I've been averaging a ten-plus point difference on the actual test versus my blind review, and I don't know what I'm doing wrong.