A lot of people think that reporting social statistics is a good idea. I mean, c’mon – we all want to know how many friends we have, how many people we know, who we talk to the most often, etc. This is fascinating, right? Yet, twist the conversation in a different direction… do we want those around us to know who we talk to the most? Probably not.
Cobot in LambdaMOO: A Social Statistics Agent is a great paper (by folks at AT&T) that addressed this issue head on. It’s an old paper, but one that’s really important for anyone thinking about the social implications of providing statistical data publicly. Basically, Cobot spent a bunch of time in the MOO, collecting information about people’s social behavior and then spewing it back out at them. Problem was that it also told you how important you were to others, which while utterly fascinating to most people caused significant divisions in a way that did not build community, but divided it.
Seems like people already do a pretty good job of noticing and remembering with whom their friends are talking to and spending time…
Yes, and no. Self-deception is very powerful. Do you really want to know who the girl you fancy spends most of her time talking with? If you were forced to think about it, you might know. But you still like to imagine her spending all of her time talking with her girlfriends, not your best friend. When she wants something to do on a Saturday night, do you want to know that you were the 7th person she called? Probably not…