Analysis of Players Transfers in Esports. The Case of Dota 2
In this work, we analyze how the esports transfer market is organized with the help of mixed methods. We assume that a combination of Social Network Analysis and Machine Learning can help to achieve deeper understanding and to find patterns which are hidden from the one-side analysis. For the research, we gathered information about transfers of Dota 2 teams made between The Internationals of 2016 and 2017 and built a network based on this data. For the ERGM, we checked the importance of belonging to one region and organization, difference of skills, and participation in TI, and for association rules, on a par with the regions, we added players roles, and a metric of their personal performance -- fantasy points. Summing up the results, we found out the importance of homophily within regions, detected presence of vertical mobility, and discover the influence of the specific players roles.