Tell HN: HN is not immune to manipulation

12 points by blunte 4 years ago | 13 comments

It started innocently. There was (is) a submission from phys.org about birds and windmills which caught my attention. Ironically, it was the odd article which didn't provide links to the actual research or paper it mentioned which caught my eye.

Then in the comments, I saw a strange post from an HN reader (who was also the submitter) which merely pasted an intro paragraph from the phys.org page. The comment and quote added nothing other than a visible "1 comment" to the post.

Looking at the submissions from this user, I saw dozens of submissions each day for the last several days. Many had a particular slant related to a particular ongoing war.

Moderator dang is one person. Is there any way we can keep HN from becoming drowned in garbage?

yesenadam 4 years ago |

> Is there any way we can keep HN from becoming drowned in garbage?

Do as the guidelines recommend:

> Please don't post insinuations about astroturfing, shilling, bots, brigading, foreign agents and the like. It degrades discussion and is usually mistaken. If you're worried about abuse, email hn@ycombinator.com and we'll look at the data.

esperent 4 years ago |

> Is there any way we can keep HN from becoming drowned in garbage?

You make a good point - that we should watch out for manipulation everywhere on the internet - but then you undermine it by ending with an unfounded hyperbolic statement. You only provided one example user out of thousands (millions?) of HN users, how do you get from there to "drowning in garbage"? Seems to me the mods and community here are doing a pretty good job already.

PennRobotics 4 years ago | |

The mentioned user is always on the front page, and I never hesitate to flag all of their non-tech/startup links, which is probably 90 percent of their submissions. They post nearly once per hour. This user and three others are the reason I posted a similar recent rant about serial off-topic posting: https://news.ycombinator.com/item?id=30907198

I can call out usernames if you really insist, but that's a bit rude when some of them are on https://news.ycombinator.com/leaders

blunte 4 years ago | |

I intentionally did not identify the poster. But if you look at recent submissions regarding birds and phys.org, you'll find them.

As for hyperbole, I think it's a valid concern. If that user is allowed to operate freely, (and assuming that user is being paid or is a bot), then we can expect many more like them to appear. Then indeed HN will be drowned in garbage submissions.

Unless the mods have analytics which identify odd patterns of behavior, they will depend on users to flag things. And if only a few users flag something, then most other users won't see it. Thus, it only gets "some" flags.

It seems more reasonable to judge a user by the volume of posts and submissions over time. Who legitmately has time or motivation to submit 10+ things per day? Or to do that several days in a short period? Even if that is a real human choosing to promote content, does that not suggest some behavior which is likely to be detrimental to HN as a gathering place?

gtirloni 4 years ago |

Flag and move on.

blunte 4 years ago | |

That's the problem though... the post I saw wasn't flag-worthy. It was a bit poor quality in my opinion, but it wasn't objectionable.

What was obvious to me was the frequency of submissions of the account, and additionally nature of the topics.

Simplistically, I would question the need for any HN user to submit more than one or two entries per day. People who submit 10+ are probably suspect.

hillcrestenigma 4 years ago | | |

Maybe there needs to be a way to flag a user if they are negatively contributing to HN.

gabrielsroka 4 years ago |

> Moderator dang is one person

He's the public voice of a team.

blunte 4 years ago | |

Ok. Even so, I think there could be some automation to at least hold submissions for review based on patterns.

Perhaps I'm lazy, but I can't imagine submitting 10+ links per day. That seems like a bit of an outlier which could be identified programmatically.