Introduction:
Sports are increasing in popularity with many sports teams with millions of fans across the world. Every fan wants to see their team succeed and win championships. In order for sport teams to improve and have chemistry coaches hold practices with drills so that the team can improve as a unit. This article goes over team sports performance analyzed through social network theory and how this can be used to design more impactful practices to improve team performance
Analysis:
In team sports functional performance is based on a complex network of social interactions which are established among teammates. In team sports weighted graphs indicate how strong certain interactions are between players. These team sport graphs are also directed since in most team sports there is a concept of passing the ball (soccer, basketball) so when player 1 passes a ball the player 2 that is a directed edge from player 1 to player 2. An example of a social graph for a soccer team is show below:
Graph of soccer team in a tactical 1-4-3-3 formation. Black circles represent players and blue arrows represent pass direction, so the origin of the arrow represents the player that passed the ball and the destination of the arrow represents the player that received the pass from the player at the origin of the arrow. The width and colour of each arrow represents the quantity of pass completed between players (thicker arrows indicate a greater quantity of passes between players). Size of circles represent players who participate more in attacking plays (bigger circles represent players who receive and perform more passes)
The graph can also be represented as an adjacency matrix with weights as follows:
Studies of team sports have demonstrated that some network properties can be related to team performance. In the case of social networks for sports team performance is a goal-oriented process of sharing information through passing the ball. One network property being a characteristic path length measures the separation between two vertices. In our case those vertices are players. So, characteristic path length can show how many passes are needed to go from one particular player to another. Clustering Coefficient as taught in class measures the degree to which nodes in a graph tend to cluster together. Clustering coefficients on the graph of sports teams can show subgroups of players who coordinate their actions more and often pass the ball to each other. High values of clustering coefficients can indicate a team’s ideologies to form functional clusters with players creating tight groups which comprise high-density ties. Degree of the directed graph of sports teams can indicate how many passes each player receives (in-degree) and how many passes each player makes (out-degree) ; this can help coaches identify “ball-hogs”. Betweenness as taught in class is the number of shortest paths passing through an edge. For a network for sport teams this can represent the amount of network flow a player controls. For the example of soccer it can show players that are often connecting defense with the midfield. Closeness of a vertex is defined as the sum of shortest path distances from all other vertices in the graph. This metric can be used for a sports teams network to provide insight into the adjacency of one play to another on the team. Players with low closeness score are adjacent to others making that player receive a pass or rotate with the nearest player more rapidly. These metrics which are obtained from the network of the sports team will inform coaches and performance analysts about the functionality of team organization and if team organization relies on just a few key players if there is a significantly unequal interaction between players. This information can prompt coaches to modify gameplans and practice drills which will improve team organization. For example, promoting the influence of different player subgroups within the team during games to balance the team effort which may lead to more success.
Conclusion:
Sports teams can be represented by a social network to evaluate performance in training and competition. Analysing the social network for sports teams can unravel many useful metrics which can provide coaches with information to form concrete practices and game plans which if done correctly can significantly impact performance of the team!
References:
Ribeiro, J., Silva, P., Duarte, R. et al. Team Sports Performance Analysed Through the Lens of Social Network Theory: Implications for Research and Practice. Sports Med 47, 1689–1696 (2017). https://doi.org/10.1007/s40279-017-0695-1