> On Aug 28, 2022, at 4:05 PM, Ofer Inbar <cos@xxxxxxxxx> wrote: > > People who won't > respond to the list to say they disagree with you, or Keith, because > this wasn't an environment they want to deal with at all. But people do that even if there were no overt insults at all. As far as I can tell there are many people who simply don’t want to deal with the cognitive dissonance that comes from watching people disagree. Or maybe some feel insulted by the fact that some people disagree with them, who don’t change their minds just because others insist that they should. After watching other moderated online discussions for awhile, I started to wonder if moderation actually improves the signal to noise ratio in practice. To me, it didn’t seem to help much if at all. What it seemed to me to do instead was to tilt the playing field to favor the kind of troll who avoided overt insults but who just repeatedly spewed the same talking points. This was not what I wanted to believe and was surprising to me when I first saw it. I’m not saying that moderation is inherently useless or worse, but that I think it’s necessary to observe the actual effects of some particular policy under carefully controlled conditions to have real confidence about what its effects of the policy are. Particularly so when, as in IETF, it’s important that participants be able to speak their minds and respectfully disagree with one another. And of course that kind of research is not easy to do well. I’ve thought about how to do an experiment with the IETF list or a similar list, in which group A would see unmoderated messages and group B would see only messages sent to the same list that passed moderation. Then see which group had more people dropping off the list. One problem would be that in a real discussion, people could tell which group they were assigned to. Another problem is what you really want to measure is not simply a count of unsubscribes, but rather which list was better at producing a technically sound consensus. But maybe others see a way to improve on that experiment. Keith
