Globalisation of the World Cup
Until the globalisation of soccer towards the end of the 20th century, the World Cup had been held exclusively in Europe and the Americas. A tight correlation existed between the winners, and the continent of the host country. Brazil’s win in Sweden in 1958 was the only exception to that trend.
Looking at all tournaments, modern era included, no European side has ever won when the World Cup has been staged in either North or South America.
Italy, Netherlands, and West Germany have each fallen at the final hurdle when the events has been held in the Americas, with the Azzurri’s penalty shoot-out defeat at USA 94 the closest any team has come to breaking that hoodoo.
The winning side has been either Brazil – winning in 1962, 70 & 94 - and Argentina – as hosts in 1978, and again in 1986. Uruguay’s two wins started the trend but came at the dawn of the competition in 1930 and 1950.
Although these statistics hint at the added difficulty faced by European sides in winning “away”, we have to distinguish this effect as separate from the expected dominance of teams which are simply better on neutral footing, rather than a more general continental home field effect.
Measuring the impact of HFA
Conventional home field advantage (HFA) is both the most widely recognised, yet least understood effect in team sports. Although random variation will inevitably produce some clubs sides that produce better results away from home, even over the course of a season, these “away specialists” are not repeatable traits. Over the long term sides perform better at home than away.
Numerous explanations have arisen to account for this ubiquitous effect, ranging from crowd influence, familiarity with your surroundings to increased levels of testosterone when defending your own territory.
I have looked at the performance of English club sides when they move to a new home ground and the amount of HFA only gradually returns to normal league levels over the course of a few seasons. In addition, HFA appears to be steadily declining in many national leagues, implying that the effect, although real is also susceptible to change over time and shouldn’t be treated as an absolute.
If we want to see if any of the many possible contributing factors that produce a HFA can also spill over to also assist teams from the same geographical catchment area, giving them a continental boost, we need to look at all of the group and knockout finals matches rather than relying on simply the identity of the side that lifted the trophy.
With due respect to the rest of the world, Brazil 2014 again appears to be a contest between Europe and South America. The big two from each confederation, Brazil and Argentina and Spain and Germany respectively, dominate the betting and it is odds on that one of these four will be crowned winners. It is only when we reach halfway in the betting that a different confederation, Ivory Coast for CAF appears.
Measuring Home Continent Advantage
When looking at the possibility of a continental effect our remit must be relatively broad. By including USA in a catch all Americas, we are certainly moving away from a like for like comparison to a World Cup played in Brazil. However, if we exclude USA ‘94 we then should also omit two competitions played in Mexico, leaving very little data.
Limited sample sizes can lead to conclusions that may fail to be repeated in future events, so to maintain a decent sample size, but with that caveat, I have included Central and North American sides and venues in the data and disregarded tournaments prior to England’s sole success in 1966, where participation levels and logistical arrangements were very different to today.
Measuring the performance of a side at its most basic level either involves tallying up the wins, draws and losses or the average number of goals both scored and conceded over a period of matches. Both methods have the advantage of simplicity.
A goal-based study delivers its verdict in terms that can be readily used later, for example in a Poisson approach, but may be susceptible to overweighting abnormal large margin results. Recording wins and draws as a percentage of total games played, where a draw is consider half a win, partly eliminates the influence of high margin wins or losses, but is less useful, subsequently. It is recorded as the success rate in the table below.
Inclusive of 1966, six tournaments have subsequently been played in Europe and four in the Americas. The remaining two were shared by South Korea and Japan in 2002 and South Africa in 2010. These are where the data is drawn from.
|Europe vs. Americas||Goals For per Game||Goals Against per Game||Goal Difference||European success rate|
|Europe in Americas||1.18||1.31||-0.13||0.43|
|Europe in Rest of the World||1.17||1.03||0.14||0.54|
|Europe in Europe||1.52||0.91||0.61||0.65|
Sample sizes range from nearly 100 contests between Europe and the Americas in Europe to just over 60 in the Americas and nearly 40 in the two neutral venues of Africa and Asia.
A steady decline in performance by European sides against the Americas can be seen both in the success rate and the average goal difference as we move from games played in Europe, through the Rest of the World to the Americas.
It appears that failure of European teams to win the trophy in the Americas is an unhappy coincidence that disguises a more widespread advantage that exists and plays against visitors from Europe.
We have taken on trust that the relatively large sample size has produced similarly talented batches of opponents from the two confederations. Legitimate objections about the inclusion of North and Central American sides can be partly allayed by similarly tiered results appearing if these sides are omitted.
These trends also persist if the results recorded by the hosts, who will benefit from conventional HFA, are also given less weight. So there are good reasons to suspect the existence of a continental effect, if we stretch the definition to encompass North and South America, which will hamper European teams and assist countries closer to Brazil in 2014.
Europe’s record against the Americas in 2002 and 2010 might be considered as a baseline performance figure on relatively neutral ground. This performance level dips by an average of nearly three tenths of a goal per game when they face teams from the Americas on their own continent.
Allowing Brazil the more familiar home field advantage, but denying an additional continental premium to the hosts and their near neighbours is likely to result in over rating the chances of the European teams in a competition that is guaranteed to see many such games in the group stages alone.