Network Theory in Soccer: How Teams Pass in the World Cup

In this paper, the authors apply network theory to soccer analysis. They focus specifically on how players on a given soccer team pass the ball. The graphs have edges directed the way the ball travels. Very conveniently, the graphs are organized to reflect the general location of the players based on the position they play. This means the graphs show not only who passes to whom, but where on the field the passes are coming from and going.

Besides showing who on the field is connected, the authors also show how strong the connections are. The thickness of the arrows between players is based on how many passes each player passed to each other. Because Player A might pass to Player B much more than Player B passes to Player A, the arrows between two players can be of different thickness.

While it’s interesting to look at the graphs, it’s also important to assign some numbers to the patterns observed. The authors of the paper define three statistics that describe a player’s importance to the passing. The first, closeness, measures how easily a player can be passed to. The second, betweenness, is the percentage of shortest paths that go through a player. The last one, pagerank centrality, is a measure of popularity based on the number of passes a player receives and who passes to him.

The authors got their data from the 2010 FIFA World Cup. Although they made tables for the final 16 teams, they only made graphs for the teams that reached the finals, Spain and the Netherlands. Interestingly, out of the 16 teams that made it to the knockout stage, Spain and the Netherlands had the highest average clustering coefficient, meaning that their passing was balanced. It involved the whole team, not just a few players. Their style is one other teams might want to consider because it allows for players having bad games. If a team relies on two or three key players and one of them isn’t playing well, that team will most likely lose.

The paper being referenced

A slideshow made by the authors for a presentation

Both of these links are to pdf’s, so depending on your browser, you might end up downloading them. Google Chrome opens them in new tabs.

3 thoughts on “Network Theory in Soccer: How Teams Pass in the World Cup”

  1. This was an interesting find on your part. It strengthened the notion in FOOTBALL that ball movement is generally through the central midfielder, which is why players at that position are often so well paid and highly regarded. Players like Gerrard, Sneijder, Xavi and Pirlo come to mind.

  2. It always makes me very happy when someone refers to the sport with its real name ie. Football. But, I agree that this is a very cool find! If anything, when trying to understand the theory in class I’ll just picture players and a football from now on.

    What I find most interesting is that the winner, Spain, and other close competitors (such as Netherlands, Argentina and Brazil) has the highest number of passes, clustering and size of clique. It also has a high-end edge connectivity, while keeping a low betweenness score. Is this the recipe for a FIFA win? Something that I often wonder about sports teams is whether or not they even like eachother as more than acquaintances/ teamates and if they are actually friends. It would be interesting to see whether that aspect, apart from simply game strategy has an impact on overall team performance.


  3. Anyone who has played football before knows the importance of communication when it comes to establishing team chemistry. It would be interesting to compare a players national results vs his club results because, aside from some exceptions, you are generally surrounded by different players when playing for your club than for your country. One major exception to the study would be a team like Barcelona who’s club team is essentially their national team. An example on the other side of the spectrum could be the Dutch national team whose players, relative to Spain’s, apply their talents outside of their home nation much more frequently. So, if you could look at a Dutch player’s performance in say England, and than compare his results to those of when the same dutchie plays for his national team, theoretically you could quantify the importance of cultural connectivity within competition or even to lesser extent as it is inherently hard to quantify, the importance of communication when playing football and how this may or may not be connected to performance.

Comments are closed.