I recently worked up a statistical ranking system for club team across Europe, found here, and decided to apply that system to the 32 teams in the World Cup. The ranking system looks at statistics across three categories that I found most correlated to winning. Additionally, in this edition of the rankings, I was able to incorporate the expected goals model from Colin Trainor’s work. I wanted to incorporate these statistics because I wanted to make sure that shot locations were taken into account. For example, the USA gave up the most shots in the group stage but they were all low efficiency shots, while Spain gave up a small number of shots but with a much higher expected goal level. By incorporating the expected goal value of each shot, it allows to better analyze both offensive and defensive performances. The stats used for these rankings are:
- Passing Score: Short Passes, Pass Success Rate, Possession %
- Defense: Goals Allowed, Shots Allowed, Expected Goals Per Shot Allowed
- Offense: Goals Scored, Shots, Shots on Target, Expected Goals Per Shot
Like the rankings for the club teams, all scores are a comparison score with the other teams in the WC. Each category is compared to the other 31 teams, weighted, and put into an overall score for each category, and then a final overall score. These scores allow us to see who performed best in each of the three categories of the game, as well as who was the most consistent in all three. As with the club score, the limitations of this system is that it does not take into account different game states and scorelines, or the quality of opposition. Here are the top statistical performers of the World Cup Group Stage.
Top Passers
| Team | Pass Score |
| Germany | 98.17 |
| Spain | 94.69 |
| Argentina | 94.14 |
| Italy | 91.63 |
| France | 90.00 |
| Bosnia | 84.96 |
| Japan | 84.11 |
| Chile | 83.03 |
| England | 82.41 |
| Ivory Coast | 81.12 |
Top Defenses
| Team | Def Score |
| Belgium | 93.26 |
| Costa Rica | 92.85 |
| Mexico | 86.83 |
| France | 65.09 |
| Germany | 64.19 |
| Brazil | 63.50 |
| Colombia | 57.52 |
| Argentina | 53.83 |
| Russia | 53.74 |
| Netherlands | 52.85 |
Top Offenses
| Team | Off Score |
| Netherlands | 92.49 |
| France | 88.00 |
| Colombia | 78.97 |
| Brazil | 74.92 |
| Switzerland | 73.33 |
| Germany | 71.26 |
| Argentina | 65.63 |
| Croatia | 60.78 |
| Algeria | 60.41 |
| Portugal | 54.15 |
Top Overall
| Team | Total |
| France | 79.24 |
| Germany | 73.81 |
| Belgium | 73.24 |
| Brazil | 71.33 |
| Netherlands | 70.98 |
| Colombia | 68.29 |
| Costa Rica | 67.98 |
| Mexico | 67.61 |
| Argentina | 66.61 |
| Switzerland | 57.91 |
Here are the overall team scores for each team
As always, I wanted to make sure that this ranking system had a strong correlation to winning. Below I have included a linear regression for my total team scores and points earned in the group stage. A respectable R-squared value of .72 will do.
Nelli Analytics 
Amazing information here. Would love to make a post soon and include them with links to you if possible? Also, who do you actually think will win the World Cup?
Absolutely, feel free to use this information and link it back here.
I’m not quite sure who I think will win right now because of how unbalanced the sides of the bracket are. Germany, Brazil, France, Chile etc all on the same side is rough. I’d say France or Germany would be my favorites to advance from that side. The other side is significantly softer, with Argentina and Netherlands really the only two “big” teams. Gun to my head, I would have to say the Netherlands would be my prediction right now.
I think models of this nature need some sore of “tournament composure” index, it seems like some teams always bed in well at tournaments (Germany, Argentina, Brazil) and others (France, England, and, strangely enough, Spain) are historically quite volatile. Of course you’d have to find some explanation for this intuition in the data, perhaps by comparing tournament performance to FIFA or ELO ranking at the tournament (some standardisation for past tournament sizes etc). Of course, you’d have to infer tournament composure for teams that have no prior experience but this time around the only debutant was Bosnia.