I’ve said this elsewhere, but I rather suspect that downvoting stuff doesn’t decrease its visibility much. I haven’t actually applied a regression to the data (though now that I think of it, that’d be an interesting problem), but eyeballing vote totals on my replies heavily downvoted vs. comparably upvoted posts at the same level of the comment tree, I don’t see much difference. That implies that about as many readers are expanding the tree as would follow it normally.
It seems obvious to me that, at least in theory, stuff can get upvoted that we don’t want here, for reasonable values of “we don’t want”. Do you disagree with this? (And in practice, I think this particular post is an instance of such stuff.)
Not in principle, but the main class of stuff that would get upvoted but which the community wouldn’t reflectively want around is stuff that exploits some kind of short-term bias. Flattery, tribal politics, that sort of thing. I don’t see a good case for putting this post into that category.
A flat up/down voting system is insensitive to degrees of dislike, too, and if the response to a post is highly uneven—lots of lukewarm positive responses and a few very extreme negative ones, say—then dropping it might be good for the forum despite positive karma. There might be a stronger case for saying this is such a post (though they should generally be rare), but on the other hand this line of reasoning leads to some nasty strategic effects downstream; I think we ought to be extremely cautious about using it as justification for removal.
I’m aware of that. And indeed we don’t get the massive dogpiles on bad ideas that we used to, or I’d expect downvoting to increase their visibility. But the argument from vote totals still seems to apply.
Probably applies better to downvoted comments than to top-level posts, though. It’s much easier to expand a collapsed comment thread than to notice the presence of a post that’s been voted off the sidebars, and while curiosity might be a motive for the former I don’t think it’s sufficient for the latter.
But the argument from vote totals still seems to apply
I am not sure—you have an unobservable characteristic of a post, let’s call it propensity to elicit replies. It is likely to be correlated with how controversial the post is. A controversial post is likely to get both many replies and many downvotes (as well as many upvotes). On the other hand a milquetoast post will get neither replies nor downvotes.
I’ve said this elsewhere, but I rather suspect that downvoting stuff doesn’t decrease its visibility much. I haven’t actually applied a regression to the data (though now that I think of it, that’d be an interesting problem), but eyeballing vote totals on my replies heavily downvoted vs. comparably upvoted posts at the same level of the comment tree, I don’t see much difference. That implies that about as many readers are expanding the tree as would follow it normally.
Not in principle, but the main class of stuff that would get upvoted but which the community wouldn’t reflectively want around is stuff that exploits some kind of short-term bias. Flattery, tribal politics, that sort of thing. I don’t see a good case for putting this post into that category.
A flat up/down voting system is insensitive to degrees of dislike, too, and if the response to a post is highly uneven—lots of lukewarm positive responses and a few very extreme negative ones, say—then dropping it might be good for the forum despite positive karma. There might be a stronger case for saying this is such a post (though they should generally be rare), but on the other hand this line of reasoning leads to some nasty strategic effects downstream; I think we ought to be extremely cautious about using it as justification for removal.
Huh? Once the post drops below −4 net, it becomes really hard to see it and costly (5 karma) to reply to it. There is a clear threshold effect.
I’m aware of that. And indeed we don’t get the massive dogpiles on bad ideas that we used to, or I’d expect downvoting to increase their visibility. But the argument from vote totals still seems to apply.
Probably applies better to downvoted comments than to top-level posts, though. It’s much easier to expand a collapsed comment thread than to notice the presence of a post that’s been voted off the sidebars, and while curiosity might be a motive for the former I don’t think it’s sufficient for the latter.
I am not sure—you have an unobservable characteristic of a post, let’s call it propensity to elicit replies. It is likely to be correlated with how controversial the post is. A controversial post is likely to get both many replies and many downvotes (as well as many upvotes). On the other hand a milquetoast post will get neither replies nor downvotes.