The curse of the upvote | Hacker News Reader

The curse of the upvote

unalone.net

42 pointsunalone26 comments

This problem has already been solved in various forms.

Social features may not be worth it if you want the best content. The guy who made delicious purposely didn't include "community" features like ratings or comments. http://simon.incutio.com/notes/2006/summit/schachter.txt

Or from the recent dark knight/godfather episode on IMDB..

"...in these types of situations their cognition or cooperation failed because the members of the crowd were too conscious of the opinions of others and began to emulate each other and conform rather than think differently." - http://news.cnet.com/8301-13515_3-10000650-26.html

Even Dawkins weighed in on this back in 1997 - http://www.simonyi.ox.ac.uk/dawkins/WorldOfDawkins-archive/D...

I think what we need is an adsense for content.

That's what I like about del.icio.us. There are some things is DOESN'T do, but that's because it doesn't have to. It tackles one problem with a few social aspects, and it handles it perfectly.

AdSense for content? What do you mean by that?

It's just been kicking around in my head. Serving targetted 'ads' that link to interesting websites might work considering the numbers of people who go to digg/reddit etc. Arranging it chronologically as everyone does is horrible. I'm sure I've missed a lot on HN because I wasn't here.

But there's a branding problem & I don't see anyone spending months on something that google/stumble/delicious can duplicate with a small script.

"Voting, for now, is here to stay."

Translation: voting is temporarily permanent. Nice writing.

I always usually write things like that.

Yeah, no.

"Are you sure it probably won't be a robot I know?"

"Definitely probably not."

Yeah, well. I try.

This has been discussed a lot, but one element you left out is that the "wisdom of the crowds" doesn't work as well if you can see how people already voted (although you hint at it with a comment about a few downvotes being followed by many more). It creates a 'cascade' rather than genuinely independent assessments.

The solution I came up with was just hiding the number of votes either way. Bump up good things and lower bad things, but otherwise keep it normal. It's not a perfect solution, but at least it stops the groupthink and leaves us with the dilution problem.

thesixtyone.com, a social music site, has a clever way of dealing with this: You can "bump" a song, but it costs you "points," which you see a return on if the song becomes popular. It turns upward momentum into a currency instead of something tossed around in large amounts.

Yeah, that's another solution. Problem is, it's unwieldy for news sites, since usually you're not submitting your own content.

But that only exacerbates the problem. People will start to vote for what they think others will like instead of what they personally like.

But that's how the stock market works. :) Trying to do better than that is futile.

The stock market wouldn't work either if users of the products companies make would act like investors. Just imagine what they would buy :-)

I think think the upvote concept has merit but I agree, it's flawed by votes for popular stories just because they're popular.

In order to avoid the groupthink problem a simple solution might be to hide the number of votes an article has received until you've voted either up or down/neutral. Obviously without a visible vote count you could still tell an article was doing well by its position on the site, but without the number of votes visible, there would be less of a tendency to vote something up just because everybody else is doing it.

Good summary of the voting problem and how the masses are dumber than small groups. Ironically, I upvoted it.

It's ironic if you really are smarter than the masses. But that's what the masses think about themselves: http://bps-research-digest.blogspot.com/2008/08/why-we-think...

It's a bit more nuanced than that. One of the social sites I briefly used counted clicks more then upvotes. Predictably, being a 30ysh year old male, I usualy clicked on links promising pictures of (more or less necked) women. In less then two weeks most of my personalized content was either porn or lolcats.

The moral of the story: while I do enjoy the ocasional porn and lolcat, it's definitely not what I want from the net (also why I didn't use the site for more then 2 weeks). It's just that it was much easier for their algorithm to categorize those things then the much more abstract "interesting", or even the somewhat less abstract "programming".

edit: It's not that social sites are doomed. It's just they're damn harder to get right then just digg, and they're still in infancy.

Yeah. I have high hopes for social sites. But it's why I prefer sites trying new models (like what Facebook was like before it went overboard) rather than sites that try to be "like Digg but for X."

It's okay. I'm a hypocrite and appreciated your upvote.

problem 1: voting sites don't work, because they attract too many users

problem 2: subscription sites don't work, because they attract too few users

startup idea: a subscription voting site, with smallest possible monthly fee ($1 pcm?) - the purpose isn't to make money, but to limit users. Would probably work best for a vertical market or special interest group.

see metafilter, $5 for a new account signup

Thanks - http://www.metafilter.com/newuser.mefi

They give other reasons, but I wonder if it is also retaining the quality of the community. Also, they seem to use public bookmarkin ("favourites"), and not voting as such. Maybe that makes a difference.

The solution to this, I think, lies in niche sites--this is why over the next few years small sites for specific audiences will become more and more important. Instead of one Youtube, we'll have dozens of video sites catering to specific niches. The same thing goes for news aggregators. As people learn to distrust the wisdom of the crowds, the Long Tail will become even more important as the thousands of smaller sites become more important than the few large ones.

Does the author think we are so boring as a community that all we ever want to read are purely factual and though-provoking articles that we must sit down for a half hour and read? Don't get me wrong, I love these kinds of articles, but hackers in general are not 100% serious people who want to seriously discuss hacking 100% of the time. Cartoons can be relevant and they might actually make me laugh too.

I love a good voting discussion. It's the most broken system (along with it's cousin, recommendation engines) on the web today.

Two comments:

1) What if, instead of quantitative information, we gather qualitative information? Slashdot doesn't ask you to vote up, it asks you to categorize the material. I like this much better.

2) Wouldn't it be fun to create an elitist site constructed solely around debating/discussing hot issues? You'd have to channel the ability of the audience to respond emotionally, but I bet you could come up with something like "Pop Wars" and it'd be a hoot. You could have ranking (special titles), virtual goods (rewards, do-dads, etc), and better yet, the system would drive to the top the most emotionally engaging and reasoned discussion of the hour. Something to pull you in and make you think (and participate). Not just a yes/no hot/cold POS. More like a battle where the audience categorizes the arguments, not approves or dissaproves of the contents of them. Email me if you'd like to discuss further.

Reply on news.ycombinator.com