It may be worth considering whether your intuitions and priors about how serious a problem trolling is is at odds with the impression of the rest of the community. Or, it may be that most of the people you have attracted here are somewhat more tolerant of some amount of trolling. It seems at least from the general voting in this thread that most of the community is not happy with even this change, let alone the other changes you are suggesting.
Biased sample if those who flee the long-replies-to-downvoted-comments threads have already left. At the point where LW starts being unfun for me to read, I panic. If my standards are too high… well, there’s worse things that could happen to a site, like my threshold for alarm being set too low.
Personally, it seems to me that it is, but that it might well be justified anyway. I’m not a big fan of the approach taken, but I’m not yet completely against it either. I’m disappointed that it was implemented unilaterally.
Biased sample if those who flee the long-replies-to-downvoted-comments threads have already left
Valid point. How can we test this?
At the point where LW starts being unfun for me to read, I panic.
Being concerned about the signal to noise ratio is reasonable, but yes this sounds like panicking. Deciding that there’s a problem is not the same thing as deciding that a specific course of action is a good solution to the problem. (I shouldn’t need to tell you that.)
The mental model being applied appears to be sculpting the community in the manner of sculpting marble with a hammer and chisel. Whereas how it’ll work will be rather more like sculpting flesh with a hammer and chisel, giving rather a lot of side effects and not quite achieving the desired aims. Sculpting online communities really doesn’t work very well.
It may be worth considering whether your intuitions and priors about how serious a problem trolling is is at odds with the impression of the rest of the community. Or, it may be that most of the people you have attracted here are somewhat more tolerant of some amount of trolling. It seems at least from the general voting in this thread that most of the community is not happy with even this change, let alone the other changes you are suggesting.
Biased sample if those who flee the long-replies-to-downvoted-comments threads have already left. At the point where LW starts being unfun for me to read, I panic. If my standards are too high… well, there’s worse things that could happen to a site, like my threshold for alarm being set too low.
Do you feel that this is an example of you being intolerant of other posters’ tolerance of trolls? If not, why?
Personally, it seems to me that it is, but that it might well be justified anyway. I’m not a big fan of the approach taken, but I’m not yet completely against it either. I’m disappointed that it was implemented unilaterally.
Valid point. How can we test this?
Being concerned about the signal to noise ratio is reasonable, but yes this sounds like panicking. Deciding that there’s a problem is not the same thing as deciding that a specific course of action is a good solution to the problem. (I shouldn’t need to tell you that.)
The mental model being applied appears to be sculpting the community in the manner of sculpting marble with a hammer and chisel. Whereas how it’ll work will be rather more like sculpting flesh with a hammer and chisel, giving rather a lot of side effects and not quite achieving the desired aims. Sculpting online communities really doesn’t work very well.