[gwc]

Is it sufficient to simply not target growth as a goal?

HN, if I'm not mistaken, has kept a general upward trend in number of users; presumably, this is accompanied by an ever-increasing variance in quality of those users and what they consider to be interesting news and good etiquette.

Don't get me wrong; I certainly agree that it is easier to make changes with an eye towards quality when you're not concerned about growth. However, I do wonder if there it a point where the dark side of network effects, which are primarily sociological in nature, will simply overwhelm any technological aspects or other moves on the part of a community's stewards / leaders.

[Xichekolas]

> ever-increasing variance in quality of those users

I'm not sure it's necessarily the quality of the users. Call me an optimist, but I think most people in general are decent and most people that bother to come here are pretty smart.

I think the problem with size is the same reason everyone else is a crappy driver but you. In a normal day of driving, you are surrounded by mostly good drivers, but a couple of them are bound to make really stupid mistakes (just by random chance). Since this is your only experience with them, you label them a "bad driver" and it overshadows all the good driving around you. If you only drive on lightly traveled roads, you are less likely to see a stupid mistake, but get on a major interstate highway in a major city and there are enough people that something is bound to happen.

Likewise with large social news sites. Every user is a 'good user' 99% of the time, but has that random moment when they do something trollish or get carried away with an argument and say something mean. (I know it has happened to me before.) When a site gets large, the probability of this happening on any given thread rises accordingly, and since everyone focuses on these instances, it seems like every thread is full of trolls and angry people.

I think anonymity also plays into it. Large communities are by default anonymous, and small ones are not. People in small towns don't cut each other off or tailgate each other, because you are likely to know the person in the other car, or at the very least be headed to the same place. (I know, I'm from a small town.) In a major city, you are never going to see that other car again, so no one cares if they act like an asshole.

[william-newman]

It is not sufficient simply not to target growth. But not targeting growth removes what is otherwise one helluva tough design constraint for a filter system. Imagine writing a spam filter when you're being paid by beancounters whose bottom line is maximizing the number of messages. Even, sometimes, beancounters who face strong temptation to maximize the number of bytes of email this quarter and worry about the longer-than-a-month-term consequences later.

Also, some kinds of filtering (and related things, like informal community norms) don't scale, so if the community hasn't grown too much, the problem might tend to be easier. Hacker-oriented mailing lists with 100 posters don't seem to be too hard to moderate, but I'm not eager to try 10,000 posters.

[stcredzero]

"Is it sufficient to simply not target growth as a goal?"

Growth is a big part of the problem. It's much the same for movies. You can either appeal to a more discriminating audience, which is inherently smaller, or go lowest common denominator. Actually, the pattern is to start out appealing to the discriminating crowd to get a "cool" cachet, then expand and go mass-market.

Another big problem is the monkey-see-monkey-do nature of Homo Sapiens. Over time, you will see more noise in the form of attempts at imitating signal instead of true signal.

[kingkongrevenge]

Metafilter simply turned off new user registration for a long time when they didn't like where things were going.