Beautiful experiment: Jeffrey Nolan has created a fake LinkedIn account, then sent out link requests, to see how many people would reflexively accept a link from someone they could not, by definition, actually know: Note that this is not the same as a Fakester -- Liz Goodman draws a distinction between a fake and a Real Fake. I'm a fake if I say I'm an outdoorsy type who loves the ocean; I'm a Real Fake if I say I'm an Elfin mage. The Fakesters were Real Fakes of the highest order -- Jesus, the City of San Francisco, Pure Evil. What Nolan is doing is much more subversive: taking what everyone has noticed -- that no one turns down friend requests -- and turning it from an observation to an attack strategy. A site with a lot of Fakesters could be fun; a site with a lot of fakes would be significantly less useful than advertised, if the system started forwarding communications request through fictitious nodes. The loss of value would come not merely because such requests might not arrive, but because if they _did_ arrive they would demonstrate to social hackers that they could get to anyone listed on the service. All you would need to do is send link invites to hundreds of people they don't actually know, and take the high yield of links generated, to see more of the network at no cost. (Something like this seems to be happening on Orkut, in fact, albeit as a result of the simplicity of link requests, without needing the person's real-world email address.)