An idle thought
I've noticed in the last few days that the frequency of @replies on twitter amongst my small circle of friends has increased dramatically.
@replies were always interesting as they grew out of the natural behaviour of the community, and were only later included as a full feature of the system.
I'm wondering if the high crossover between friends and use of twitter as SXSW as an organising tool might be due to this. It would be an interesting stat to track - changes in the 'flocculence' of the site as replies cluster around events with large groups of active users. It might be an interesting visualisation exercise, particularly if it were possible to map cel location to incidences of flocculence.
Were I able to write scripts to extract the data, and map it against dopplr co-incidences, I would.
But I can't, so someone else is welcome to the idea.