The science on rating systems is inconclusive.115 There’s no strong evidence to suggest that having three or five or ten or fifty rating points makes a difference. Our 41-point scale came out of our engineering DNA. It felt satisfyingly precise to be able to draw a distinction between a 3.3 and a 3.4 performer. And when you averaged ratings over many quarters, you could calculate ratings finely enough to sort a 3.325 from a 3.350 performer.
