This seems like it mostly tests whose code has the most test coverage, not which lines of code are most important?