I await the results of this experiment with interest.
My gut feeling is that:
A) To first order this won't really work. You can't really be anonymous in a closed group. In a small closed group
you can't be anonymous at all (the degenerate case, two people, is obvious; writing style is an identifier in larger groups) and in a really large group the stakes of letting your mask slip are too high and will squelch serious conversation, while the lure of the big crowd will drive trolls and comedians.
B) To second order this could
work great. The second-order advantage of this is the useful social fiction of anonymity. Anonymity doesn't have to be cryptographically strong to be useful. Think: a masked ball. The masks aren't that highly effective - that seven-foot-tall masked woman is pretty easy to ID, mask or no - but if everyone plays along the game can go well. Plus, you can get plausible deniability, which can be very useful in social situations. ("No, it wasn't me who confessed that secret in anonymous chat. It must have been some other native speaker of Finnish from a basketball team in Toledo, Ohio.")
C) But, alas, I'm not sure use case (B) is significant enough to drive a product on its own. Nor that it will be easy to get the site's culture off on the right foot.
A great experiment; I hope it goes well despite my worst fears. We certainly need many more alternatives to the "stand naked in front of the world on the Googleable Internet" model of online conversation.
Even in smaller groups though, people may be able to consciously subvert their mannerisms and still get their point across. Not saying it will be easy, but I wouldn't rule it out.
I agree with the plausible deniability point, but i think it can theoretically apply equally to small (and by small i mean at least 10 ppl) and large groups alike. There are many ideas that are not too controversial, but just touchy enough that plausible deniability would be sufficient to overcome the anxiety barrier. This would only work for comments that a user would have no problem being suspected of saying, as long as noone could categorically prove it.
My gut feeling is that:
A) To first order this won't really work. You can't really be anonymous in a closed group. In a small closed group you can't be anonymous at all (the degenerate case, two people, is obvious; writing style is an identifier in larger groups) and in a really large group the stakes of letting your mask slip are too high and will squelch serious conversation, while the lure of the big crowd will drive trolls and comedians.
B) To second order this could work great. The second-order advantage of this is the useful social fiction of anonymity. Anonymity doesn't have to be cryptographically strong to be useful. Think: a masked ball. The masks aren't that highly effective - that seven-foot-tall masked woman is pretty easy to ID, mask or no - but if everyone plays along the game can go well. Plus, you can get plausible deniability, which can be very useful in social situations. ("No, it wasn't me who confessed that secret in anonymous chat. It must have been some other native speaker of Finnish from a basketball team in Toledo, Ohio.")
C) But, alas, I'm not sure use case (B) is significant enough to drive a product on its own. Nor that it will be easy to get the site's culture off on the right foot.
A great experiment; I hope it goes well despite my worst fears. We certainly need many more alternatives to the "stand naked in front of the world on the Googleable Internet" model of online conversation.