This project explores the impact of collaboration patterns on the performance of videos and the individual growth of creators on Bilibili, a popular Chinese video-sharing platform. Collaboration videos, where two or more creators work together and share the video on their own accounts, are becoming increasingly popular on the platform. By analyzing the collaboration patterns of creators and the performance of their videos, this study aims to identify which types of collaborations lead to the most successful videos and the greatest growth for individual creators. The project uses a data-driven approach, collecting and analyzing information on collaboration frequency, roles, and video performance. The results of this study will provide insights into how creators on Bilibili can best collaborate to create successful videos and grow their individual presence on the platform.