hmm, I was thinking >The implicit assumption that this is a bad thing is grounded in the assumption that in a large enough set of something there should be con

hmm, I was thinking >The implicit assumption that this is a bad thing is grounded in the assumption that in a large enough set of something there should be considered to be a normal distribution of high quality, medium quality and lower quality members in the set, unless one can show the distribution is biased in some way.