World Cup and Graphs
With the World Cup 2022 coming around this winter in Qatar, let’s see how graph theory, or at least graph construction and how conclusions can be made from it to predict results. Predicting football or any sports match outcome has been a topic of interest for a very long time, especially since technically people can make profit from it. Although the author acknowledges that the method suggested by it is not 100% accurate, it does indeed seem to have good reasoning behind why it may be of use in making a more educated guess. The concept is simple; each node is a player and each directed edge is the path of the pass from one player to another. Given the graph, the author gives insight into what to look for. First is looking at total passes by a team to figure out the weight of each node, which will tell us which team is better at play-making. The second is looking at which team has higher weights at mid-level nodes, which indicates strong build-up play. The third is looking for a team with fewer revisited nodes, which signifies fewer redundant passes and more attacking potential. The fourth is on loops, where fewer loops imply a high press build-up. And the last is regarding the direction of the passes/edges, where the team is attacking more if there are many passes that are directed toward the opposition goal.
This is an example of how the construction of graphs can simplify the phenomenon, a game in this case, and makes it easier to interpret. An example of another interpretation may be, if there were many long edges (from defender/goalkeeper to striker), it would mean that the team preferred a more direct, long ball style of attack rather than building up. Or as mentioned above, if there is a player with a high weight value, that would mean that the player is the main play-maker and should be marked properly to cut out their offense. This would be similar to cutting off a “well-connected” node from a graph. Although these analyses seem useful, again note that this is a post-match analysis of the game and that football games are between two very different pairs of teams so the graphs may be very different per game for a team.