The importance of domestic football leagues to international performance: predicting FIFA points

Abstract

This paper estimates the importance of domestic professional and semi-professional football clubs to the performance of national team sides in earning FIFA points. It is likely the case that the pathway to further development and a spot on the national team for a young player is, at least initially, through a domestic football league. Yet the stage of development of football leagues differs markedly across nations. This paper finds that the presence and development of these leagues is a significant contributor to FIFA points for the national side, but with diminishing returns. Most recently, the largest effect is found for the nations that play in the African (CAF) and Oceania (OFC) zones, while the other four zones (AFC, CONCACAF, CONMEBOL and UEFA) share smaller positive effects. Even in the presence of limiting economic and demographic factors, the road to international football success is through the development of domestic leagues.

Acknowledgements

Thanks to Pamela Wicker and Craig Kerr for valuable comments that improved the quality of the paper. Any errors and omissions are my own.

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

1. The top four teams in the AFC qualify for the World Cup out of 43 teams, most of whom do not rank highly. The fifth place team can qualify by defeating the fifth place team in the South American zone (CONMEBOL). The top team in the 10-team OFC zone must defeat the fourth place team in the North and Central American and Caribbean zone (CONCACAF) to qualify for the World Cup. The chances of qualification from the OFC zone are slim. Australia qualified in 1974, while New Zealand qualified in 1982 and 2010.

2. FIFA rankings are not without controversy and criticism. The ranking for each country is calculated based on a point system for victories and the quality of opponents. The ranking is based on a weighted average of points earned for the last three years with declining weights. The ranking calculation was changed by FIFA in 1998 and again in 2006 in response to criticisms of its accuracy and meaningfulness. A good reference is McHale and Davies, ‘Statistical Analysis of the Effectiveness’.

3. Canadian Soccer Association Annual Report 2014, http://www.canadasoccer.com/files/CanadaSoccer_AnnualReport2014_EN_web.pdf.

4. Hoffman, Lee, and Ramasamy, ‘Socioeconomic Determinants’.

5. Macmillan and Smith, ‘Explaining International Soccer Rankings’.

6. Houston and Wilson, ‘Income, Leisure and Proficiency’.

7. Leeds and Marikova Leeds, ‘International Soccer Success’.

8. Foer, How Soccer Explains the World.

9. Ibid.

10. Ibid.

11. Miyazaki, ‘Football Technology Transfer’.

12. Source: http://www.mlsplayers.org/salary_info.html (accessed November 3, 2014). These salaries include those for designated players whose salaries can be much higher than the average.

13. Source: http://blog.fieldoo.com/2014/02/league-review-nasl-reviving-the-past-glory-of-pele/.

14. Easton and Rockerbie, ‘Overtime! Rules and Incentives’.

15. Dixon and Coles, ‘Modelling Association Football Scores’.

16. This is the difference between two Poisson distributed random variables.

17. Miyazaki, ‘Football Technology Transfer’.

18. White, ‘A Heteroskedasticity-Consistent Covariance’.

19. Why just ten teams? Canada is a large country in terms of land mass and its major markets are spread out across the country. The only domestic professional league that operates at a national level in Canada is the nine-team Canadian Football League. A Tier II professional soccer league with ten teams will probably exploit the most viable financial markets in Canada and might struggle to survive. The Canadian Hockey League operates across the country and is composed of the Western Hockey League, Ontario Hockey League, and the Quebec Major Junior Hockey League. There are many teams but all are amateur.

20. Hausman, ‘Specification Tests in Econometrics’.

21. The FIFA zone fixed effects were not included in this regression.

22. Hausman, ‘Specification Tests in Econometrics’.