I figure you’d audit it by examining the results, and if bias isn’t detectable in the results
I was just asking for more info on how you’d examine the results for bias since it would need to be pretty extreme like the example you gave to be identifiable as something worth investigating.
All good if you don’t have a more specific answer, I was just curious what your personal thoughts were here.
Honestly, they probably like the feeling of “power” (as sad as that sounds) that comes with moderating a community as large as theirs on Reddit.