[frik]

Sadly it's true. Some companies game the HN ranking algo, and employ shills to divert the discussion of their product. At least two companies (that are infamous for such shady tactics anyway) do it actively regularly on HN.

The first thing to prevent it, would be to change the HN ranking algo so that stories with more comments than votes are not automatically punished and forced many pages backwards from the frontpage. Add a report button to report users. Implement an admin interface to monitor certain users and ban them.

[dang]

You've made these accusations many times but never supply a drop of evidence. We've repeatedly responded [1,2,3] and asked you to stop, yet still you persist. I appreciate your concern for the quality of HN, but at some point this becomes abuse in its own right.

It's unfortunately common for users to feel absolutely certain that other commenters are astroturfing when they merely happen to disagree about company X or issue Y. The underlying assumption is: 'no one I disagree with could possibly be commenting in good faith—they must be disingenuous'. This is a cognitive bias. Nearly always, when we investigate these accusations we find nothing—nothing except that the accuser really dislikes $BIGCO.

Real astroturfing and shilling do exist, I've personally poured countless hours into combating them on HN, and I can tell you from long experience with the data that they don't look anything like what you're positing.

1. https://news.ycombinator.com/item?id=11844253

2. https://news.ycombinator.com/item?id=11988639

3. https://news.ycombinator.com/item?id=12322393

[truetraveller]

First, I want to thank the guys behind HN and Dang for being very open about any HN vested interests. I have no doubt HN tries hard to be honest. No complaints here.

That being said, I am very suprised you mentioned "drop of evidence". It is very upsetting you said that. There is hard evidence.

The fact is: There are may dishonest people in this world. They may justify their honesty, but that's not the point. See the PLETHORA of dishonest reviews (with proof) on Amazon, Yelp, Ebay, etc. And HN is the perfect place to game comments/upvotes/topics, and you bet it is. Firstly, it's simple to game HN. Actually, too easy. And the returns are really good. The audience are high-income earners, and intelligent people. The topics can be very niche (which means highly targeted). In fact the mentality that "HN can't be gamed" is better for the gamers, since other HN users will really believe most comments are genuine.

HERE'S THE DEAL:

I will do a project "for free" for the public good. This will hopefully establish proof. If I am allowed by HN legally, I will show my study of how I can easily, massively game HN. And you will not but able to detect it. I will not make money of this, but I will provide detailed statistics about traffic and CTA clicks. I will provide a dollar value of the fruits of my gaming, if it was to be real. I would be just one person doing this in my spare time. When the fruits are so sweet, PR companies will employ full-time paid individuals and maybe even teams.

P.s. Dang, I know you try hard to remove "shill" comments /upvotes /topics. Yes, you catch the amateurs. But the whole point of shilling is to blend in be undetectable. You don't and simply cannot catch those. And on a seperate, there was a study that "doctors who believe they cannot be gamed/bribed, are actually the most gamed/bribed".

[dang]

I said there was no evidence in a specific user's claims, not that there was no evidence anywhere. See my third paragraph.

There are two problems. One is that astroturfing and shilling exist. The other is that some users are too eager to see an astroturfer under every bed and a shill in every pot.

Both problems are destructive and we need to deal with both and not pretend that one subsumes the other. On HN the approach is simple: (1) if you think you see abuse, please let us know at [e-mail] so we can investigate; and (2) don't accuse other users of astroturfing and shilling unless you have evidence, and keep reminding yourself that an opposing view (e.g. them liking/hating $BIGCO while you hate/like it) is not evidence.

[truetraveller]

Yes, accusing a specific company certainly requires proof. So I apologize if I was defending that.

You say "astroturfing and shilling exist". It is likely a "very high" percentage. "Very high" does not need to be 60%, but it is relative. Even 10%-20% is very high, which I think the ratio might actually be. It is just too attractive--the bang-for-buck ration is just too sweet.

Please note, I don't say this with pleasure. I'm an honest shop and it is painful seeing competitors astroturfing.

[frik]

I sent details to HN per mail and other means. One of the companies is MSFT (others here named the other companies), they created manys accounts around BUILD confernce 2015. Before that event HN was a pro-Apple, pro-open source with little MSFT news - something you would expect from startup and SV investors. If you would have a proper interface, you could watch such activity in real time and act. Some third party interfaces like http://hckrnews.com/ show also stories that got hidden despite high commitment from the community. Almost every day I find insightful stories that vanished off radar. And it visible that not so nice stories about certain companies get immediate reaction by publishing a minor news story that very rapidly gets traction on HN frontpage and the not so nice stories is spammed with comments and flags to push it off the frontpage.

Disclaimer: I have nothing against a certain company per se, I just don't like how the do their PR in a very shaddy way, and how they destroy things (like their own products for very short term greed). And I would prefer if HN algo doesn't punish stories if there are more comments than votes - as this is the most common way for them to "hide" stories.

Edit: Now it makes sense, it explains why MSFT is so interested in HN (woos Y Combinator startups into their eco-system, present a polished new MSFT, PR is trying to hide ugly truth), it fits my observation of a massive user increase (green accounts) around BUILD 2015 event, etc "Today, Scott Guthrie and I joined Sam Altman, to announce a partnership with Y Combinator, one of the world’s leading startup accelerators.": https://blogs.msdn.microsoft.com/stevengu/2015/02/09/y-combi... and "Microsoft woos Y Combinator startups with $500K in Azure cloud credits" http://venturebeat.com/2015/02/09/microsoft-woos-y-combinato... , "Microsoft offers $500k in Azure credit to woo Y Combinator startups" http://www.geekwire.com/2015/microsoft-offers-500k-azure-cre... "$500k of Azure credit for YC startups" http://blog.ycombinator.com/500k-of-azure-credit-for-yc-star... "Microsoft Wants To Buy Love In Silicon Valley" https://techcrunch.com/2015/02/10/microsoft-wants-to-buy-lov... ... as Alex Wilhelm of TechCrunch wrote "It will be interesting to see what percentage of the current Y Combinator class chooses Azure over AWS". What's then answer? It certainly changed HN, that's my impression.

[dang]

This is what I mean about there not being a drop of evidence in what you say. Alleging that Microsoft created accounts on HN is a serious charge. How do you know this? You don't. You know none of these things. You're merely narrating your own perspective and finding an assortment of "details", as you call them, to fit it (some silly corporate partnership, a few stories you saw do this or that on HN's front page—Now it all makes sense!). This is not evidence, this is how imagination works.

It's really time now for you to stop posting like this. It's past tedious, and destructive of the community.

[nkurz]

It's really time now for you to stop posting like this. It's past tedious, and destructive of the community.

I'm not sure if you read it, but the recent "Who Buried Paul" is a wonderful exposition of how a plethora of weak evidence can lead to a false conclusion: https://news.ycombinator.com/item?id=13902012. What's wonderful about it is that it's written about a politically noncontroversial topic, and thus (hopefully) allows for more reasonable conversation. Might be a great candidate for recycling.

More controversially (and thus less appropriate for direct HN discussion) Scott Adams has a nice piece on confirmation bias, using Pizzagate as the example:

So let me tell you what a mountain of evidence is worth.

Mountain of Evidence Value = zero.

In the normal two-dimensional world in which we imagine we live, a mountain of evidence usually means something is true. So why am I looking at the same mountain of evidence as the believers in pizzagate and coming to an opposite conclusion?

The difference is that I understand what confirmation bias is and how powerful it can be. If you don’t have the same level of appreciation for the power of confirmation bias, a mountain of evidence looks like proof.

Here’s what I know that most of you do not: Confirmation bias looks EXACTLY LIKE a mountain of real evidence. And let me be super-clear here. When I say it looks exactly the same, I am not exaggerating. I mean there is no way to tell the difference.

http://blog.dilbert.com/post/153821538056/about-pizzagate.

[truetraveller]

Dang is right, accusations against a specific person/entity are serious and require solid proof. Could even be illegal.

[wickawic]

I could be wrong but I felt this happened on the Spotify thread that is currently on the front page. All the inital reaction was negative, about how the piece was a 'submarine,' etc. Now the top 5 comments are about Spotify's new killer feature and how happy the customers are...

[dang]

I took a look at the thread I think you're referring to, and both those comments and their upvotes are by established users with no signs of fakery that either we or our software could pick up on.

What you observed is probably a different phenomenon: early comments in a thread often start off negative, and only later do other users show up to post positively—their motivation being to balance things out. Often the positive comments to get the most upvotes in the end, but the whole process takes a while to play out. If you've ever run across a thread where the top comment says, "I can't believe how negative this thread is!" and proceeds to make a positive case with great fervor, this dynamic is probably why.

[dpweb]

I dont think you could stop upvote fakery. Maybe like highway speeding just try to control it enough to keep things somewhat safe.

Old, established accts on all the sites can be, and are, sold to people for purposes of fake upvoting. Maybe limit the number of upvotes one can do in a day. Doesn't seem a major inconvienence and makes the upvotes more valuable.

[whitepoplar]

Please name these companies!

[randomname2]

This probably refers to Uber and Peter Thiel and companies associated to him?

If they're manipulating HN they're not very good at it. Lately I'm seeing smear pieces against them appear on the front page on almost a daily basis and the tone in these threads is overwhelmingly toxic rather than positive.

[cowpig]

this--plus evidence too if you can find it please!

[DanBC]

> Add a report button to report users.

You can flag individual comments. And you can also email the mods with your suspicions.