Hi,
Extrapolating the trendline from the first 3 weekends would have come to a very similar result.
I am afraid there is no actual statistical proof that claims that the increase was caused by FOSDEM.
I would actually expect a slight decrease to be honest (all developers watching talks instead of actually coding). and maybe a delayed effect in a few days / weeks that would show an increase.
Intresting would also be to look only at the repos that were (slightly) related to people /rpojects that were actually on FOSDEM (impossible to do I guess)
Sorry to spoil the enthousiasm here :)
Nathan