Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Tell HN: There is a flame-war detection system
8 points by ColinWright on Jan 27, 2020 | hide | past | favorite | 5 comments
I'm seeing this often enough to think it's worth telling people about.

HN has a an automated simple proxy for flame-war detection. One characteristic of flame-wars is that people get into a to'n'fro over the issue, and the number of comments balloons. However, no one else is really interested, so they don't upvote the submisson, and the participants can only upvote a submission once, so the number of points doesn't increase.

Result is that a simply proxy for a flame-war is the number of comments on a submission out-stripping the number of votes.

A side-effect of that is that if there's a mildly interesting submission that lots of people comment on, but very few upvote, then the flame-war penalty will be triggered, and the submission will sink like a stone, never to be seen again.

Like this one: https://news.ycombinator.com/item?id=22158218

It's a reasonably interesting collection of suggestions and comments from a fairly wide range of the HN community, but it's now lost, because while people commented, they didn't upvote.

So this is a public service announcement ... if something is interesting enough to comment on, consider upvoting as well, so others can see your comment, and you can then see theirs.



> Result is that a simply proxy for a flame-war is the number of comments on a submission out-stripping the number of votes.

I would be more interested in an echo-chamber detection system. A simple heuristic could be down votes greater than 3x the number of comments. The idea behind an echo chamber being that a post or comment is removed from circulation via excessive down votes without any corresponding commentary qualifying that level of voting.

The result of echo chamber detection would be:

* Presence of an echo chamber, such that the comments present don't qualify the quantity of down votes or that themselves present hostility or greater than tranquil disagreement. This is the online equivalent of a heckler's veto.

* Presence of a contribution that is probably deserving of moderation due to hostility more than disagree-ability.

On HN this would rarely work, because an echo chamber scenario results in a maximum of 4 down votes.

I have found that echo chambers are a group behavior that feeds itself in that some users contribute to the group's hostility simply out of conformity or other mental laziness. They may even leave evidence of such when their comments clearly indicate they are advocating for a position without reading the thread they are commenting into.

Responding to echo chambers in unexpected ways would make for interesting social experiments. Consider what a hostile group would do, for example, if the group hostility instead raised the presence/visibility of the contrary contribution and the vote count is artificially exaggerated.


Clickable: https://news.ycombinator.com/item?id=22158218

(This post, and this link, partly so I'll be able to find that submission again, and its friend from earlier https://news.ycombinator.com/item?id=22105561 )


I'd been wondering why the submission you used as an example dropped off the front page so quickly this morning. I'm upvoting it and yours now. Thank you.


In a flame war the ratio of the number of people who commented to the actual number of comments would be greater as usually the people who feel strongly about the topic would comment more often. Wouldn't this be a better metric than simply number of likes to comments?


Speaking as one who has implemented such things (in completely different contexts) ...

Possibly yes, but it's harder to compute, and the improvement might not be enough to make it worthwhile. This metric is trivial to compute, doesn't require crawling over the entire comment tree, and is "Good Enough(tm)".

The cases where it does wrong is where lots of people chip in with drive-by comments, but don't bother to upvote the submission. If it's interesting enough to comment, why is it not interesting enough to upvote? And why are non-commenters not upvoting?

Personally I find the dynamics here on HN baffling, but again, as a simple member of "the community", I suspect this metric is best "Bang for Buck".

(... edit ...)

As a case in point, the item about bathrooms in houses[0] has over 200 comments, but fewer than 100 upvotes. There are 144 unique commenters, and roughly half have upvoted.

[0] https://news.ycombinator.com/item?id=22157414




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: