User talk:MusikAnimal (WMF)/NPP analysis
AfD questions[edit]
MusikAnimal (WMF), sorry to be the bothersome one who asks you for all sorts of numbers, but I think it would be interesting to see what percentage of pages deleted at AfD were initially created by new users. It would give us a way of comparing what the daily AfD load would be before and after ACTRIAL. I'm not sure if this is possible to pull, but I thought I would mention it. TonyBallioni (talk) 18:38, 28 June 2017 (UTC)
- @TonyBallioni: No problem! I think we could infer this from the existing data. This pie charts shows the number and percentage of articles created by non-autoconfirmed users from January through March 2017 (at the time of writing) that were deleted via AfD:
Graphs are unavailable due to technical issues. There is more info on Phabricator and on MediaWiki.org. - ...versus all non-autopatrolled users:
Graphs are unavailable due to technical issues. There is more info on Phabricator and on MediaWiki.org. - So you could deduce that roughly 55% (1,114 / 2,009) of articles that are deleted via AfD are created by non-autoconfirmed users. This does not account for pages deleted at AfD that were created by autopatrolled users, but from my unpublished research the deletion rates of such articles are considerably lower. Hope this helps. MusikAnimal (WMF) (talk) 21:12, 28 June 2017 (UTC)
- However it's important to note the date range pertains to the time the articles were created, not deleted. So many of these articles we processed may later be deleted via AfD. The stats for your purposes should still be fairly accurate, though, because I've found that most (upper 90th percentile) of deletions happen within 90 days of the time the article was created. MusikAnimal (WMF) (talk) 21:18, 28 June 2017 (UTC)
- Thanks. I should have thought of that. Yes, that makes a lot of sense, and is what I would have expected. Thanks again for all the hard work you have put into this. TonyBallioni (talk) 00:35, 29 June 2017 (UTC)
- Also Atsme and Godric, both of you might be interested in these numbers. TonyBallioni (talk) 15:20, 29 June 2017 (UTC)
- Wow! Definitely something WMF needs to see. The figures unequivocally support ACTRIAL. Atsme📞📧 15:55, 29 June 2017 (UTC)
- Well, considering that MA pulled them, I think they know :). Yes, I agree. It will also help with the AfD backlog and get more eyes on the articles that need it there. TonyBallioni (talk) 15:57, 29 June 2017 (UTC)
- While the recent progress have been phenomenal, only if the WMF has paid an early heed to the numbers produced by ScottyWong!But MusicAnimal has surely done an exceptional and commendable job!Winged Blades Godric 16:43, 29 June 2017 (UTC)
- Wow! Definitely something WMF needs to see. The figures unequivocally support ACTRIAL. Atsme📞📧 15:55, 29 June 2017 (UTC)
- However it's important to note the date range pertains to the time the articles were created, not deleted. So many of these articles we processed may later be deleted via AfD. The stats for your purposes should still be fairly accurate, though, because I've found that most (upper 90th percentile) of deletions happen within 90 days of the time the article was created. MusikAnimal (WMF) (talk) 21:18, 28 June 2017 (UTC)
Updating[edit]
Following up here from WT:NPR, I'm sure whatever you said here was correct, but I'm a bit dull so I need clarification. When you said 90 days to present do you mean stats based on July 17 to April 17 ? The current graphs look like they started in March because that was when the 90 day time period elapsed so we would have a good idea of the total deletion numbers by then. TonyBallioni (talk) 00:28, 18 July 2017 (UTC)
- Pinging MusikAnimal, just to make sure you saw this (no rush in response needed.) TonyBallioni (talk) 17:28, 19 July 2017 (UTC)