Trust works best in small communities. A few bad apples can ruin the whole experience as the community grows. Within no time the community degrades if it is just based on trust. A saner way to do this will be to trust people to do the right thing, but verify that they are not doing anything malicious.
A saner way to do this will be to trust people to do the right thing, but verify that they are not doing anything malicious.
Absolutely. And Mefi's definitely been of a size for a long while now where trust-but-verify has been an essential part of keeping it working. Our approach these days is pretty much the same as Matt's approach years back when it was just him running the place: welcome people in with the benefit of the doubt but keep an eye out for trouble -- whether of the malicious or the clueless variety.
And we see both: keeping spam in check (even after the $5 speedbump) is a daily project, though thankfully not a high-volume one (we probably ban a couple of spammers/linkfarmers/astroturfers/SEO-marketers a week), and we're daily working with new users who are varyingly conspicuously not getting how the place works or what it's for or what the local tone and community expectations are.
One nice thing is that we've been able to take long experience with dealing with spammy behavior in particular and use that to build a pretty decent local toolset to keep a close eye on problem areas with a minimum of effort, so what spam does actually try to manifest these days tends to get spotted and nixed with quick turnaround. Which by itself is good news but doesn't say anything directly about the community, but the implication is important: because we can get on top of one of the major sorts of bad behavior very responsively, it leaves us free to continue to let everybody else on the site have the sort of general community freedom that Matt talks about in those guidelines re: trust. Because our small administrative team can effectively and promptly corral the problems that do occur, we don't have to clamp down on the basic posting and commenting rights of the vast majority of users acting in good faith.
It's really rare that, with the exception of spammers and self-linking douchebags, we have to forcibly tell someone to leave. I always figure we're doing something right if people on both sides of whatever argument people are having think we're favoring the other side.
The really big thing we "do" such as it is, is spend a lot of time in MetaTalk, the part of the site that is dedicated to community discussion of etiquette, bugs and sitewide norms. All of our moderator decisions can, if people want, be discussed there. This can be a pretty under-the-microscope why-did-you-say-it-this-way sort of thing, but it usually involves a lot of people talking about why the guidelines are the way they are.
And if you just want to read the site, or answer a question on Ask MetaFilter, you can ignore the navel gazing part of the site entirely.
My favorite thing about MeFi is the high conversion factor. That is, lots of people meet online but then go on to become friends or associates or colleagues or coworkers in real life. I love internet land, but I think there's got to be crossover between that and the big blue room and we have a lot of that which I think, in turn, reinforces community trust and cohesion.
Metafilter does this, actually. When you sign up you pay a once off fee of $5 (the meta-filter). If you do something especially bad, they shut your account down and don't return the $5 (AFAIK).
And while we absolutely don't spite the income from the signups -- it's a nice side effect of the signup-throttling process and helps with costs some -- it's really not nearly enough to cover payroll for three full time mods and a full-time programmer/developer. The bulk of our revenue is from ads running in the archives.
Metafilter is quite an outlier in terms of online communities that have survived, thrived, and kept a high signal-to-noise ratio.
There were also problems in that the initial setup of k5 was completely user-driven and user-generated with votes on the mod queue determining what was good. Early attempts to 'control' the crapflooding issue was to create the "diary" section, which functioned as license to post anything you wanted. The $5 fee was a reversal of that approach AND didn't add anything for users (Scribd?), so people mostly left.
Another thing that probably drove the decline of K5 was the rise of blogs for everyone. Why post at K5 when some bunch of goons could remove what you regard as worth publishing while on your own blog it went live always?
That wasn't the entire community, dont get me wrong, but it was enough to slow activity to the point where it wasn't very interesting anymore.