Open
Description
In practice comparators are high, the distribution of comparisons is skewed, and there's an interaction between comparator level and performance level. Also, once social comparison tiered-comparators are complemented the distribution of candidates will change.
Detailed examples are needed. And better insight into accepted vs. selected candidates and their scores will help.
The goal is to have moderators for positive and negative gaps relative to various comparators be similar and in the range of 0 to 1.