The home advantage refers to the phenomenon whereby the home team in sports wins more often than the visiting team does (Schwartz and Barsky 1977). In football, Pollard (1986) found that the average home team in the English Football League would win around 64% of the points at home. Sport spectators generally believe that their support is causing the home advantage (Smith, 2005; Wolfson et al. 2005), but reviews of empirical studies on the home advantage concluded that more factors are likely to have an influence on this phenomenon (Courneya and Carron 1992; Nevill and Holder 1999). These reviews conclude that the two key variables that are likely to have the strongest effect are familiarity with the stadium (and the associated relaxed feelings that go along with being in a familiar environment) and the support of the crowd.
In this article, I explain how I tried to disentangle these effects, to see whether crowd support or familiarity are likely to be more important for the home advantage in football. This article is based on research published in the Journal of Applied Social Psychology (Van de Ven 2011). More detailed information can be found in that article.
Both familiarity and crowd support might influence the home advantage in multiple ways. Familiarity with the field could lead players to perform better, because they are better adjusted to the turf or the lighting. Familiarity also leads to lower levels of stress for the home team (found in ice hockey by Carré et al. 2006). Supportive crowds can motivate players to do their best more and they can also influence a referee’s decision, leading to a referee bias favouring the home team (Boyko et al. 2007, but see Johnston 2008 for a critique). All these factors can contribute to home team performance, but they could also have negative effects. For example, crowd support can sometimes lead the home team to choke under the pressure (Baumeister 1984) and misbehaving crowds were found to exert a negative influence on performance of the home team in basketball (Thirer and Rampey 1979).
It thus remains unclear what the relative effect of both familiarity with the location and of crowd support is on the home advantage. It appears that there is an ideal situation in which these effects can be disentangled, namely same-stadium derbies such as AS Roma versus Lazio (who share the Stadio Olimpico) and AC Milan versus Internazionale (who share the Stadio Giuseppe Meazza, better known as the San Siro). In these derbies, both teams are equally familiar with the venue, but the “home” team brings more supporters to the game (as most tickets are sold as season tickets). Thus, if the crowd support is most important, there should still be a normal home advantage for the team designated to be playing at home. In contrast, if familiarity with the venue is more important for the home advantage, there should be no home advantage in these same-stadium derbies as both teams are equally familiar with the stadium.
The analyses
The data used for this analysis are a total of 64 Italian Serie A games of AC Milan, Internazionale, AS Roma, and Lazio Roma in the seasons 1988-1989 until 2003-2004. We left out later seasons, as in the seasons 2004-2005 and 2005-2006 several teams were punished for match-fixing in those seasons. For the four mentioned teams, we compared their home advantage in the same-stadium derby to their normal home advantage, against a team that was of equal quality as the team they shared a stadium with that season. For example, in the seasons 2003-2004, Internazionale accumulated 59 points in the league. We therefore compared the home advantage of AC Milan against Internazionale, with the home advantage of AC Milan against Parma (who was closest in standing to Internazionale that season, with 58 points). This allowed us to control for team quality.
Results
In normal games against teams that are of equal strength as the ones that share their stadium, a clear home advantage existed for the two Milan-based and two Rome-based teams. At home they won on average 1.64 points per game, while in away games they only won 0.91 points per game. This difference of 0.73 points was a clear and sizeable home advantage (p < .001).
For same-stadium derbies, however, no home advantage existed whatsoever, as there was no difference in performance for the designated home and away team. In those games, the designated home team won on average only 1.14 points, while the visiting team won 1.42. Statistical analysis show that there was no effect of who played at home in those games (p = .179). Also note that if anything, the pattern was in the opposite direction, namely that the away team did better than the home team.
Conclusion
The current data show that no home advantage exists in games in which the visiting team is equally familiar with the stadium as the home team is, even if the home team has the most crowd support. This suggests that crowd support is not a necessary precondition for the home advantage. It also suggests that being familiar with the field might be more important. This includes the specifics of being familiar with field size or lighting conditions, but probably also to a generalized feeling of being more relaxed and at ease in home games.
Note that one alternative explanation, the non-existent travel time for the visiting team is unlikely to have affected these results. Research suggests that travelling only has a very small effect on the home advantage, notably if time zones are crossed or because games are typically played with less time in between (Pace and Carron 1992). The fact that a home advantage also exists in countries such as the Netherlands, even though the maximum travel time between two clubs is less than 3¼ hours, makes this an unlikely explanation.
With this article, I do not intend to make the point that crowd support does not matter at all. Rather, it seems that there is much more that can be gained in knowledge here. There are clear positive effects of crowd support (via, for example, a referee bias), but apparently the potential negative effects can be just as large. Finding these possible negative effects of crowd support can be a valuable line of research to further investigate why players and teams perform better in home games.