9 ms·
Hi dang, I want to say first that I think you do a great job of moderating the site, which I imagine is an enormous and difficult task. Is it always "enough" f
by eindiran 7y ago
Hi dang, I want to say first that I think you do a great job of moderating the site, which I imagine is an enormous and difficult task.
Is it always "enough" for users to flag an article for it to be removed? I remember the highly contentious Amelia Bassano article from a few months ago and in that case, the article wasn't removed (perhaps by your specific intervention). Given that this article seems to be on-topic and was rapidly gaining points (20ish in the first 30 minutes after being posted), it seems like the community was legitimately interested in the topic. I'd go so far as it say that this seems like a case of flag-based brigading to prevent the story from gaining traction. So I guess my question is, when is flagging by users enough to kill a story and when is it not enough?
- dang 7y agoNot always. We turn off flags in some cases, if the article is intellectually interesting and substantive and there's a chance the community can discuss it sanely. You can't go just by points, unfortunately. Users upvote lots of things that are either off-topic or otherwise bad for HN. Indignation routinely attracts upvotes [1], and not every riler-upper makes for a good HN submission; most don't. The site guidelines, flagging, and moderation are necessary counterweights to the voting system, which can't regulate itself. HN can't live by upvotes alone. Life would be easier if it could! In this case, I took a quick look last night and only saw the baity title ("How Silicon Valley enables online slave markets") and a video. That's a terrible combination. Most users won't watch a video, and the combination of that and provocative subject matter makes a shallow, divisive discussion inevitable. I missed that there was a substantive article below the video, though. At least I assume I missed it, since it seems unlikely they would have added it later. That, together with the fact that there's a better title now, makes a significant difference, so I'm going to turn off the flags on one submission of that story (there were several) and put it in the second-chance pool (described at https://news.ycombinator.com/item?id=11662380 https://news.ycombinator.com/item?id=11662380). Please do not post about "brigading" and similar abuses unless you have specific evidence of abuse. This is an important site guideline [2]. In this case it was clear from the few comments people did post that some users objected to the story—or more likely the title—as unsubstantive. I know you disagree with that, but other users having an opposing view is not an abuse, it's the way any sufficiently large population sample works. Internet users are too quick—orders of magnitude too quick—to jump to the conclusion that sinister manipulations like brigading or astroturfing or government disinformation are responsible when they see things they don't like on the internet. This has a poisonous effect on discussion, so we ask people not to do it on HN. Lots more explanation at [3]. Also, if you'll review the site guidelines, you'll note that they ask you to email hn@ycombinator.com instead of making a post like this on the site. For one thing, if you post on the site, odds are we won't see it. We don't come close to seeing everything that gets posted here, and I only saw this by chance. But also, it's off topic and repetitive to make posts like this to HN itself. A story getting flagged is not an "interesting new phenomenon", to quote the guidelines again. 1. https://hn.algolia.com/?dateRange=all&page=0&prefix=true&query=by%3Adang%20indignation%20upvotes&sort=byDate&type=comment https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que... 2. https://news.ycombinator.com/newsguidelines.html https://news.ycombinator.com/newsguidelines.html 3. https://hn.algolia.com/?sort=byDate&dateRange=all&type=comment&storyText=false&prefix=true&page=0&query=by:dang%20astroturf https://hn.algolia.com/?sort=byDate&dateRange=all&type=comme...