I've actually got this mostly up and running on the test stats site, though it's based on activity in the last 28 days. Aqua Stats for example where I list it at the top. I can easily tweak the algorithm a little.
I've actually got this mostly up and running on the test stats site, though it's based on activity in the last 28 days. Aqua Stats for example where I list it at the top. I can easily tweak the algorithm a little.
28 days seems high, but then again, I haven't crunched the stats at all to see what the distribution is. If a project had known durations for the WUs, one could use that as part of the calculation (e.g. a one hour per WU project might be set to several days whereas a two week per WU project may need a month or more for a WU to be returned). I don't know that the average turnaround time would suffice though.
Spring 2008 Race: (1st Place)