Hacker News Reader

Comment by ajiang | Hacker News Reader

Isn't this a problem that can be easily fixed in the same way that HN looks for vote rigging? Seems like one week of work in putting in a fraud detection / collusion / vote weighting system. Also seems to be just a minor growing pain, a feature just ignored so far because it hasn't been important to work on, but once it becomes a bigger concern by the community, I can't think of a good reason why it wouldn't be resolved.

Pandering is hard to stop, but again as the community matures, you'll likely see counter culture develop in the community in the same way HN commenters complain about the first comment to any article being a negative response about the submitted content ;)

View on HN

> Isn't this a problem that can be easily fixed in the same way that HN looks for vote rigging? Seems like one week of work in putting in a fraud detection / collusion / vote weighting system. Also seems to be just a minor growing pain, a feature just ignored so far because it hasn't been important to work on, but once it becomes a bigger concern by the community, I can't think of a good reason why it wouldn't be resolved.

It would be relatively simple to stop: Hacker News implements a behavior where upvotes on direct links to a HN submission from a non-HN referrer do not count. Of course, there's a workaround (not disclosed here) to this behavior : if PH implemented a similar system, everyone would just use the workaround, as there's a strong financial incentive for the average person to upvote-beg (Hacker News has flagging which strongly offsets any upvotes for inappropriate articles using this method, while PH downvotes are equal power to upvotes). Additionally, there's a strong disincentive for PH to stop upvote begging (each upvote beg tweet is a free name drop for Product Hunt), which is why I'm not fond of the "bullet point" instead of "tell the user that upvote begging is bad"

The problem is they've made it pretty clear they won't bother as its been mentioned multiple times and the general response is "Oh, we'll add a FAQ bullet point" or the like and then it gets blown off.

"May I incentivize people for upvoting my product?

Please don't. People should upvote things they genuinely like or find interesting, not because they're bribed with free (metaphorical) candy. :)"

http://www.producthunt.com/about

That is as close as it gets.

We have anti-vote-ring and anti-vote-spam in place. Similar to HN actually. They and also people of the reddit team help us a lot.

Additionally we have a moderator team that additionally looks for in-balanced behaviour.

I understand that a lot of people are asking on twitter for upvotes - so far we are able to filter those in the rankings. We are considering to communicate the policies even more public - but step by step.

Disclaimer: CTO @ PH.

When was it implemented?

Maybe it was recent enough none of us noticed the impact...

@opendais might be.

Sorry i answer so loose/un-specific.

One of the advices of HN/reddit was to not communicate those things too openly in public. It's a bit against my nature but it makes sense from what i can tell so far.

It is fine. :) We all have our reasons for making the choices we do.

Hopefully going forward it'll put an end to the impression that PH is easily and frequently gamed then. :)

Interesting. This needs to be stated publicly and transparently.

Your assertion that "HN/Reddit are secretive" doesn't work for PH since PH puts a larger emphasis on real-life identity and impact. (HN has recently been more public with how it operates too.). Case in point, my parent post has 12 points. This is the perception of Product Hunt.

I didnt disagree with you. And also you are right about the fact that we need to better communicate our policy regarding upvote requests.

So far we have had the upper hand with ranking adjustment. But the fact that you believe it's mainly gamed - shows that we have to do a lot of work in improving our communication.

Reply on news.ycombinator.com