Bad replies aren't necessarily the least popular. Any number of people would prefer blue check (basically signed) replies than pseudonymous drivebys whose entir

Bad replies aren't necessarily the least popular. Any number of people would prefer blue check (basically signed) replies than pseudonymous drivebys whose entire participation in twitter is sharp replies that get massively upvoted. Also, popularity based in upvotes from nobody, unverified accounts is usually inorganic. Scoring upvotes based on how many of them are from verified or likely authentic accounts can only improve user experience (for content providers, not professional reply guys.)