Benchmarking a Bug Scanner

We ran a tournament pitting Detail's findings against thousands of comments from code review bots.