Group projects are an important part of preparing software engineering students for the transition to working in industry. However, group projects can be difficult for instructors to monitor, particularly when trying to judge whether students are contributing and collaborating sufficiently. The use of version control systems such as Git in such courses provides an opportunity to analyze code structure through abstract syntax trees (ASTs), rather than relying only on basic metrics such as commits or lines of code reported by tools such as GitHub Insights. In this experience report, we present three case studies drawn from a dataset of 204 repositories from a second-year software engineering group project course. The case studies illustrate the deficiencies of using such basic metrics and built-in repository management interfaces. These tools can aid in the evaluation of students, but often misrepresent their contributions. Further analysis of code repositories and refinement of instructor-facing visualizations show promise in presenting a more representative overview of how students are contributing within group projects, supporting fairer evaluation and earlier instructional intervention. Upon acceptance, we will release the analysis scripts, visualization code, and a small mock repository so that other instructors can adapt the approach.