I managed to take down arbitrary posts on Hacker News

Nasty stuff, thanks for sharing it in the open.

Were they able to run each experiment with the same botnet? Or is a fresh one required every time?

> The IP has already registered not long ago

> The IP comes from a known VPN provider

> The IP comes from previously flagged locations

I noticed this when registering. Residential IPs are treated differently. You don't get shadow-banned at all with plain old vanilla IPs. In-fact you can make your first post under a residential IP, and it will show up in /newest and you don't have to wait in purgatory for about 5 days for it to show up.

One other option is to use a really obscure VPN service that not many people use and has a greater chance of not being on a blacklist, but that is rare. It's also rare to get a SOCKS5 proxy that isn't flagged as suspicious in some spam database.

Then there's the shady practice of using hacked residential routers to make your posts, but who does that?

> lowering the author’s credibility in the eyes of the algorithm.

This is one of the known flaws of Hackernews

It has a way of earning greater weight to users but if you fall out of that there is no way back in. It is easier to fall out when the account is new, a short string of negative karma or flags and that account is burned. Browser session as well. IP not as much but the mods do watch when an IP address eventually gets linked to a malbehaved user to apply greater scrutiny.

From their perspective this is fine, because people aren't that persistent. They don't have a way to quantify what an improved system looks like that allows for re-earning good standing, so to them there’s no rationale to develop that, as long as people just move on or stay shadowbanned talking to themselves forever.

> I was surprised to learn that they actually had a decent algorithm for detecting possibly fraudolent and/or irrelevant content, as well as a weighted voting system, where the vote of a couple users with lots of karma are more important than hundreds of votes from low karma users.

> votes of suspicious users are still displayed normally, even though they do not count as much as a normal user’s vote, if not at all.

That's interesting.

There are a few notable groups of bad actors on HN but obviously the really good ones you wouldn't know about, which is a bit of a teapot.

The lack of a downvote button on posts does stop a lot of attacks.

Interesting write-up. I don't think the final exploit leads the way to a useable hack, but the journey says a lot.