Hacker News new | past | comments | ask | show | jobs | submit login

> You broke the site guidelines egregiously in this thread with flamewar, snark, and even personal attack.

Please do explain the algorithm that resolves to a True statement here. I can't find a link to the codebase for the sentiment and opinion analysis models.

> We ban accounts that post like this because it destroys what the site is for.

I don't see warnings on other interlocutor accounts. Seems very one-sided, which is of questionable moderation. Not just on this comment thread but many others. I don't see these warnings anywhere. Again, this points to subjectivity. I searched for subjectivity in moderation on the link you provided and it returned no results.

Please can you also share the algorithms used in moderation for HN?

Thanks very much!




The "algorithm" is that I read your comment and saw that it was obviously breaking the rules. I wish I knew how to write software to do that correctly, but I don't.

If you'll review https://news.ycombinator.com/newsguidelines.html with a sincere intent to use this site as intended, it should be easy to see how you weren't doing that.

Re other commenters: everyone always feels like the mods are singling them out personally and treating the other side with kid gloves (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...). That's a universal bias, same as it always feels like the cops are singling you out for a speeding ticket when lots of other cars were going fast too.

If you see a post that ought to have been moderated but hasn't been, the likeliest explanation not that we're against you—it's simply that we didn't see it. You can help by flagging it or emailing us at hn@ycombinator.com. https://hn.algolia.com/?dateRange=all&page=0&prefix=false&qu...


> The "algorithm" is that I read your comment and saw that it was obviously breaking the rules. I wish I knew how to write software to do that correctly, but I don't.

Got it. Maybe I will give it a shot and share it. I'm not sure which types of bots you're already using but there's ways to orchestrate consensus networks on opinion dynamics to derive a "threat" level (or whatever metric name you want to call it). And that could be used to do things like auto-cut tickets to inspect threads for guideline violations.

IMO, there was a misunderstanding about the term "you": there is an individual "you" and a collective "you".

> Re other commenters: everyone always feels like the mods are singling them out personally and treating the other side with kid gloves

I don't feel like I'm being singled out, more like I feel like there is inconsistency in moderation. Probably because it's hard and there are limited resources and the bots that you do have have thresholds set super judiciously to avoid false-positive flagging (or not; I don't actually know).


Ok, I hear you about the word "you" - it's easy to confuse that in general. I often tell people "I don't mean you personally" (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...).

At the same time, if "you" can be misinterpreted as a personal attack, it likely will be - so the burden is on the commenter to disambiguate (https://hn.algolia.com/?dateRange=all&page=0&prefix=true&que...).

Moderation here is always going to be inconsistent, simply because we don't see everything that gets posted, or even 10% of it - there's far too much. If you see a post that ought to have been moderated but hasn't been, the likeliest explanation is that we didn't see it. You can help by flagging it or emailing us at hn@ycombinator.com.




Join us for AI Startup School this June 16-17 in San Francisco!

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

Search: