I'm distinguishing it between the "intended" way of using the site by making posts and then the community organically votes on it, and some unintended "manipulated" way where content is artificially voted upon or otherwise given an advantage over other posts.
They are trying to train their user base to moderate content correctly. If it's against the sites policy, users should down-vote the content and report it.
I don't think it's quite malicious but if I'm a mod and it's common to suppress anything election related then I could see that applying here. But according to the comment you linked their algorithm only targets upvote/downvote wars, I did not observe that happening here at all.
Considering this is a low quality listicle without much content and is posted by a new account the same username as the domain, it seems somewhat likely that it's using fake votes to get as many points as it did?
Yup, it's manual intervention by dang or other mod. Happens every day so you shouldn't be surprised. Front page is not entirely what we vote for, it's also a function of what dang likes and dislikes.
In fact, one of those posts were by the submitter of this post. On the surface, this looks like a way to game the system of collecting points by posting things that are bound to be voted up a lot.
The methodology here is quite simplistic, and suspect.
The way they “objectively” decided on “quality” of posts, by which to judge the ‘before’ picture of a poster was via some machine learning analysis of the keywords in the post in question, implicitly assuming that their keyword analysis was a better judgment of quality than the explicit voting on the site. This seems like a very poor assumption to me.
It’s quite plausible (to me) that the posts’ keywords in the negatively-voted posts looked okay to their algorithm while the content still turned out to be trolling or bullshit (hence the downvotes). It would be quite reasonable to assume that posts full of trolling or bullshit were by posters who were inclined to future trolling/bullshit posts.
I’d be very interested to see what results they’d get if they ran the clock in reverse. I.e. pulled some posts of similar “quality” based on their metric but different voted scores, and then looked at the several posts before that, from those posters.
I'll admit it bothers me a bit. I don't have a problem with people who participate here submitting their blog posts but as a fairly active member of this community I'm a little offended by someone who doesn't contribute in any other way.
That said I don't suspect foul play here simply because someone who was using voting bots or some other form of cheating would probably do it more often (most of this guy's post never get above 1)
We work similar to HN as we have spam rating on votes and devalue and derank posts based on those.
Essentially if you ask for upvotes there is a (almost too high) chance that our system will notice and punish your post for this. Sadly a lot of really good products drop because of this.
Regarding those blogposts - we don't really speak up if they are right or wrong (for obvious reasons) but those kind of articles exist for any system/website/mechanic/process where users believe they can get a benefit if they manipulate.
The article is pretty light on how these platforms do so.
Do they not simply bubble up popular posts: ones with likes, retweets, and comments? That might mean a politicized post, but it also could be a positive one about the kindness of a stranger or some cute picture of your kids.
These platforms may well be incentivizing enraging posts. I just don't know how they're doing it, and this article did not help me understand it.
Seems that way, but it could also have been automatically penalized for other reasons. You'll never really know because there's little transparency into how posts are ranked and penalized. Also, I've heard (and assume) mods have the power to penalize posts so mods may influence rankings without anybody knowing.
It doesn't happen mechanically, but I think you get about equivalent social behavior - for example Metafilters orders posts by time rather than popularity but there's a great deal of social policing that goes on there and piles upon opinions that are at variance with community norms.
The specific types of posts that are treated specially are well documented, including in this thread. It's 'suspicious' mostly because you (seemingly) haven't had a chance to look into it.
reply