In football (soccer) information flows through a ball-passing network where some players have predominance over others in which regards the control of information (e.g., some players pass more often or connect pairs of other players more frequently). However, social network studies have been using the concept of shortest-path to model players' interactions. However players do not pass the ball uniquely through their shortest path. When we apply the concept of random walk instead of shortest path we are considering that connections between players are performed in an random way. This is not also true because connections are not random at all. How can I resolve this problem?