Thank you for your thoughtful response. As long as you don't hold that credibility can only come from external accreditation, and that in some categories it doesn't make sense to talk about credibility at all, I'm far less concerned about the potential fallout from such a change. As for the problem you're solving for, have you considered crowd sourcing it? For example, what if you allowed each Medium member one daily "golden ticket" vote, where we identify our favorite article that we read that day (that we didn't write, and that we indeed read that day). But also, one "thumbs down buzzer" for what we felt was the worst piece we read that day. In both cases, I'd use machine learning to identify bad-faith voters, who try to game the system. I'd then take away their voting right, or at least significantly lower the weight I give their votes. Hopefully, over time, this system will pick out the writers who consistently do well by their readers, and the ones who regularly resort to click-baiting without delivering on their promises.