When doing a study on a large number of video games, it may be difficult to cluster them into coherent groups to better study them. In this paper, we introduce a novel algorithm, that takes as input any set of games S that are released on Steam and an integer k, and cluster S into k groups. Each group is then assigned a distinctive name in the form of a Steam tag. We believe our tool to be valuable for gaining deeper insights into the video game market. We show that our algorithm maximises an objective function that we introduce, the naming score, which assesses the quality of a clustering and how distinctive its name is.