I'm afraid there isn't much written about the buttons we're speaking about there, but this tidbit may be pertinent: "Users should vote for a story because they

I'm afraid there isn't much written about the buttons we're speaking about there, but this tidbit may be pertinent: "Users should vote for a story because they personally find it intellectually interesting" Granted, we are talking about comments, not stories, but presumably the same logic holds? What do you find intellectually interesting about low value content? Doesn't the premise of downvoting "bad" content violate the notion of it being intellectually interesting, thus violating the intent of the voting system? But let's assume that comments are meant to be treated differently. All we really have to go on with respect to comments is from pg[1]. He indicates that the buttons have no meaning; use it however you please. He only considers it abuse if you press the button without reading the comment first, which, again, goes back to it being an analytics system. In fact, there already exists a flag button for dealing with "bad" content which doesn't reveal analytics data to the user. pg was an early pioneer in developing methods to stop spammers[2], so no doubt he, more than most, understands the importance of not leaking participation data to spammers. If they know you're reading, they're going to keep on doing it. Only when they think their audience has disappeared will they stop. And, so, downvoting their content is ultimately nonsensical, even if nobody is going to ban you if you do. Given that the buttons have no meaning, there is no policing. It is fundamentally impossible to determine your voting intent by observing your action of pressing an unlabelled button. But there is an assumption that you will use the buttons in good faith around content you find interesting, not to feed the trolls. [1] https://news.ycombinator.com/item?id=117171 https://news.ycombinator.com/item?id=117171 [2] http://paulgraham.com/spam.html http://paulgraham.com/spam.html