Films and television shows provide a rich source of cultural data and form an integral part of modern life. However, the video medium remains difficult to analyze at scale effectively and its study has generally attracted limited research attention. We propose a method of summarizing the audio and visual aspects of entertainment videos, through the automatic extraction of dominant colors from video frames and textual categories from dialogue. The colors and dialogue are displayed in a visualization that allows the user to explore the video, highlighting both high-level and low-level patterns from the data. Focusing on the hit television series Game of Thrones, we show how our visualization supports the detection of scene changes and plot points, providing a new perspective for both scholars and casual viewers.