In many sports, coaches and athletes rely on annotated video footage to improve player performance, reduce the risk of injury, and analyse the game. This is also the case for rugby. Stakeholders at all levels, from amateur to professional, invest a lot of time and resources into annotating these games. With video quality and camera angles in the amateur games not being at as high a standard as the professional ones, there can be difficulty in identifying certain labels. We would like to help fix some of the problems. Challenges from the amateur footage arise from varying lighting and weather conditions, as well as player occlusion (especially when games are recorded from a single camera angle). From the work we have done so far, identifying the player position has proven to be one of the most difficult labels to manually annotate. We would like to be able to create a tool that could keep track of their positions automatically. This project would involve researching and applying computer vision techniques to real-world data to see how they behave, as well as designing a method for keeping track of which players are on the pitch and whether they have switched positions, etc. The ideal candidate for this would be someone who has an interest in sports, computer vision, and tackling real-world data issues.