Let’s face it, not all reviews are of the same quality or utility. And since there are so many, to help readers prioritize we give each review a ‘ranking’ to help separate the best from the rest.
Our review rating scale is from 8 (best) of 1 (worst) – but we only use 1, 3, 5, and 8 – so there are nice clear distinctions between the quality of reviews at each level. Of course this is subjective; we do our best.
While there is no definitive meaning to the levels, functionally:
8 = Outstanding
5 = Substantial
3 = Minor
1 = Limited Value
The idea is that each move up the scale is roughly* twice as useful as those at the level below it. The goal is that all reviews at the same level are grouped and all the reviews at any level are distinct from those at the levels below and above.
To be clear, this is a ranking of how interesting / useful / important the review is, not the reviewers rating of the work.
(* this is based on the Fibonacci sequence and methods used to rate tasks in planning methods used in agile development.)