> I'm sure you could run the same experiment with e.g. prestigious math journals. There is probably a significant overlap between authors and reviewers. In fac

> I'm sure you could run the same experiment with e.g. prestigious math journals. There is probably a significant overlap between authors and reviewers. In fact, how else are you going to find qualified reviewers but to find people who were reviewed well? Does it make sense to get people whose works are bad to review other people's?