Topic on Talk:Page Curation

Currently we have a very simple colour scheme - yellow for unpatrolled and white for patrolled. I'd like to make this more complex in various ways. but the simplest would be a warning red colour for possible badfaith articles. This could be done with a few pretty simple rules:

1. any article created by an account whose previous 5 article creations include one deleted as either G3 or G10
2. any article containing certain obvious trigger words such as the ones used by friends of gays.

Of course many of these articles will turn out to be innocuous and uncontentious, or articles about the sort of rock groups who use strong language in the album titles and even band names. But in those cases marking them as patrolled or tagging them as A7s would immediately turn them from red to white without the author ever knowing that their article had been flashing red on someone's screen. If the interface defaulted to displaying red articles first regardless of their place in the queue and whether the patroller was otherwise working at the front or the back then we might actually speed up our deletion of G3 and G10 articles - and I would hope we could all count that as a positive.

I'm gonna have to say that I don't understand what "friends of gays" means here in this context. Could you elaborate?

Sorry I thought it was one of our better known memes. The essay got moved to meta. meta:Friends of gays should not be allowed to edit articles

Oh, okay. I was confused there for a moment...

Anyways, to the meat of your comment.

I agree wholeheartedly. I have been advocating for us to basically "embed Cluebot" into our core software - not necessarily to do reversions, but to tag and mark revisions as "suspect". This would be something that would show up in the Zoom's "List View" most obviously, but with more detailed reports in the Zoom view.

Your first suggestion - checking the revision vis-a-vis the editor's previous contributions - is probably not doable. The system doesn't track why an article was deleted, merely that it is no longer there. Further, even discovering that an editor has deleted pages is probably a difficult call to make.

I appreciate that only admins and researchers have access to deleted revisions, so the initial concept of one of their last few creations has been deleted G3 or G10 might not be easy to code. As it would compromise the idea of deleted edits only being viewable by admins I suppose in theory it should require assent from legal and consensus from the community. But I can't imagine that either would object.

An alternative way to do this would be to check the editor's talkpage, and ideally talkpage history, looking for recent G3 or G10 warnings. That wouldn't pick up everything, some of the more childish stuff just gets summarily deleted without comment, and some G10s get initially tagged as A7 or BLPprod, I doubt if many admins change the warning message even when we delete under a different code than the tag.

Well I do, but as the NPPers refuse to be educated I`'m beginning to wonder if it's worth bothering.

I'm inconsistent as regards giving feedback to patrollers. But my experience, such as it is, is not quite as negative as yours. Some patrollers seem eager to get it right and happy to learn. Others just see me as that "hemp clad patchouli smoking sandal wearing rabid inclusionist" from the Article Rescue Squadron and get annoyed when I explain that "would probably be deleted at AFD" is not a speedy deletion criterion. WereSpielChequers 16:22, 23 September 2011 (UTC)