After using these splendid stats for quite a while, I've come across a peculiarity I don't quite understand. Which users are included in the BOINC combined team member list may seem almost arbitrary, but I think I've found some system to it. It doesn't yield sane stats, but at least it's seemingly consistent.
The BOINC combined team member listing for my team, sorted by last week's production, looks like this:
http://boincstats.com/stats/boinc_user_stats.php?pr=bo&st=0&or=10&teamid=e24fa5f1901c2c38540e9adef2e3b0a1
Now look at two projects for one particular user:
http://boincstats.com/stats/user_graph.php?pr=abc_main&id=1748
http://boincstats.com/stats/user_graph.php?pr=pg&id=4077
Last week's production (at the time of writing) in these two projects should have earned him the #2 spot in the team member listing above. But he's nowhere to be found. Why? I think it's because he's not a member of that team in his *first* project (sorted alphabetically), nevermind the rest.
To sum it up, I'll introduce a user named "User" and a team named "Team":
The way I think it works today:
If User is member of Team in project #1 (alphabetically), then include User's credit from all projects.
If not (regardless of other projects), then don't include user's credit from any projects.
A better way IMO:
If User is member of Team in any project, then include User's credit from all projects.
The best way IMO:
For each project in which User is member of Team, include User's credit from that project.
Do not include User's credit from other projects.
Am I on to something?
Cheers!