How were influencers identified in this case study?

We used conventional SNA techniques to identify influencers in our dataset. Given that there is no agreement on which individual measure to utilize when identifying influencers, we used six different measures of centrality:

  1. Closeness

  2. Betweenness

  3. Eigenvector

  4. Indegree

  5. Outdegree

  6. Page-rank

All the sets of players identified by each centrality measure are intersected with each other to identify the players that are considered central for each of the six measures. Influencers are players who satisfy all these six conditions.

We then plotted the resulting influencers onto a network graph where the nodes represent the players, and the color of a node indicates the community (module) the node belongs to. The resulting super graph is depicted in the figure below. The size of the nodes is proportional to the importance of a player. Hence, influencers display a much bigger size than normal players.

Get hands-on with 1200+ tech skills courses.