Last modified: 2012-07-09 19:49:11 UTC
Adding a meta-distinction to the db between patrols undertaken with PT and patrols undertaken in Special:NewPages would give us a quantitative metric for adoption and the lack thereof, with incredibly fine granularity. Just an additional field in the db of 0 or 1 (pagetriage=0 or pagetriage=1) and we've got a metric for measuring socialisation.
Well we do have a pagetriage_log table which keeps detailed stats on all use of pagetriage, but it doesn't have any data on use of Special:NewPages. There are a few possible solutions to this: 1. Add a new field to the recentchanges table indicating where the patrol action came from. Downside: The recentchanges table is pretty big and slow; adding another field to it will make this worse. 2. Log all patrol actions to pagetriage_log regardless of whether they came from pagetraige or not. Downside: This makes parsing out the pagetriage data more complicated, and the table bigger and slower. 3. Create a new table just for logging patrols done from Special:NewPages. Downside: Not much, although some might complain that creating a new table for this is overkill.
Marking this as fixed. Records in pagetriage queue has one of the following status: unreviewed, reviewed ( patrol from PT ), patrolled ( patrol from Special:NewPages ) and auto-patrolled, we can use these data to make socialization analysis