic
and cultural factors involved and the potential for referee bias.
3
Introduction
Although there is now an extensive literature on home advantage in football, most studies
have focused on data from England, or from other western European countries, and many
have been univariate in design (Pollard, 2008a). The purpose of the present study was to
address both of these shortcomings by carrying out a broad survey of home advantage
worldwide and then to examine possible explanatory variables using a multivariate approach.
The only previous study to investigate home advantage in football over a wide range of
countries on more than one continent was by Pollard (2006a) in which home advantage was
compared in the national leagues of all countries in Europe and South America, together with
a handful of countries from other parts of the world, a total of 72 countries. A clear pattern of
unusually high home advantage was found in the Balkan nations, especially Bosnia-
Herzegovina and Albania, as well as in the Andean nations in South America. It was
suggested that a heightened sense of territorial protection in the more isolated regions of these
countries might be a contributing factor. A study of home advantage in basketball in 35
national leagues in Europe also showed high values in Balkan countries, suggesting the
previous finding was not confined to football (Pollard & Gómez, 2013). Home advantage has
also been compared between men’s and women’s football leagues throughout Europe
(Pollard & Gómez, 2012). After controlling for differences in competitive balance between
the leagues of 26 European countries, a country’s score on a measure which quantified the
status of women in a country (the ‘Gender Gap Index’), was shown to be a significant
predictor of the difference between home advantage in the male and female leagues.
Compared with other sports, a meta-analysis by Jamieson (2010) found that in football home
advantage was significantly greater than in each of the nine sports against which it was
compared.
4
In order to broaden the scope of the present study as much as possible, the approach
was first to quantify home advantage in the men’s national domestic leagues of each country
in the world for which home and away records were available. A multivariate analysis would
then be carried out in order to relate home advantage to a set of characteristics for each
country that, in the light of previous research, might be expected to be having an influence.
The challenge was to find meaningful proxy variables for these effects, since it was unlikely
any could be measured directly in a consistent way for each country. The ultimate aim was to
assess the significance and to quantify the magnitude of each explanatory effect.
In a review of home advantage in football, Pollard (2008a) listed a set of plausible
explanations for the advantage, together with a review of the published evidence for and
against each. To set the present study into the context of what is already known about
football’s home advantage, these explanations, together with the main references relating
specifically to football, can be briefly summarized as follows:
Crowd effects The size, density, intensity and proximity of crowd support should provide an
advantage to the home team through their effects on both sets of players, and possibly the
referee. Although an analysis of the top nine levels of play in England convincingly showed
that leagues with small average crowds have lower home advantage, the relationship with
crowd size was not a linear one (Pollard, 2006b), nor was it in an earlier study by Nevill,
Newell and Gale (1996). For example in England home advantage appeared to be very
similar, with no relationship to crowd size, for the top four leagues in which average crowds
ranged from under 4,000 to over 30,000. Moreover, within a single league, no study has been
able to demonstrate a relationship between home advantage and crowd size. Nevertheless, in
European competition Goumas (2013b), after controlling for confounding variables, did find
a significant positive association between crowd size and home advantage. With regards to
crowd proximity, the absence of a running track has been shown to be related to an increased
5
home advantage in several countries including Greece (Armatas & Pollard, 2012) and
Germany (Dohmen, 2008).
Travel effects The distance the away team has to travel can magnify home advantage.
Pollard, Silva and Medeiros (2008) showed this to be the case in Brazil where distances are
large, but cautioned that differences in climatic conditions might also be playing a
confounding role. In Australian soccer, Goumas (2013a) has shown home advantage to be
influenced by the number of time zones crossed.
Local derbies Games between teams of the same city require less travel and have more
balanced crowd support and hence should lower home advantage. This has been shown in
London (Clarke & Norman , 1995), Istanbul (Seckin & Pollard, 2008), and in Lisbon, Paris,
Madrid and Milan (Pollard & Gómez, 2009).
Familiarity with local conditions This is likely to favour the home team. It has been shown
to apply both to the playing surface (Barnett & Hilditch, 1993) and to pitch dimensions
(Clarke & Norman, 1995). Extremes in climate or altitude are other possibilities (McSherry,
2007), and even choice of the type of ball (Dosseville, 2007). The slight drop in home
advantage that has been observed when a team moves to a new stadium (Loughead, Carron,
Bray, & Kim, 2003) also suggests that familiarity with the home venue is a factor in home
advantage.
Referee bias There are now many studies providing evidence of referee bias in favour of the
home team (e.g. Dawson, Dobson, Goddard, & Wilson, 2007). Although this is likely to be
the result of a subconscious reaction to the home crowd, the possibility of deliberate bias as a
consequence of pressure from outside sources such as gambling interests can no longer be
discounted in the light of recent revelations about match-fixing worldwide.
Territoriality An aggressive response to a perceived invasion of one’s territory is a natural
human reaction (Neave & Wolfson, 2003). Applied to football there is evidence that teams
6
playing in isolated locations, especially in front of an ethnically or culturally distinct home
crowd, and in countries with a history of violent conflict may experience increased home
advantage (Pollard, 2006a; Pollard & Seckin, 2007; Pollard & Gómez, 2009).
Special tactics The more cautious approach often favoured by the away team may
unwittingly transfer a psychological and territorial advantage to the home team and thus
influence home advantage (Lago & Martin, 2007; Tenga, Holme, Ronglan, & Bahr, 2010).
Rule factors Since all leagues now award three points for a win and play under the same set
of rules, the effect on home advantage of rule changes, detailed by Pollard and Gómez
(2009), does not need to be considered in the present study. Likewise, Allen and Jones
(2012) concluded that other rule factors, such as home shirt colour, are unlikely to affect
home advantage in football
Team composition Partially as a result of the Bosman ruling of 1995 it is now much easier for
players to change clubs both domestically and internationally. As a result players often now
perform, sometimes for only short periods of time, in cities and countries to which they have
little attachment or sense of ‘home’. Pollard and Gómez (2009) suggested this as a
contributing factor to the sharp decline in home advantage that was found in France, Italy,
Portugal and Spain from the mid 1990s.
Psychological factors Home advantage has been shown to have existed in football at least
since the dawn of the professional game in England in the 1880s (Pollard & Pollard, 2005).
There is a real likelihood that since players, coaches, referees and fans are all aware of its
importance, it becomes a self perpetuating phenomenon. Other psychological and
physiological considerations are discussed by Neave and Wolfson (2004), while Thelwell,
Greenlees and Weston (2009) have shown that psychological skill usage by players is more
frequent at home than away.
7
Not all these explanations could be easily incorporated into the analysis which was
based on league tables as opposed to individual teams and specific games. As a result, the
approach was to look for suitable proxy variables that would capture the essence of these
explanatory factors. The procedure and rationale for the choice of these variables are
described in the following section.
Methods
Scope and sources of data
The Fédération Internationale de Football Association (FIFA), the governing body of world
football, has 209 members, each representing a different country or territory. These are
divided into six confederations: Asia (AFC, 46 countries), Africa (CAF, 54 countries),
Central and North America (CONCACAF, 35 countries), South America (CONMEBOL, 10
countries), Oceania (OFC, 11 countries) and Europe (UEFA, 53 countries). A few countries
(e.g. Australia, Israel, Guyana) play outside their geographical continent. The object of the
study was to make a valid comparison of home advantage in the national leagues of as many
of these countries as possible using data for the six most recently completed seasons, as at
January 1st, 2013. This meant either 2006-2007 to 2011-2012, or 2007 to 2012 depending on
the timing of each country’s league season. The primary source of data was the open access
website www.soccerway.com which displays home and away final league tables for many
countries over many years. When the required table was not available a second open access
website www.rsssf.com was accessed. Although home and away tables are not given on this
website, it is more comprehensive in its coverage, usually giving a full list of the results of
each game played in a season, from which the information required to calculate home
8
advantage can be extracted. For some of the smaller or more remote countries sufficient data
could not be obtained from either website. The final data set consisted of 157 countries.
Before considering the reasons for excluding 52 countries, the quantification of home
advantage needs to be defined.
Home advantage
Home advantage was calculated by the well established method used by Pollard and Pollard
(2005). In any league in any season, home advantage was defined as the number of points
gained by all home teams expressed as a percentage of all points gained in the league, home
and away. Thus 50% represented no home advantage and the higher the value above 50%,
the greater the advantage. Since FIFA has mandated that all leagues award three points for a
win, one point to each team for a draw and zero points for a loss, the method of calculating
home advantage was identical for all leagues. Ideally the league should employ a balanced
playing schedule in which each team plays each other team the same number of games at
home and away.
Reasons for exclusion
The general principal adopted was to include as many countries as possible, eliminating
leagues only when the validity of a home advantage calculation was clearly compromised.
To insist on a full six seasons of perfectly balanced competition would have meant the
elimination of the majority of countries. Thus the following guidelines were used, with the
proviso that a country needed at least two full seasons of data to be included. Most of the
countries eliminated were small island nations of the Caribbean and Pacific, or nations in the
more remote parts of Africa and Asia. Apart from the obvious situation in which a country
did not have a national league, the main reasons for exclusion were as follows:
9
All or most games in the league were played at the same stadium Many teams in many parts
of the world share the same stadium, so that when games are played between such teams
home advantage has no meaning. It was decided that if more than 50% of teams in a league
played in the same stadium, the league should be excluded, since the concept of home
advantage could not have existed for a substantial proportion of games. In San Marino,
although several stadiums are used, teams are randomly assigned to the stadiums before each
round of matches, a related reason for exclusion.
Unbalanced playing schedule In some countries each team played each other three times in a
season, either once at home and twice away, or twice at home and once away. Since this
slight loss of balance is built into the structure of the league, the effect on the overall home
advantage of the league was likely to be minimal. Such leagues were not excluded. A more
serious situation occurred in leagues where some teams played greatly different number of
games at home and away. Leagues in which at least one team played more than 60% of its
games either at home or away were excluded.
League results unavailable It was not always possible to obtain the full results from a league
needed for the calculation of home advantage.
Teams divided into regional groups Instead of a fully national league, some countries divide
teams into groups, with group winners advancing to a mini-league or a play-off to decide the
national champion. When the groups were regional and games were played only between
teams within the same group, the league was excluded since neither group represented the
entire country. In other situations both groups were included. This could either be when
10
each group covered the entire country, or when an inter-locking schedule was played with
teams from each group playing teams from the other group as well as their own, such as in
the United States.
These guidelines resulted in the exclusion of 52 countries, mostly for more than one
reason. Since many countries changed the format of their leagues from year to year, not all
included countries had a full six seasons of data. The final selection of 157 countries,
classified by confederation, and showing the number of seasons, is shown in Tables 1 – 6.
Excluded countries are listed at the foot of each table.
***Tables 1 to 6 near here***
Quantification of variables
For each country, a value for the dependent variable, home advantage, was obtained by
combining the results of all games played over the six seasons. The aim of the study was to
relate this measure of home advantage to the following predictor variables:
FIFA world ranking This is a measure of the relative success of each country in international
competition. It could therefore be used as a proxy for the strength of each national league,
which in turn could be used as an indicator of crowd size and support. Partial justification for
this was based on an analysis of the countries of Europe for which attendance figures were
available at www.european-football-statistics.co.uk open access web domain. The Spearman
correlation coefficient between FIFA rank and average crowd size rank was +0.693. The
need to make use of FIFA rankings in this way was due to the fact that attendance data was
available for only a few countries outside Europe. The rankings for December 2009 were
11
used, this being immediately after the conclusion of qualifying games for the 2010 World
Cup finals, as well as being near the mid-point of the study.
Distance This variable was used as a proxy for travel effects. It was defined as the maximum
geographical distance in kilometers between any two teams in a country. Since the
distribution of distances was highly skewed, its natural logarithm was used in the analysis.
Although this was not an ideal measure, it was preferred to the area of each country since
many countries have large sparsely inhabited regions devoid of professional football. The
time and difficulty involved in calculating a marginally more precise measure, such as the
average travel distance between each pair of teams within a country was not considered
practical.
Altitude Human athletic performance is thought to start being affected at altitudes above
about 1,500m and thus could heighten home advantage in countries with teams playing at
high altitude. Altitude was incorporated as a binary variable, coded as 1 if two or more teams
in the league had a home stadium above 2,000m (there were eight such countries), and 0
otherwise.
Climate It is possible that travelling from one climate zone to another might have an adverse
effect on the away team and thus influence home advantage. For each country, this effect
was quantified as the number of different climatic zones (1, 2, 3 or more) based on a
modification of the Köppen system of climate classification and excluding parts of a country,
such as a desert or polar region, containing no teams in the national league. The altitude and
climate variables were intended to capture some aspects of familiarity with local playing
conditions.
12
Ethnic fractionalization Since previous studies had suggested that the degree of ethnic
division within a country might be a factor in home advantage related to territoriality, a
measure that captured this concept was sought. The closest such variable available for a wide
range of countries has been developed and quantified by Alesina, Devleeschauwer, Easterly,
Kurlat, and Wacziarg (2003) and named ‘ethnic fractionalization’. A score, on a scale of 0 –
1, was available for all but seven of the 157 countries (six small island nations and Palestine),
the higher the score the greater the fractionalization.
Civil war Because the effects of civil conflict, both lingering and current, had also been
proposed as a factor in home advantage, countries were coded by this binary variable
depending on whether or not a civil war had taken place at some time during the last 20
years, a somewhat arbitrary time period but one that should ensure memories and scars from
recent conflict. A civil war was defined as an armed conflict between two or more factions
within a single country taking the lives of at least 10,000 people. The necessary information
was obtained from the UCDP Conflict Encyclopedia (Uppsala Conflict Data Program, n.d.)
and the UCDP/PRIO Armed Conflict Dataset (Gleditsch, Wallensteen, Eriksson, Sollenberg, &
Strand, 2002; Themnér &Wallensteen, 2012). Twenty countries qualified. Since civil wars
are likely to invoke feelings of territorial protection, this variable was also included as a
possible proxy for territoriality
Corruption perceptions index In order to avoid the possibility of crowd disturbance following
a home defeat, accompanied by endangerment of officials, referee bias in favour of the home
team, enhanced by pressure or even intimidation from local officials, is a plausible factor in
some countries. Bribery is also a possibility. This is obviously difficult to assess, let alone
13
quantify. However it could be argued that the likelihood of this occurring might be related to
the degree of corruption perceived to be present in a country, especially if gambling interests
are involved. This has been quantified by Transparency International with a measure called
the ‘corruption perceptions index’ (http://cpi.transparency.org/cpi2012/results/). This is on a
scale from 0 – 100, the higher the score the less corruption, and was available for all but ten
of the countries (nine small island nations and Palestine). It was hoped that this variable
would capture the aspects of referee bias not solely due to the influence of the crowd.
Local derbies Home advantage is known to be reduced for local derbies, games in which
teams from the same city play each other. Although this represents a small minority of
games in most leagues, professional football teams in some countries are concentrated in the
capital city with a significant number of games qualifying as local derbies. A binary variable
was constructed, coded 1 if more than 50% of teams in the league were from the same city
and 0 otherwise. This variable captured 27, mostly small countries, but also with Uruguay
and Paraguay as notable inclusions.
Competitive balance The amount of competitive balance among the teams in a league has
been shown to influence home advantage when quantified as the percentage of points won by
the home team (Pollard & Gómez, 2012). This is because if two teams differ greatly in
ability, this difference is going to far outweigh the effect of home advantage in determining
the result of the game; the stronger side is likely to win both at home and away, thus masking
the home advantage effect. It therefore follows that the more evenly matched are the teams
in a league, the bigger is likely to be the effect of home advantage. A measure of league
competitive balance needed to be incorporated into the analysis in order to control for its
effect. Trandel and Maxcy (2011) reviewed the development of such a measure based on a
14
comparison of the amount of variability in the winning percentage of teams in a league with
what would be expected if all teams were of the same ability. Their refinement of this
measure adjusting for the existence of home advantage was used in the present study. This
produced a value to estimate the competitive balance in a league each season for each
country, the lower this value the more competitive balance. For each country their mean
value of this measure was used. Ten of the 11 least balanced leagues were in former Soviet
republics in Eastern Europe, the Caucasus and Central Asia. Cyprus was the other such
country.
Hypotheses
In the light of findings from previous studies on specific explanations for home advantage in
football, the following hypotheses were formulated and predicted that:
1. The higher its FIFA ranking, the greater the home advantage in the country.
2. Home advantage would be higher in countries with teams playing at high altitude.
3. The greater the distance between teams in a country, the higher the home advantage.
4. The greater the number of climate zones within a country, the greater the home
advantage.
5. The more ethnic fractionalization within a country, the greater the home advantage.
6. Home advantage would be higher in countries that had experienced civil wars.
7. The more perceived corruption in a country, the greater the home advantage.
8. Home advantage would be lower in countries with a large number of local derbies.
9. The greater the competitive balance in a country, the higher the home advantage.
Analysis
15
In order to assess the extent to which the variation in home advantage could be explained by
the predictor variables described above, a stepwise multiple regression approach was used
(Minitab 16 Statistical Software, 2010). An analysis of residuals was made to check the
validity of the regression assumptions. In view of the stated hypotheses, all tests that
predicted a specific effect were one-sided.
Results
Complete lists of countries, classified by continental confederation, together with their home
advantage values are shown in Table 1 - 6. The countries are listed in descending order of
home advantage magnitude. In seven countries, more than 70% of points were won by the
home team. The top four countries were each from a different continent, with Nigeria having
by far the highest home advantage in the world (86.82%), followed by Bosnia-Herzegovina,
Indonesia and Guatemala. Other countries with values above 70% were Bolivia, Algeria and
Ghana. The results are also shown on a world map (Figure 1). The advantage appeared to be
especially high in several regional groups of countries. These were in western Africa around
the Gulf of Guinea (mean home advantage, 69%), the Andes (67%), northern Central
America (67%) and the eastern Balkans (67%), all well above the world average of 60%.
There were also some pockets of low home advantage. For example, the three Baltic
republics averaged only 53% and the Arabian Peninsula also comprised countries with home
advantage mostly well below the world average.
***Figure 1 near here***
In the regression modeling ‘ethnic fractionalization’ was not a significant predictor of
home advantage. In addition it became clear that the climatic variable (number of climate
16
zones) was largely a function of both the size of the country (closely related to ‘distance’) and
variation in elevation (closely related to ‘altitude’). This meant that once distance and
altitude were included in the model, there was little additional predictive value when
including climate. The final model showing the six significant predictors is shown in Table
7. An examination of the residuals did not reveal any problems with assumptions needed for
model and testing validity.
***Table 7 near here***
The importance of controlling for competitive balance was confirmed, a factor that
clearly needs to be incorporated when comparing the home advantage of different leagues.
The occurrence of a civil war and the perception of corruption were both significant
predictors. A civil war increased a country’s home advantage by an average 3.66 percentage
points. ‘FIFA ranking’, the proxy for crowd size, was also a highly significant predictor, as
was ‘local derbies’, the situation in which a national league contained over 50% of teams
from the same city. In this case, home advantage was reduced by an estimated 3.00
percentage points. ‘Distance’ and ‘altitude’ were also significant. The eight countries in
which at least two teams play at high altitude had home advantage increased by an average
2.96 percentage points. The model accounted for 43% of the variation in home advantage
between countries. However it did not provide a satisfactory fit for the four countries with
very high home advantage (Nigeria, Bosnia-Herzegovina, Guatemala and Indonesia), all of
which had high standardized residuals. Myanmar was the only country with home advantage
significantly lower than predicted by the model.
Discussion
17
There was considerable variability in home advantage between countries, less than half of
which could be explained by a multiple regression model containing proxies for the main
known explanations for home advantage. The nature of the study, with home advantage
estimated from home and away league tables and not from individual teams or matches,
meant that the variables constructed were always going to be somewhat imprecise proxies for
their intended effects. However the fact that most of them contributed a significant amount to
the observed variation in home advantage suggested that their effects are real and a more
precise method of variable quantification would lead to model producing a better fit.
Nevertheless the inability to fully explain the very high figures in several countries was an
indication that there are other factors at work that still need to be identified.
With regards to individual countries, Nigeria stands out with an extraordinarily high
home advantage value of 86.82%, eight percentage points above any other country. Two
factors that may not have been adequately captured by the variables provide possible
explanations. The three main ethnic groups in Nigeria (Hausa, Yoruba and Igbo) are grouped
in specific regions of the country. Most teams and their supporters will identify with one of
these groups, or in some cases with a smaller ethnic group. With traditional distrust and
conflict between groups, stirred also by a religious component, there arises a similar situation
that existed in the Balkans where an unusually high home advantage was found following the
break-up of Yugoslavia and attributed to the concept of territoriality (Pollard, 2006a). The
corruption issue is also likely to be a factor, two Nigerian journalists being in no doubt. ‘The
home team is always favoured one way or the other and the referees are often bribed to swing
matches in their favour’ (Ude, 2012) and ‘The Nigerian league is notorious for .... match-
fixing, bribery, hooliganism .... the match referees and their assistants have been
compromised to guarantee victory for the home team’ (Alao, 2012). As for the reasons for
18
the alleged corruption, gambling interests are a possibility but Ude (personal communication,
November 29, 2012) went on to state: ‘The match officials and commissioner are scared of
what might happen if the home team loses. Things like crowd violence.’ Even the chairman
of the Nigerian National League openly admits that league referees are regularly bribed by
club officials (Ngobua, 2013).
The failure of the model to capture regional ethnic division might also be playing a
role in the inability to fully account for the high home advantage in other countries such as
Indonesia, Bosnia-Herzegovina, Guatemala and Ghana, all above 70%. Although the
variable ‘ethnic fractionalization’ was included in an attempt to incorporate ethnic division,
this was more a function of the number of different ethnic groups in a country and not the
extent to which they clustered as dominant groups in specific geographical areas. Other
factors present in countries with high home advantage and perhaps not fully captured by the
derived variables include travel distance (Indonesia, where teams also come from five
different islands), altitude (Guatemala and Bolivia) and civil wars (Indonesia, Guatemala and
Bosnia-Herzegovina). An earlier conclusion that territoriality was the main explanation for
the high home advantage in the Balkans and Andes (Pollard, 2006a) might also apply to
Indonesia and Guatemala, as well as Nigeria and Ghana and other countries in western
Africa. A more detailed analysis of game by game results would be needed to explore this
issue further.
Many countries with low home advantage were small island nations or other countries
with a low FIFA ranking. Exceptions, for which there were no obvious explanations, were
Ireland and Northern Ireland (both below 54%), and the Baltic nations of Latvia, Lithuania
and Estonia although for these three countries average crowd size, all below 700, was very
low in relation to their FIFA ranking. Home advantage was also low throughout the Arabian
Peninsula with the notable exception of Yemen (63%), compared with an average 53%
19
elsewhere in the peninsula. Since Yemen was the only country in the region with teams at
altitude, with a score on the civil war variable and with the worst regional corruption
perception rating, its relatively high home advantage value was in fact accurately predicted
by the model. Uruguay and Paraguay (both under 55%) were noticeably lower than
elsewhere in South America. As noted earlier, the overwhelming majority of teams in these
countries were from the capital cities of Montevideo and Asunción. Thus almost all games
were local derbies for which home advantage is known to be lower. An additional related
factor is that teams playing in capital cities have been shown to have lower home advantage,
not just as a consequence of local derbies (Pollard & Gómez, 2009).
The ambitious scope of the study and the nature of its somewhat imprecise
explanatory variables meant that the investigation was largely exploratory, hopefully serving
as a launching pad for subsequent more targeted and localized research. Nevertheless new
insights into a number of the explanations into the causes of home advantage in football have
surfaced. Distance and altitude are factors that both seem to be playing a role. These will be
better understood through analysis of individual games, rather than league tables, especially
in countries with extreme values on these variables; for example, Australia, Indonesia and the
United States for distance and the Andean countries for altitude. The exact role played by
crowd size continues to be difficult to pin down. Again a game by game approach is needed,
together with the inclusion of other location variables such as the existence of a running
track.
The explanations with wider cultural implications are more challenging. There is
definite evidence of an ethnic component to home advantage, but exactly how it operates and
how it might be quantified remain uncertain. Likewise the effects of armed conflict on a
national or regional psyche and how this might induce or augment a feeling of territorial
protection with implications for home advantage are not easy to formulate. How this feeling
20
manifests itself is a matter of speculation, although Neave and Wolfson (2003) have shown
that testosterone levels of players increase more before a home game than away, a finding
that has been replicated for ice hockey (Carré, Muir, Belanger, & Putnam, 2006) along with
increases in other physiological and psychological variables. Smith (2005) suggested that a sense of
territory surrounding the game location extends beyond the players to both the fans and the media, all
of which contributes to home advantage. There is plenty of scope for more research into the
psychological and physiological state of players before a match and to what extent this might vary
across different cultures and ethnic groups.
It was not possible to directly address the possibility of referee bias from the use of
league tables, but the finding of a relationship between the perceived magnitude of corruption
in a country and home advantage was worrying. Opportunities for many types of illegal and
unethical activities in professional football, including the bribing of players and officials,
have recently been spelt out in detail (“Villian’s guide to football”, 2013). It is quite possible
that, in some countries, this could be on a large enough scale to affect home advantage.
There are other components of home advantage that were not addressed in the study;
for example, the relationship between players, teams, their supporters and the local
community. Smith (2003) has explored how this might influence home advantage and how
changes that have taken place in the game over the last 20 years have affected these
relationships.
There are also implications for the coaching staff of teams throughout the world.
There is no doubt that home advantage exists everywhere football is played, but how best
should an away team prepare to minimize its effect, and how best should a home team try and
maximize its advantage? Attempts to address this question have been made (Pollard, 2008b;
Wolfson & Neave, 2004), but increased knowledge of the reasons for home advantage will
doubtless produce more specific strategies.
21
Finally there are now an increasing number of excellent websites with ever expanding
data on football statistics from around the globe. These present exciting possibilities for low
cost research into many football-related activities, not least home advantage.
References
Alao, M. (2012, December 23). Random musings on the NPL. Soccertalk. Retrieved January
24, 2013 from, http://soccertalknigeria.blogspot.com
Alesina, A., Devleeschauwer, A., Easterly, W., Kurlat, S., & Wacziarg, R. (2003).
Fractionalization. Journal of Economic Growth, 8, 155-194.
Allen, M. S., & Jones, M. V. (2012). The home advantage over the first 20 seasons of the
English Premier League: Effects of shirt colour, team ability and time trends.
International Journal of Sport and Exercise Psychology,
http://dx.doi.org/10.1080/1612197X.2012.756230.
Armatas, V., & Pollard, R. (2012). Home advantage in Greek football. European Journal of
Sport Science, iFirst article, 1-7, doi: 10.1080/17461391.2012.736537
Barnett, V., & Hilditch, S. (1993). The effect of an artificial pitch surface on home team
performance in football (soccer). Journal of the Royal Statistical Society A, 156, 39-
50.
Carré, J., Muir, C., Belanger, J., & Putnam, S. K. (2006). Pre-competition hormonal and
psychological levels of elite hockey players: relationship to the “home advantage‟.
Physiology and Behavior, 89, 392-398.
Clarke, S. R., & Norman, J. M. (1995). Home ground advantage of individual clubs in
English soccer. The Statistician, 44, 509-521.
22
Dawson, P., Dobson, S., Goddard, J., & Wilson, J. (2007). Are football referees really biased
and inconsistent?: evidence on the incidence of disciplinary sanction in the English
Premier League. Journal of the Royal Statistical Society A, 170, 231-250.
Dohmen, T. (2008). The influence of social forces: Evidence from the behavior of football
referees. Economic Inquiry, 46, 411-424.
Dosseville, F. E. M. (2007). Influence of ball type on home advantage in French professional
soccer. Perceptual and Motor Skills, 104, 347-351.
Gleditsch, N. P., Wallensteen, P., Eriksson, M., Sollenberg, M., & Strand, H. (2002). Armed
Conflict 1946–2001: A New Dataset. Journal of Peace Research, 39, 615–637.
Goumas, C. (2013a). Home advantage in Australian soccer. Journal of Science and Medicine
in Sport, http://dx.doi.org/10.1016/j.jsams.2013.02.014.
Goumas, C. (2013b). Modelling home advantage in sport: A new approach. International
Journal of Performance Analysis in Sport, 13, 428-439.
Jamieson, J. P. (2010). The home field advantage in athletics: A meta-analysis. Journal of
Applied Social Psychology, 40, 1819-1848.
Lago, C., & Martín, R. (2007). Determinants of possession of the ball in soccer. Journal of
Sport Sciences, 25, 969-974.
Loughead, T. M., Carron, A. V., Bray, S. R., & Kim, A. J. (2003). Facility familiarity and the
home advantage in professional sports. International Journal of Sport and Exercise
Psychology, 1, 264-274.
McSherry, P. E. (2007). Altitude and athletic performance: statistical analysis using football
results. British Medical Journal, 335, 1278-1281.
Minitab 16 Statistical Software (2010). [Computer software]. State College, PA: Minitab, Inc.
(www.minitab.com)
23
Neave, N., & Wolfson, S. (2003). Testosterone, territoriality and the ‘home advantage’.
Physiology and Behavior, 78, 269-275.
Neave, N., & Wolfson, S. (2004). The home advantage: psychological and physiological
factors. In D. Lavallee, J. Thatcher, & M. V. Jones (Eds.), Emotion and Coping in
Sport (pp. 131-148). New York: Nova Science.
Nevill, A. M., Newell, S. M., & Gale, S. (1996). Factors associated with home advantage in
English and Scottish soccer matches. Journal of Sports Sciences, 14, 181-186.
Ngobua, D. (2013, July 3). Why it is difficult to eradicate corruption in Nigerian football –
Inyama. Weekly Trust. Retrieved August 4, 2013 from http://weeklytrust.com.ng.
Pollard, R. (2006a). Worldwide variations in home advantage in association football. Journal
of Sports Sciences, 24, 231-240.
Pollard, R. (2006b). Home advantage in soccer: variations in its magnitude and a review of
the inter-related factors associated with its existence. Journal of Sport Behavior, 29,
169-189.
Pollard, R. (2008a). Home advantage in football: a current review of an unsolved puzzle. The
Open Sports Sciences Journal, 1, 12-14.
Pollard, R. (2008b). Away disadvantage: five steps for coping. Soccer Journal, 53(4), 53-54.
Pollard, R., & Gómez, M. A. (2009). Home advantage in football in South-West Europe:
Long-term trends, regional variation, and team differences. European Journal of
Sport Science, 9, 341-352.
Pollard, R., & Gómez, M. A. (2012). Comparison of home advantage in men’s and women’s
football leagues in Europe. European Journal of Sport Science. iFirst article, 1-7, doi:
10.1080/17461391.2011.651490.
Pollard, R., & Gómez, M. A. (2013). Variations in home advantage in the national basketball
leagues in Europe. Revista de Psicología del Deporte, 22, 263-266.
24
Pollard, R., & Pollard, G. (2005). Long-term trends in home advantage in professional team
sports in North America and England (1876 – 2003). Journal of Sports Sciences, 23,
337-350.
Pollard, R., & Seckin, A. (2007). Why is home advantage in South-east Europe the highest in
the world? In Y. Theodorakis, M. Goudas, & A. Papaioannou (Eds.), Book of long
papers, 12th European Congress of Sport Psychology (pp. 53-56). Halkidiki, Greece:
FEPSAC.
Pollard, R., Silva, C. D., & Medeiros, N.C. (2008). Home advantage in football in Brazil:
differences between teams and the effects of distance traveled. Brazilian Journal of
Soccer and Science, 1, 3-10.
Seckin, A., & Pollard, R. (2008). Home advantage in Turkish professional soccer. Perceptual
and Motor Skills, 107, 51-54.
Smith, D. R. (2003). The home advantage revisited. Journal of Sport and Social Issues, 27,
346-371.
Smith, D. R. (2005). Disconnects between popular discourse and home advantage research:
What can fans and media tell us about the home advantage phenomenon? Journal of
Sports Sciences, 23, 351-364.
Tenga, A. P. C., Holme, I., Ronglan, L. T., & Bahr, R. (2010). Effects of match location on
playing tactics for goal scoring in Norwegian professional soccer. Journal of Sport
Behavior, 33, 89-108.
Thelwell, R. C., Greenlees, I. A., & Weston, N. J. V. (2009). The influence of game location
and level of experience on psychological skill usage. International Journal of Sport
and Exercise Psychology, 7, 203-211.
Themnér, L., & Wallensteen, P. (2012). Armed Conflict, 1946-2011. Journal of Peace
Research, 49, 565-575.
25
Trandel, G. A,. & Maxcy, J. (2011). Adjusting winning-percentage standard deviation and a
measure of competitive balance for home advantage. Journal of Quantitative Analysis
in Sports, 7(1), Article 1. Available online at http://www.bepress.com/jqas.
Ude, A (2012, Sept 11). Nigerian premier league: can we ever have the same passion for our
local teams? Connect Nigeria.com. Retrieved December 7, 2012 from,
http://connectnigeria.com
Uppsala Conflict Data Program (n.d.). UCDP Conflict Encyclopedia. Retrieved January 14,
2013 from www.ucdp.uu.se/database.
Villian’s guide to football. (2013, July 13). Economist, pp. 53-54.
Wolfson, S., & Neave, N. (2004). Preparing for home and away matches. Insight, 8(2), 43-46.
26
Table 1. Home advantage in Asia (AFC), 2006 – 2012.
Country
n
GP
HW
HD
HL
HA (%)
Indonesia
4
1260
773
281
206
74.31
Uzbekistan
6
1266
720
264
282
68.59
Vietnam
6
1092
567
284
241
66.34
China
6
1410
658
443
309
63.82
Yemen
6
1020
495
273
252
63.08
Iraq
4
1677
759
509
409
61.61
Iran
6
1770
748
592
430
60.11
Turkmenistan
5
598
302
107
189
60.05
Syria
4
728
330
199
199
59.90
Sri Lanka
2
263
114
80
69
59.52
Australia
6
687
310
184
193
59.35
Tajikistan
6
846
435
119
292
58.87
Malaysia
6
1040
500
210
330
58.76
Pakistan
6
1136
514
283
339
58.40
India
6
858
364
256
238
58.15
Thailand
6
1572
656
465
451
57.23
Kyrgyzstan
6
520
251
85
184
56.81
South Korea
6
1264
539
332
393
56.33
Oman
6
792
326
229
237
56.22
Brunei Darussalam
3
233
115
29
89
55.82
United Arab Emirates
6
792
350
175
267
55.66
Japan
6
1836
790
445
601
55.60
Jordan
6
643
267
173
203
55.47
Saudi Arabia
6
892
379
222
291
55.38
Palestine
4
585
250
129
206
54.06
Lebanon
6
770
312
197
261
53.62
Singapore
6
1113
473
233
407
53.19
Myanmar
2
314
131
70
113
53.10
Taiwan
3
120
53
19
48
52.20
Qatar
6
801
311
191
299
50.81
Philippines
2
132
50
32
50
50.00
Kuwait
6
464
174
104
186
48.60
n = Number of seasons; GP = games played; HW = home wins; HD = home draws; HL = home losses; HA =
home advantage. Countries excluded: Afghanistan, Bahrain, Bangladesh, Bhutan, Cambodia, Guam, Hong
Kong, Laos, Macau, Maldives, Mongolia, Nepal, North Korea, Timor-Leste.
27
Table 2. Home advantage in Africa (CAF), 2006 – 2012.
Country
n
GP
HW
HD
HL
HA (%)
Nigeria
6
2041
1508
424
109
86.82
Algeria
6
1538
854
443
241
72.05
Ghana
6
1440
811
385
244
71.61
Togo
3
719
348
237
134
66.72
Cameroon
6
1274
603
380
291
63.60
Morocco
6
1440
618
527
295
62.77
Tunisia
6
1150
526
350
274
62.19
Zimbabwe
6
1438
659
415
364
61.35
Gabon
4
596
275
168
153
61.30
Angola
6
1267
581
362
324
61.21
Ethiopia
4
968
424
311
233
61.05
Sudan
6
966
461
230
275
60.46
Mozambique
6
1092
476
342
274
60.33
Kenya
5
1200
524
373
303
60.27
Uganda
6
1582
700
465
417
59.92
Seychelles
5
386
193
71
122
59.80
Egypt
6
1343
577
424
342
59.78
Namibia
6
792
361
209
222
59.62
Guinea
3
446
194
135
117
59.60
Libya
5
924
410
250
264
58.68
Benin
3
609
250
201
158
58.49
Burkina Faso
6
966
385
336
245
58.20
Zambia
6
1427
584
454
389
57.64
South Africa
6
1440
599
427
414
57.13
Congo DR
5
305
141
63
101
57.04
Cape Verde
6
173
77
41
55
56.90
Malawi
5
1050
454
274
322
56.88
Senegal
5
818
299
309
210
56.22
Rwanda
6
866
369
219
278
55.74
Tanzania
5
760
305
226
229
55.55
Lesotho
5
1034
418
299
317
55.40
Côte d'Ivoire
6
896
340
283
273
54.18
Mali
6
1145
459
309
377
53.93
Gambia
6
750
252
292
206
53.52
Mauritius
6
676
273
160
243
52.41
Botswana
6
1440
583
331
526
52.14
Mauritania
6
830
305
238
287
51.20
Swaziland
6
844
313
215
316
49.81
n = Number of seasons; GP = games played; HW = home wins; HD = home draws; HL = home losses; HA =
home advantage. Countries excluded: Burundi, Central African Republic, Chad, Comoros, Congo, Djibouti,
Equatorial Guinea, Eritrea, Guinea-Bissau, Liberia, Madagascar, Niger, São Tomé and Principe, Sierra Leone,
Somalia, South Sudan.
28
Table 3. Home advantage in North and Central America (CONCACAF), 2006 – 2012.
Country
n
GP
HW
HD
HL
HA (%)
Guatemala
6
1332
772
375
185
74.32
Haiti
6
1245
570
416
259
64.06
U.S.A.
6
1499
720
417
362
63.16
El Salvador
6
1080
491
349
240
63.02
Honduras
6
1077
506
312
259
62.69
Costa Rica
6
1180
559
333
288
62.68
Mexico
6
1836
832
531
473
60.82
Nicaragua
6
786
378
179
229
60.26
Cuba
5
827
376
215
236
59.27
Canada
6
834
381
181
272
57.04
Jamaica
6
1188
476
384
328
56.98
Grenada
5
444
196
100
148
55.84
Panama
6
958
397
263
298
55.69
Saint Kitts and Nevis
6
524
225
119
180
54.65
Bermuda
6
438
196
84
158
54.63
Belize
6
365
156
81
128
54.14
Suriname
6
704
306
141
257
53.73
Puerto Rico
5
251
112
43
96
53.38
Dominican Republic
4
231
98
49
84
53.26
Trinidad and Tobago
6
699
296
144
259
52.84
Guyana
3
176
73
35
68
51.52
Cayman Islands
4
336
128
76
132
49.36
n = Number of seasons; GP = games played; HW = home wins; HD = home draws; HL = home losses; HA =
home advantage. Countries excluded: Anguilla, Antigua and Barbuda, Aruba, Bahamas, Barbados, British
Virgin Islands, Curaçao, Dominica, Montserrat, Saint Lucia, Saint Vincent and the Grenadines, Turks and
Caicos Islands, U.S.Virgin Islands.
29
Table 4. Home advantage in South America (CONMEBOL), 2006 – 2012.
Country
n
GP
HW
HD
HL
HA (%)
Bolivia
6
792
475
164
153
71.84
Peru
6
1588
836
408
344
66.94
Colombia
6
1944
976
548
420
65.78
Brazil
6
2280
1141
616
523
64.89
Venezuela
6
1709
832
484
393
64.18
Ecuador
6
1236
606
321
309
63.15
Chile
6
2005
963
490
552
61.16
Argentina
6
2280
1019
635
626
59.50
Paraguay
6
1584
645
426
513
54.58
Uruguay
6
1425
589
329
507
53.12
n = Number of seasons; GP = games played; HW = home wins; HD = home draws; HL = home losses; HA =
home advantage. Countries excluded: None.
30
Table 5. Home advantage in Oceania (OFC), 2006 – 2012.
Country
n
GP
HW
HD
HL
HA (%)
Fiji
6
490
243
78
169
57.97
New Zealand
6
392
183
65
144
55.27
Tahiti
6
469
212
76
181
53.49
Cook Islands
3
126
52
20
54
49.16
Papua New Guinea
5
268
102
55
111
48.20
n = Number of seasons; GP = games played; HW = home wins; HD = home draws; HL = home losses; HA =
home advantage. Countries excluded: American Samoa, New Caledonia, Samoa, Solomon Islands, Tonga,
Vanuatu.
31
Table 6. Home advantage in Europe (UEFA), 2006 – 2012.
Country
n
GP
HW
HD
HL
HA (%)
Bosnia-Herzegovina
6
1440
994
230
216
78.53
Macedonia
6
1084
590
259
235
67.79
Albania
6
1172
636
273
263
67.25
Croatia
6
1314
680
313
321
64.84
Slovakia
6
1122
559
292
271
64.05
Norway
6
1324
667
328
329
63.91
Kazakhstan
6
1108
567
254
287
63.68
Austria
6
1080
523
292
265
63.13
Italy
6
2280
1078
631
571
62.25
Czech Republic
6
1440
686
384
370
62.04
Belgium
6
1608
776
408
424
61.96
Romania
6
1836
880
475
481
61.89
Bulgaria
6
1440
731
296
413
61.85
Netherlands
6
1836
913
412
511
61.83
Montenegro
6
1188
576
291
321
61.69
Spain
6
2280
1117
536
627
61.66
Poland
6
1440
685
372
383
61.47
Greece
6
1440
687
365
388
61.34
Hungary
6
1440
703
334
403
61.29
England
6
2280
1074
595
611
61.12
France
6
2280
1028
679
573
61.08
Russia
6
1440
648
418
374
60.53
Turkey
6
1836
859
463
514
60.26
Switzerland
6
1062
499
261
302
60.10
Georgia
6
1021
489
231
301
59.96
Sweden
6
1382
642
349
391
59.92
Serbia
6
1248
573
322
353
59.64
Portugal
6
1440
638
392
410
58.71
Germany
6
1836
829
459
548
58.35
Ukraine
6
1440
642
360
438
57.73
Finland
6
1124
494
291
339
57.55
Cyprus
6
1092
491
260
341
57.46
Denmark
6
1188
516
309
363
57.05
Slovenia
6
1080
474
267
339
56.81
Azerbaijan
6
916
405
217
294
56.58
Moldova
6
1179
518
283
378
56.45
Faroe Islands
6
810
367
173
270
56.45
Scotland
6
1188
514
287
387
55.81
Israel
6
1314
539
370
405
55.63
Belarus
6
1165
483
317
365
55.57
Armenia
6
784
356
148
280
55.17
Wales
6
1454
633
318
503
54.82
Lithuania
6
917
402
195
320
54.81
Iceland
6
750
322
171
257
54.69
Luxembourg
6
1092
459
245
388
53.51
Ireland
6
1101
442
283
376
53.28
Latvia
6
805
340
162
303
52.46
Northern Ireland
6
1271
518
270
483
51.48
Estonia
6
1080
465
172
443
51.08
Malta
6
582
229
130
223
50.56
n = Number of seasons; GP = games played; HW = home wins; HD = home draws; HL = home losses; HA =
home advantage. Countries excluded: Andorra, Liechtenstein, San Marino.
32
Table 7. Results of multiple regression analysis with home advantage as the dependent
variable.
Predictor
Coefficient
SE
t
P
Constant
64.115
3.527
18.18
<0.001
FIFA rank
-0.032
0.009
-3.59
<0.001
Log distance
0.682
0.384
1.77
0.039
Altitude
2.959
1.725
1.72
0.044
Civil war
3.662
1.138
3.22
0.001
Corruption perception
-0.053
0.021
-2.52
0.006
Local derbies
-2.997
1.111
-2.70
0.004
Competitive balance
-2.601
1.073
-2.42
0.008
33
Figure 1. World map of home advantage
