I thought most of a pull request was noise. I measured 1,000 of them. I was wrong.

I'm a CSE student, and reviewing pull requests always felt slower than it should be. My theory was...