I've noticed the site statistics for Cognitive Sciences fluctuating a lot more from week to week than I would expect if the statistics represented the entire life of the beta (757 days and counting from the time of this post). This leads me to suspect that the statistics Area 51 displays are calculated on a period of recent activity stretching much less far back than that — maybe only a few months, weeks, or days for all I can tell. Is this the case?
Edit: it's two weeks for
questions per day, according to the cursor-hover tooltip. Is this the case for other statistics too? Only
questions per day displays this hover text. Is there a source for overall data that goes back further than this BTW? It would be nice to analyze longer-term trends without having to track this statistic manually.
I can see why you'd want to do this. Activity changes quite a lot over the life of a beta, I'm sure...You wouldn't want to count low activity in the early stages against a site that's been around for two years and developed a strong following. Hence no objections to the practice are implied by the question; I only ask for a little more transparency about the process. How else can we hope to make meaningful inferences about these statistics?
Treat this as a feature-request if you like the idea of adding a little info to the statistics themselves, such as some fine print at the bottom about the time frame that displayed statistics represent, but I would be content with just an answer here.