Joho the Blog
An Entry from the Archives

Why do good community/groups begin to fall apart when the number of particpants becomes greater than 150?

Better yet, how can we try to organize web communities/groups optimized to 150 so they don't self destruct?

Link to a better explanation.

The really exciting thing about "communities" online and the technologies that facilitate them is the speed and flexibility with which they can be created and destroyed. On the web, communities can spring up in a few days, for a specific purpose, and then die away casually. Unlike in real life, online we can form and unform ad-hoc specific purpose groups: to protest for or against the war; to organize a burning man camp; to solve a coding problem. In real life relationships are sticky, complex, require continuous maintenance, and are meaningful (if they are real relationships) well beyond the hot issue of the moment.

David,

LOTS of angles to this. I'm on deadline, so not much time to contribute right now alas (but I'll check back later).

2 quick things I'd reference in your position, plus one source suggestion:

1. Eyebees (which I'm sure you probably will anyway).

2. (self-serving plug) - 'AND' logic and the way 'and' acts as a transitive verb (Reed's law stuff).

Also - suggest firing an email to Stowe Boyd re: all his IM stuff...

/m

Some thoughts:

Is group-awareness necessary for a group? The web gives precedence to groups over groupings in the sense that sites that lots of people go to have more influence, but the population of espn.com visitors a group? I don't think of myself as being part of group in that sense, only as an individual user of their services.

But epsn.com visitors can't really decide to do anything as a group. They may all decide to go to page x on the site, but that's an accumulation of individual decisions, not a group decision.

My current, very dodgy thinking about what the internet "is" is that it's a series of agreements. I agree that espn.com has something interesting or useful, they agree to provide it.

Since it's easy to think of a group as a collection of people who've agreed about something (even if it's to disagree about something), what the internet does for groups is to facilitate their creation, propagation and disintegration. (But then groups are a side effect of the agreement nature of the internet, not it's main value.)

It has the nice ability to let individuals assign themselves to groups without having to participate (e.g., newsgroup lurkers and Joho subscribers who never e-mail back) or to remove themselves at will.

I don't think that groups per se make the internet more valuable. For me, the exponential value of a network comes from the increased numbers of individual providers and users. Again, I don't think that all visitors to a given site make a group.

And there are plenty of opportunities for friends or geographies to form groups. Regional newsgroups, wikis, blogs, mailing lists, Yahoo groups, and so on. The challenge is finding the people who are going to agree to be a group.

Not much on the future of groups, huh?

Pre-internet, I met with groups in church basements, libraries, function rooms, the Y. We had to have officers, agenda, bylaws, 501(c)3 status, membership, presenters, schedulers, advertizing people. To raise dues to pay the janitor to stay late we needed tax-deductible status and so forth. Now, if you have an issue, you search the web for a group and find one or, failing that, found one. Anyone with a web connection can set up forums, messageboards, wikis, shared space. Advertise by posting notices on newsgroups, just as we used to thumbtack filing cards to grocery store bulletin boards. People can find your group through web search engines, too. Groups have neither geographical nor temporal limitations. Each can contribute in proportion to their abilties and take in proportion to their needs.

I've been involved in online communities since mainframe chats in the mid-70s, and the same arguments come up over and over. Off-Topic vs. On-Topic, what's appropriate to post, who is or is not allowed in the group, whether the group is opened or closed. Some groups become self-policing, others have moderation enforced upon, others break down into chaos.

New tools have made gathering a community and contributing to it easier, yet problems remain. Slashdotting a major site (or a free book) is still a problem of mob control. Random posters, flames and spam. Daypop and Technorati and others let us see where the mob's congregating, but do they add to the discussion? How do we find Quality in the mass of stuff out there? Amazon's book rankings and Google's search rankings have found magic forumulae. Is there something similar for groups?

I can't see the web of people who whom I've sent out email. We can't even do anything with the rich social web created by second degree buddy lists. What would the Internet look like if we looked at it from the group point of view?

You might want to check this out.

What would the internet look like if we looked at it from the group point of view : A very small brain with a very buddhist mind.

I'll go now.

Peter

semantic games for staticians.
try comparing to communalism as a social dynamic. or just for the lark Occum's inverse\
good luck

We go online and form groups. We mess up. We blame it on the medium. But how much do we value our skills with groups? How dearly do we hold facilitation (by a facilitator or group self-facilitation). How much do we practice clear communication, and more, strong listening.

Online groups offer us a treasure of connection across lines that we cannot often cross in our physical bodies. But the environment still requires very human skills of group communications.

In all the flurry of talk about social software, who is talking about social systems and processes?

More at http://www.fullcirc.com/community/communitymanual.htm

And if you are interested, I'd love to tell you about the group of Armenian's I just trained to run online events in Armenia. But now I gotta turn off this machine. My hands are hurting. Have a great panel. I was thrilled to see you were talking about groups.

The 150 magic number: is based on the cognitive limits of social interaction, the "theory of mind" that says that people can only keep a certain level of social relatedness in their minds. See Steven Pinker "How the Mind Works."

Re: "why doesn't the web feel like groups" -- I don't really agree. Groups are inherently webby, messy, and morphing. There are only five or so companies in the world that are over 400 years onld, and the US is the world's oldest unrevolutionized nation. Small groups grow, break apart, and reform.

The real message is that the tools we use to interact on the web don't do a very good job of representing groupness to us. The comment David made about the second order buddy list is a good example -- there are no generally accepted approaches to represent social networks well.

Phenomena like Ryze have a piss poor user interaction, although growing a cult-like status. The IBM Community Tools project or Microsoft's threedegrees may represent indications of where real-time communities may alleviate the inherent dumbness of 'slowtime' -- that is to say asynchronous -- experience of group membership. The future of groups on the web is all about shifting to real time.

Take a look at http://www.aworkingmodel.com/images/0303-2_Accelerating_Collaboration_030401.pdf for more on this.

David -- If you feel like you have not enough to say for the conference, I'd be glad to go (wink).

The 150 magic number: is based on the cognitive limits of social interaction, the "theory of mind" that says that people can only keep a certain level of social relatedness in their minds. See Steven Pinker "How the Mind Works."

Re: "why doesn't the web feel like groups" -- I don't really agree. Groups are inherently webby, messy, and morphing. There are only five or so companies in the world that are over 400 years onld, and the US is the world's oldest unrevolutionized nation. Small groups grow, break apart, and reform.

The real message is that the tools we use to interact on the web don't do a very good job of representing groupness to us. The comment David made about the second order buddy list is a good example -- there are no generally accepted approaches to represent social networks well.

Phenomena like Ryze have a piss poor user interaction, although growing a cult-like status. The IBM Community Tools project or Microsoft's threedegrees may represent indications of where real-time communities may alleviate the inherent dumbness of 'slowtime' -- that is to say asynchronous -- experience of group membership. The future of groups on the web is all about shifting to real time.

Take a look at http://www.aworkingmodel.com/images/0303-2_Accelerating_Collaboration_030401.pdf for more on this.

David -- If you feel like you have not enough to say for the conference, I'd be glad to go (wink).