The chat room/forum problem | Hacker News Reader

The chat room/forum problem

scobleizer.com

51 pointsbdfh4225 comments

[cue rambling]

So you want to see what interesting people do. How do you do that? You go to the places interesting people go to and stand around them, looking over their shoulders to see what they're up to. All is nice and well, except others are also interested in what the interesting people do, so more and more people are starting to look over their shoulders, and the place gets too crowded. Now the interesting people tire of all the people staring at them all the time, so they find a new place to hang out - who wants to be where everyone else is, after all?

They find a nice little place where they're just about alone, but soon a few of their hangarounds start finding their way to the new place, and it all repeats.

After a few rounds of this, the interesting people start going to private clubs with strict guest list policies, and the problem of hangarounds disappears. Now they can be interesting without distractions, but nobody will actually know any of the interesting stuff they're doing, since it's all in private, and they cease to be interesting.

So... What what I trying to say? I guess it's a variation of the curse of being a public person: people take interest in what you're doing and want to interact with you, but they do it all the time and not just when it suits you. In the end you either go Britney Spears or you go Howard Hughes.

You're not addressing the other half of Scoble's point, which is not just lamenting that forums inevitably break down, but that blogs and Twitter and other such explicit lists offer a way out. I've been making this point for about eight years now, and while I (obviously) will still participate in forums such as this, I do it with the certain knowledge that it is only a matter of time before it falls apart and there is nothing I can do to prevent that, only delay it sometimes.

Meanwhile, my Google Reader remains as high quality now as it was when I first started using RSS nearly ten years ago. Not a whiff of degradation, because when a blog goes south, that blog gets evicted. If anything it's better.

It's a different type of community, and it isn't perfect in every way, but it scales radically better. A forum falls apart sometimes in the mere tens of active users, certainly by the thousands, but an RSS feed is just sitting there, serenely serving out.

(This is part of the reason I'm so against things like SideWiki, which threaten to bring forum user economics to every part of the web. Yes, to some extent you can just "not use it", but even the overflow of crap from the forumification of everything can reduce the value of a blog to you. This is also why I don't think it is necessary for a blog to have a comment section; why forcibly attach a forum to something that is superior in its own way?)

I'll try to make a point that I had in my head while I was writing my ramble. In the forum case, the method of finding interesting persons is to go where they are. It's the same mechanism that drives the crowds to popular clubs and pubs - people want to be where it's cool to be. That will lower the percentage of cool people there in that place, and the cool people will move on to the next place. The clubs try to counter this by having guest lists and bouncers that only let you in if you're cool enough (that's not an easy thing to do online).

In the RSS case, you're not trying to find cool people by going to their favourite place, instead you go to their homes. Which is all nice and well, except you have another problem: how do you find out where the cool people live? Especially if you get all your input from the world through the RSS feeds, you're well on your way to getting a very limited view of the world (sort of like only watching Fox News, or one of the other more or less biased news outlets). Your RSS sources will act as gate keepers by only linking to what they find interesting, which can lead to a narrowing spiral.

In short: RSS is very useful, but make sure you're not only listening to a single subset of opinions. To prevent this I do what I suspect what most people here do: I keep track of my favourites using RSS and look for new stuff on HN, Reddit, etc.

While true (and that last bit is indeed exactly what I do), I would observe that viewpoint narrowing is not an RSS or forum thing. You can choose to read a wide array of viewpoints and visit diverse forums, or you can read a narrow array of RSS feeds and visit forums that basically consist of people who all agree and occasionally fend off forays from someone who disagrees. It's not an RSS or forum issue, it's a personal choice issue. Monolithic forums exist everywhere, not least of which is comment sections on some blogs.

I'm trying to develop http://www.cawtech.demon.co.uk/outer-circle/core.html around the idea of shared hierarchical white lists.

this avoids narrowness because you can set the tree depth deep enough to put lots of people on your whitelist.

Scobleizer complains about the other people inviting other people into your life. Outer Circle suffers from this problem, but you do get to control who gets to invite other people into your life.

I don't thinks it's a matter of the superior people vs the rabble. I'm sure a forum of 100 random YouTube commenters would be infinitely better than the YouTube comments section. It's a broken windows problem -- you see a forum where it's all lolcatting and meme-dropping then you either ignore them or join in -- at some point, there's no fixing a broken forum.

I'm sure Scoble is a nice enough guy and all, although I've never found a reason to read him.

After this one article, I still haven't found a reason to read him.

Dude. Dude. We've been talking about this topic over here on HN for at least 2 or 3 years. Welcome to the party. Wish you had brought some new concepts with you.

Worst yet, the wrong conclusions seem to have been drawn. It's not just "people that teach" it's "things I will enjoy experiencing" because nobody is in learning mode all of the time.

Curious responses. Most HN'ers seem to be saying "same old same old". But I thought this was a very interesting and thought-provoking article.

In any case, the question of how to provide a good (for various values of "good") social environment on the web is a difficult and important one. The more people that think about it and talk about it, the better.

(And, on a rather unrelated note, I printed the article out, so that I could read it more carefully later, thereby learning that Linux FF 3.5 -- which came with my recent upgrade to Ubuntu 9.10 -- has that wonderful print-preview interface that Win FF has had for years. That's cool, and it removes almost [but not quite] the last reason for me to keep my Win machine around. Still, I wonder why it took them so long; after all, the Win & Linux versions of FF have been more or less identical in most other ways.)

It's "same old same old" because people have been writing variations on this article for over ten years.

Its the age-old question of being able to maintain a high signal-to-noise ratio. The bottom-line is that the more "democratic" a forum/board/discussion media gets, the more clueless it becomes. There doesn't seem to be a way to get away from that.

"The bottom-line is that the more "democratic" a forum/board/discussion media gets, the more clueless it becomes."

The voice in the back of my head reading that is shouting" but what about the "Wisdom of Crowds"? I think the problem we are seeing here though is that we are starting with a very "smart crowd" and only adding people with less knowledge of the subject.

All examples in the article had self selected crowds that later got diluted. It is the common problem, not only HN.

The "Wisdom of the Crowds" is simply a specialization of the Law of Big Numbers to average out the noise and leave the signal. As such, it has two weaknesses: (1) if all you have as input is garbage, no matter how much garbage you put in, you will get garbage as output, e.g. "the Emperor's nose" (2) every action has to be independent, or the math behind it is no longer sound; due to the anchoring effect and social pressure, early noise tends to be amplified.

"The Wisdom of Crowds" - as the book explains - applies only to questions with concrete answers, and it works because groups are wrong in a predictable distribution so that they average out to right.

I, for one, believe the state of social software on the internet would be better if people actually read the book.

The issue discussed in the article is not being democratic or not, but let me choose who I want to pay attention to. If you visit a blog you know who is talking, and the same is true for twitter.

Correct ,but there is also the focus on the ensuing discussion, feedback.

Popular blogs can have a large comment population and it is quite hard to follow the ensuing relevant discussion. Twitter to some extent alleviates that but then Twitter is not really a discussion forum.

Twitter partly solved the spammer problem as well. Nobody will follow a spammer. The only thing I hate about Twitter is too much trivial RT's

But there's a problem with Twitter's model, if you join twitter too late, you don't know where to start, too much fresh ideas to understand how twitter works quickly, and you have no friends to get attention to.

>> Nobody will follow a spammer.

I'm pretty sure people do, they get tricked all the time into following spam accounts. It's pretty common.

Forums break down after they the number of users passes the Dunbar Number. It's only when you have so many users that the community loses cohesion that this happens. In many of the communities I participate in, forums simply split every time they get too large.

I agree, in general, that there is some threshold above which forums break down, but I suspect that your guess that it's Dunbar's Number can't be right.

Are you counting lurkers? Visible participants tend to be the tip of an enormous iceberg below the surface, and it's not always the same set that happen to participate in any given thread.

Also, must the threshold be the same for every participant? Dunbar's Number has to do with a cognitive limit on social relationships. There are many who pay little attention to the identity of those who post, focusing instead on what is being said. If I'm not tracking the people, why would Dunbar's Number be a limiting factor in what I can get out of a forum?

You are right that it isn't exactly the Dunbar number. In a forum with a normal participation curve (what percentages of users have what activity level) the breakdown is around 300. In forums where everyone participates actively it breaks down a lot sooner. Lurkers are basically disregarded.

On a site like Hacker News where I don't recognize many posters at all, you are right, it doesn't really matter. Except that whatever "culture" HN is aiming for might be diluted.

Moderation can change this.

What it boils down to, is that if you create a community, then there will be people there that you do not like.

This is not a new problem with software. It happens even in person. If you have public meetings in a cafe or a library, with your small group of friends, they will want to invite others. Or somebody might come over to you and start talking.

Groups are always full of people you won't 100% agree with, or who will "dillute" the thing that you think you love about the group.

And people leave groups, too. Or they change.

Twitter, on the other hand, is not really social software. It's not groups.

Instead of people going to the cafe to meet - and, god forbid, having to deal with their friends' friends habits, or strangers - it's as if everyone has their own cafe, where only they can speak, all the time. And sometimes, like ships bumping in the night, those cafe-bubbles overlap.

Which is why Twitter is great for a lot of things, but not building the kind of community that Scoble is lamenting.

Because to form that deep type of connection, you REQUIRE "off-topic" chat, and adjusting to each others' idiosyncracies and social engagement styles. It requires compromise, and annoyance, and fighting and making up.

Without those things, you can't have a true relationship.

Relationships, among two people or groups, are messy and changeable.

Their frustrations are the very reason that they are valuable.

An interesting explanation of why discussion forums remind me of fashion and why HN's time is ticking:

Both systems can be modeled as a few cool types trying to stay away from some clueless followers. They move someplace else -- and are followed by the clueless. So they have to move again. And again...

I guess I qualify at least as a clueless follower here. That is better than fashion, where I'm not even a clueless follower. :-)

Ah well, it is a stark reminder. I'll take responsibility and just read in the future, unless it is a subject I know quite well.

it isn't boolean. there is a distribution ranging from bleeding edge to last guy in the room to know. it would be cool to see the popularity of things graphed over time and compare distributions (twitter may make this possible).

Reply on news.ycombinator.com