Competition format, championship uncertainty and stadium attendance in European football – a small league perspective

Abstract

To increase the number of regular season games, small football leagues are often organized as quadruple round robin tournaments with teams playing each other four times. Theoretically, however, the more games played, the less uncertain is the championship race, reducing fan interest in the league. This article uses data from Austria and Switzerland to study the relationship between competition format, championship uncertainty and attendance demand empirically. Results suggest that a team still in contention to win the championship positively affects attendance, while it is not the specificity of the competition format that per se contributes to less championship uncertainty.

Keywords:

• competition format
• competitive balance
• uncertainty of outcome
• football
• Austria
• Switzerland

JEL Classification:

• D12
• L83
• Z2

Acknowledgements

Previous versions of this article were presented at the 6th Annual Meeting of the European Sport Economics Association (ESEA) in Antwerp as well as the 22nd Annual Meeting of the European Association for Sport Management (EASM) in Coventry. We would like to thank participants for their valuable comments and suggestions. Special credit is due to Oliver Budzinski, Alex Krumer and Nicolas Scelles as well as an anonymous referee and the editor David Peel who significantly helped to improve the article. We would also like to thank the Master students in the Sport Management program in Tuebingen (class 2012/13) for their help in collecting parts of the data during their research project seminar 2013/14.

Notes

¹ A recent review of the literature related to European professional football is provided by Pawlowski (2013). For further studies (also) in other sports, see Downward et al. (2009).

² A different approach is the implementation of play-off games after the regular season as applied in the Belgian Jupiler League (see Goossens et al. 2012 for an analysis of the effects of this competition format).

³ This argument is supported by Olson and Stone (2014) who find that an underdog’s chance to win a championship series against a dominant team decreases with an increasing number of games played.

⁴ FC Red Bull Salzburg in Austria and FC Basel in Switzerland realize more than one third of the overall league turnovers.

⁵ Before that season, the Swiss championship was made up of 12 teams, where the first eight teams of the regular season competed in the championship play-offs and the teams in 9th to 12th position competed with the top four teams of the 2nd division in a play-off/play-down round.

⁶ The country coefficient rankings are based on the results of each country’s clubs in the five previous UCL and UEL seasons. More information about the rankings and their calculation can be obtained from www.uefa.com.

⁷ FC Basel has been recently particularly successful also in the European competitions, reaching the round of last 16 of the UCL in 2011/12 and the semi-finals of the UEL in 2012/13.

⁸ The ‘Austrian pot’ is a regulation to promote the more active participation of young Austrian players in the 1st division football games. According this, each club has to have a specific number of players holding Austrian citizenship at match in order to get rewarded. The more minutes these players play, the higher the bonus the teams receive. For players under 21 years with Austrian citizenship, the teams are awarded double bonuses.

⁹ Total revenues were collected from the balance sheets of the two clubs.

¹⁰ Prize money information is obtained from the annual reports for the UCL and UEL as issued by UEFA.

¹¹ The gross sample is N = 1800 games. Yet in Austria (N = 1) and in Switzerland (N = 2), few matches were played without the presence of fans (league sanctions). Moreover, Neuchâtel Xamax FC club declared bankruptcy on 26 January 2012 and was consequently excluded from the Swiss league in the season 2011/12 (N = 18). Therefore, the net total sample for our empirical estimation consists of N = 1779 games. As there is no published attendance data that distinguish between season ticket holders and regular ticket purchasers for every single game, we were not able to correct for season tickets in our estimations. Yet, the absence of this kind of information is common for many empirical papers which focus on European football (e.g. Peel and Thomas, 1992; Czarnitzki and Stadtmann, 2002; Forrest et al., 2004; Buraimo and Simmons, 2008).

¹² Data for the betting odds are gathered from www.betbase.info which collect odds from 48 bet companies.

¹³ Data were gathered from www.freemeteo.com by using the nearest weather station for the home team city.

¹⁴ The shortest car distance between the cities was measured using www.maps.google.com.

¹⁵ Results of the robustness checks are available upon request.

¹⁶ We suggested also controlling for a ‘derby effect’ directly. A ‘derby’ takes place when the competing teams share a common geographical boundary and therefore attract traditionally greater audiences. We considered including a dummy variable in our models to depict those games. Yet in Austria and Switzerland, there are only two single rivalries (the Vienna and Zürich derby) that fall within this definition of ‘derby’. Therefore, we decided not to include this variable.

¹⁷ To test the robustness of these findings, we try to proxy a double round robin tournament format in the Austrian and Swiss league by splitting the seasons into two halves and considering only the first half – where all teams have the same starting point and play against each other twice (home and away) – for estimating HIUCS*. The results confirm our general findings since HIUCS* values based on the first half of the season are even higher (indicating less championship uncertainty) than those based on the entire season in both leagues (Fig. A1, Appendix).