David wrote: I was only talking about the general principle, but yeah Neil's idea is probably better, David, this is an extract from my user page (it isn't particularly interesting at the moment, I keep meaning to do more samples, but it's better than nothing); I selected 4 random pages of 50 newpages, then deleted what was obviously speedy deletable, here are the statistics; a.. 200 new articles total. b.. 89 were from anonymous IP users. c.. 41 anonymous IP articles were deleted. d.. 0 registered articles were deleted. 21% of all new articles were easily deletable (all from anonymous users), 46% of all IP user articles were deleted. Note 1: More of the anonymous IP user's articles could have been deleted, but not strictly under the speedy deletion criteria, and they will eventually be deleted through AFD or edited to remove all the original content anyway. Some registered users articles will probably suffer the same fate, although none met the speedy delete criteria. Note 2: 200 articles isn't a very big sample, but it is probably enough to be representative. Note 3: There could have also been other admins deleting articles, so it is possible that an even greater proportion of them were speedy deletable.

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 David Gerard wrote: I can't give you hard numbers, but speaking from personal experience: When RC patrolling, I never look at new articles created by registered users, but I always look at every new article created by an anon. Looking at registered user created articles is almost always a waste of time; they're almost 100% good. Anon-created articles could be as high as 30% speedies/copyvios, and another 30% get some form of {{cleanup}} tag from me. Ryan