With millions of users and intense interactions among them, Twitch.TV has become the largest online gaming community in history. So far there is not much research studying this new merging and prosperous online community. In this paper, we examine various entities, such as steamers and gaming franchises, in Twitch and how they interact with each other. Using Twitch's API services, we first collect data about Twitch entities and then use this data to answer two interesting questions: i) what game franchises attract similar audiences, and ii) how connected/segregated are different predominate subcultures of gamers. To capture the structural relationships, we classify game streamers into three types and develop two novel community visualization methods. These methods help to gain insight into the connectivity and trends of this online community. Our research uncovers valuable information on better understanding gaming communities.