Unexplained game: can economists predict football victory


In 2014, The Washington Post published an article with the title “Why economists can’t predict the lead in the world Championship.” The publication pointed to the fact that several of the strongest teams in the world are countries with much weaker economies: for example, in Spain at that time unemployment was around 25%, prices in Argentina in 2013 increased by 28% and, according to forecasts, by the end of the year could jump by another 38%. The stakes in the match between Germany and Portugal favoured the victory of the first to a much lesser extent than one would expect, focusing on economic and social indicators of the two countries.

Insidious statistics

A number of researchers, e.g., Andrew Bernard and Meghan Buss (.pdf) in 2004 managed to show that the socio-economic indicators, such as the size of the population and the GDP allow you to predict the results of countries at the Olympic games. But these models include different types of sports respectively, in the country can beat the leaders alone and show weak results in others.

In the Championships, where teams compete in one sport, to predict the outcome much more difficult. Statistical analysis shows that neither the cost of the team, nor the salary of the coach (and then, whether he is a foreigner or a countryman of their players) do not affect the result, reported in July 2014 Daniel Kaufman, a visiting senior research fellow in the global economy and development Brookings Institution. Neither the size of the population, nor “luck” in the distribution (getting into a group with “weak” opponents) at the world Championships do not have a significant effect, concluded the expert.

In 2014, ING noted that if as a prognostic parameter for victory in the world Cup to use the total value of all players in the team, should have won the team of Spain is its squad of 23 players ING was estimated at €675 million In the second and third place in the list of the most expensive teams were Germany (€609 million) and Brazil (€507 million). The top five was rounded out by France (€453 million) and Argentina (€410 million). In the end, the first and second place went to Germany and Argentina, the third — the Netherlands (11th place for value). Brazil lost to Holland in the match for third place, Spain is not even out of the group.

Conflicting testimony

In 2007, in the article “prediction of the relative strength of national football teams” (published in Social Science Quarterly) Garry Gilad and Paul Dobson of city University London (first in 2008-2012 acted as a consultant on the statistics of football club “Chelsea”) explained 70% of variation in international rankings of football teams the following set of factors: the number of regularly playing football men, the length of football traditions in the country, the welfare of the population, the proportion of players playing for foreign football clubs, and climatic conditions.

On significant positive effect of the migration of players to foreign clubs on the results of national soccer teams have also pointed out Rita Berlinski, Jaron, Shockart and Johan Swinnen from the University of Leuven in 2010.pdf). For this, they have constructed the migration index of players, which takes into account the quality of foreign clubs, in which they act.

As did Gilad and Dobson

Football tradition was measured as the number of years between the country’s membership in FIFA and 2003 (mid-study by the authors between 2000 and 2005). The wealth of a nation is expressed through GDP per capita. According to the authors ‘ calculations, between 2000 and 2005, about 41% of the players included in national teams, were presented by foreign clubs: almost 77% was played in the richer countries, just over 86% in countries with higher FIFA rankings. Football success is reflected in the country it increased in the FIFA rankings — it takes into account the results of the matches, number of goals scored and goals conceded, bonus points for games against strong opponents and away matches.

Data on the effect of income on its football success, in fact, are extremely controversial. Robert Houston and Dennis Wilson in his article for Applied Economic Letters in 2002 showed that the ratio of soccer success and national income of the state follows the law of diminishing returns: every additional increase in income per unit ceteris paribus brings a slightly smaller increase football performance. But Peter Omondi-Ochieng from Louisiana state University in an article for the journal Team Performance Management in 2015 based on the 203 national teams came to the conclusion that the gross national income allows you to predict engagement with a team of football experts, but not the results of the performance of the team.

The evaluation of the Brookings Institution, 2014, at all important were two factors: the quality of democratic governance in the country (estimated based on World Governance Indicators calculated by the world Bank) and the number of fans who come to support the team. The quality of governance, say the authors, affects the degree of teamwork in relation to individual characteristics of the players. “During this tournament [world Cup 2014] Costa Rica, Chile, France and the United States have demonstrated good governance at the team level — in contrast to Cameroon, Ghana, Italy, and Spain, each of which showed a weak result despite the individual stars”, wrote Kaufman.