|
A waiting period before posting or commenting, and then making a few decisive and prominent examples of corrections before they get out of hand likely goes a long way to preserving the cooperation of participants at making the forum enjoyable. I was just going on in another thread about how curation and moderation are like a cooperate/defect equillibrium or game theory, where the job is to promote cooperators in the mission of creating a quality vibe and isolating or filtering defectors from it - as opposed to measuring individual posts against a set of rules. It may be a deeper idea, as it's less about about enforcing rules and more about rewarding strategies. There was an HN moderator post about how he indexed on the effect of someone's posts, which seems like an immensely efficient way to leverage forum participants' revealed behaviors. Downvotes are a coarse measure, but a flame war starting thread is visually obvious and detecting them scales really well because you can see and even measure sentiment in responses. Flame wars are also costly to the people in them, which makes them an honest and high value signal of the quality of the discussion a comment produces, because they have to spend time and effort to vent spleens, where downvotes are much lower information. I remember thinking it was a very elegant technique. This strategies-and-effect driven approach is different from those Nudge Unit types because the goal is measured by producing a quality of something shared, instead of say, misleading people in service of driving a metric. (they think automatically signing you up for things you have to cancel is wise virtuous, but using lotteries that people actually willingly play are somehow unethical. Behavioral economists are just touts with airs, imo.) We know that complex rules yield poor behaviors and reward arbitrage and defection. Simple rules reduce to being aribitrary, and really, rules generally just produce more rules. Guidelines and public examples (positive and negative) make people self moderate, which itself has shown to scale pretty well (e.g. HN). The Dunbar reference in the article made me think that the best moderation probably doesn't scale much past that. But beyond the Dunbar number, perhaps it's more accurate to say that a forum is no longer a mere thing, but becomes a culture. The thing that doesn't scale is, a forum you can moderate with rules, a community you can moderate with standards, but when a community graduates to a sub-culture, it's likely that this higher level managing its strategies is the way to maintain its equillibrium. |