4 ms·

We read all comments that get sufficiently negative scores. So I have a decent handle on how bad that kind of bad comment is. What worries me more is bad (= m

by pg 15y ago

We read all comments that get sufficiently negative scores. So I have a decent handle on how bad that kind of bad comment is. What worries me more is bad (= mean and/or stupid) comments that actually get upvoted.

redthrowaway 15y ago
Are you most worried about snarky or rude comments occasionally getting upvoted, or the seeming proliferation of content-free comments? I ask because the rude/mean upvoted comments, though disturbing, are thankfully fairly rare. The content-free comments, on the other hand, seem to be becoming more frequent and less downvoted. Do you think the problems are distinct, or do they both follow from some other change in the community?
tokenadult 15y ago
What worries me more is bad (= mean and/or stupid) comments that actually get upvoted. Thanks for explaining what data points you and the curation team observe closely, and what trends you have noticed. You mentioned above about the goal of increasing comment quality that "There's a lot we already do about it, and probably other techniques that are yet undiscovered. This is somewhat uncharted territory." I'll renew a suggestion here that I think would be helpful. I've been a moderator one place or another online since 1993, and I've seen a lot of online discussion forums attempt to acculturate newbies to civil, thoughtful discussion. One of the most helpful practices that I have observed over the years is making sure that people who use factual assertions to back a position in a discussion are accountable for where one would look up the facts to verify them and ensure they are not made up. (A HUGE percentage of all real-world discourse about controversial issues is based on made-up factual premises.) With that in mind, I'm trying to encourage a culture here of asking people for sources when they make a factual assertion, and providing sources in a first comment when there sources on the issue. (I try to stud most of my comments with links to sources to exemplify this.) So any time any of us see a comment that is dumb, an essential thing to do is to ask the person who made the dumb comment what the basis for the comment is. ("Do you have a source for that?" or the more abrupt "[citation needed]" as occasion demands, or "What about [detail of assertion]?" if that is better for raising the quality of the discussion.) Another thing to do is to share the growing online literature about evaluating sources, such as my very favorite link for Hacker News discussions, http://norvig.com/experiment-design.html http://norvig.com/experiment-design.html and links about constructive online participation. http://www.paulgraham.com/disagree.html http://www.paulgraham.com/disagree.html A lot of stupid comments (and, alas, I have a particular user's example in mind) result from the convenience of reading blogs rather than reading serious scholarly literature. And the paywalls around serious scholarly articles only exacerbate this problem. So I routinely upvote any comment that points to a reliable source, and especially comments that teach other HN participants how to find and access reliable sources. In general, promoting an information-based culture here will have two effects that will help reduce the problem of dumb comments: (1) it will raise the karma and visibility of the better comments that include good information from reliable sources (or that ask for that as follow-ups to worse comments), and (2) it will be off-putting to people inclined to make it up when they are put to their proof for the things they say.

by pg 15y ago

redthrowaway 15y ago

Are you most worried about snarky or rude comments occasionally getting upvoted, or the seeming proliferation of content-free comments? I ask because the rude/mean upvoted comments, though disturbing, are thankfully fairly rare. The content-free comments, on the other hand, seem to be becoming more frequent and less downvoted. Do you think the problems are distinct, or do they both follow from some other change in the community?

tokenadult 15y ago

What worries me more is bad (= mean and/or stupid) comments that actually get upvoted. Thanks for explaining what data points you and the curation team observe closely, and what trends you have noticed. You mentioned above about the goal of increasing comment quality that "There's a lot we already do about it, and probably other techniques that are yet undiscovered. This is somewhat uncharted territory." I'll renew a suggestion here that I think would be helpful. I've been a moderator one place or another online since 1993, and I've seen a lot of online discussion forums attempt to acculturate newbies to civil, thoughtful discussion. One of the most helpful practices that I have observed over the years is making sure that people who use factual assertions to back a position in a discussion are accountable for where one would look up the facts to verify them and ensure they are not made up. (A HUGE percentage of all real-world discourse about controversial issues is based on made-up factual premises.) With that in mind, I'm trying to encourage a culture here of asking people for sources when they make a factual assertion, and providing sources in a first comment when there sources on the issue. (I try to stud most of my comments with links to sources to exemplify this.) So any time any of us see a comment that is dumb, an essential thing to do is to ask the person who made the dumb comment what the basis for the comment is. ("Do you have a source for that?" or the more abrupt "[citation needed]" as occasion demands, or "What about [detail of assertion]?" if that is better for raising the quality of the discussion.) Another thing to do is to share the growing online literature about evaluating sources, such as my very favorite link for Hacker News discussions, http://norvig.com/experiment-design.html http://norvig.com/experiment-design.html and links about constructive online participation. http://www.paulgraham.com/disagree.html http://www.paulgraham.com/disagree.html A lot of stupid comments (and, alas, I have a particular user's example in mind) result from the convenience of reading blogs rather than reading serious scholarly literature. And the paywalls around serious scholarly articles only exacerbate this problem. So I routinely upvote any comment that points to a reliable source, and especially comments that teach other HN participants how to find and access reliable sources. In general, promoting an information-based culture here will have two effects that will help reduce the problem of dumb comments: (1) it will raise the karma and visibility of the better comments that include good information from reliable sources (or that ask for that as follow-ups to worse comments), and (2) it will be off-putting to people inclined to make it up when they are put to their proof for the things they say.