Key points are not available for this paper at this time.
In parallel to the exponential growth of the gaming industry, video game live-streaming is rising as a major form of online entertainment. Gathering a heterogeneous community, the popularity of this new media led to the creation of web services just for streaming video games, such as Twitch. TV. In this paper, we propose a model to characterize how streamers and spectators behave, based on their possible actions in Twitch and, using it, we perform a case study on the Star craft II streamers and spectators. In the case study we analyze a large amount of data collected in Twitch. TV's chat in order to better understand how streamers behave, and how this new form of online entertainment is different from previous ones. Based on this analysis, we were able to better understand channel switching, channel surfing, and to create a model for predicting the number of chat messages based on the number of spectators. We were also able to describe behavioral patterns, such as the mass evasion of spectators before the end of a streaming section in a channel.
Nascimento et al. (Wed,) studied this question.