Picking the World Cup winner – follow the money

England to beat Belgium in the final?

Missing the football, searching for a sign ...

Home alone last Friday evening with no football on TV for the first time in two weeks, I was forced to turn to analysing the results in the group stages of the World Cup to fill in the time. I was keen to see if there were any patterns in the results that might provide a guide as to the likely winners of the trophy. Could I convince myself that England could lift the trophy and dull the still raw pain of Stoke City’s relegation?

While Germany’s first elimination in the group stages since 1938 had captured the headlines, I was more intrigued by an article I had read that suggested it was becoming harder for countries outside of Europe to compete because of the resources available in the top European leagues. There did appear to be some evidence to support this hypothesis. With Senegal the last African team to be eliminated, there is no African team in the knockout stages for the first time since 1982. By contrast, 10 European countries qualified, a significant increase in the six teams from the continent that made it in out of the group stage in both 2010 and 2014.

... it seems to be about the players ...

This shaped my analysis. Is there evidence that supports the idea that experience in the major European leagues helps explain results in Russia 2018?

The group results do support the suggestion that there is a relationship between successful teams and the number of players in their squads from the four richest and highest quality European leagues (the English Premier League, Germany’s Bundesliga, La Liga in Spain and Serie A, the top Italian division). I used the value of TV rights as a guide to income and the UEFA coefficients for quality to select these four competitions. While Ligue 1 in France has signed a new TV deal from 2020 that will propel it ahead of Spain and Italy, for now the French League is around 40% behind the other leagues on both measures.

Of the 16 teams that qualified for the knockout stages, 13 were in the top two in their group judged by the number of players in their squad from the top four European leagues set out above [1]. This amounts to 80% of the qualifiers which seems to represent a reasonable predictive power – put simply, the teams with the best players did best in the group stage.

The three teams to miss out were Germany, Poland and Egypt. Germany’s failure is the real outlier. On paper, Germany was much stronger than its rivals in the group but, as we know, football is played on grass and Germany’s players underperformed. 

Poland’s elimination is easier to explain with reference to the relationship. Group H was the one with the narrowest gap between the team with the most players form the top four leagues and the least. Poland had 13 players, Colombia 12, Japan 10 and Senegal 8. By contrast, England had 23, Belgium 19 and Tunisia and Panama both had only 1. Before we get too carried away, it is worth noting England had the easiest qualification path judged in terms of quality of opponents.

The final outlier, with only one player from the top leagues, is Russia. Reassuringly, for those of us who love the romance of football, home advantage and togetherness appear to have helped the team. Certainly Russia’s success in the last-16 game against Spain – the only country other than England with all 23 squad players drawn from a top four league – was a committed, battling display. But it may be that Russia has a higher quality squad than the initial number suggest. Twenty one of Russia’s squad play in the Russian league, which is sixth in UEFA coefficient terms and it might be that this league is more competitive than we realise and this has given the players an unexpected edge. It is also the case that Uruguay with 10 players from the top four leagues was way ahead of the group, and Egypt with five and Saudi Arabia with three were not that strong.

… even as the tension mounts ...

The quality of playing squad seems therefore to have been a strong predictor of success in the group stages, but what about the knockout stages? Is this the point at which unpredictability takes over?

The results from the round of 16 would suggest not. Only two teams with the lower number of players from the top four leagues managed to win. One was Russia and even then only on penalties. The other “surprise” was Sweden beating Switzerland.

It is fair to say that Belgium vs Japan and England vs Colombia stretched the model but all the other games went as player numbers would suggest. Uruguay vs Portugal, France vs Argentina and Croatia vs Denmark were close and as the gap in the number of players between the teams was one, two and three in the three games respectively, this is no surprise. Six out of eight games correctly predicted is a success rate of 75% which together with the group stage, suggests player quality is a useful prediction tool.

… and the final looms.

There does appear to be some evidence that money talks and the huge resources that the clubs in the top four leagues are able to deploy is feeding through into performances. The beneficiaries are those countries with the most players from these four competitions.

So what does the analysis of player numbers suggest for the rest of the tournament?

If the relationship identified above continues to hold then France will lose to Belgium in one semi-final and England will triumph over Croatia. Then in a tight final, England will prevail over Belgium. 52 years of hurt could be coming to an end.

Simple! However, it is worth remembering Russia have beaten the odds twice already and should therefore be watched closely as the wildcard. Nevertheless, it seems to me that the BBC’s Team of the Year award is already decided.

Come on England!

[1] It is worth noting, the numbers of players per squad are estimates as the available sources provide slightly different estimates of players reflecting views on loans, timing of the analysis etc. But the numbers do not move by enough to change any of the conclusions on individual teams.