On 12/02/06, Sam Korn smoddy@gmail.com wrote:
On 2/12/06, geni geniice@gmail.com wrote:
Umm evidence?
Touché! I guess I cannot substantiate this other than with basic logic, that there are just too many articles for a small core of users to have edited. Even if one in ten were of any quality at all (perhaps even just one in a hundred), it is still too many for a small group to have reasonably edited.
Lets see.
Geni has 8188 edits, 3329 article edits, over 3351 pages. Sam has 8814 edits, 3542 article edits, over 4708 pages. I have 8344 edits, 4647 article edits, over 4369 pages.
Lets assume, since those numbers are broadly similar, that we average out to a statistically normal "heavy user", who has...
8442 edits, of which 3839 (45%) are to articles. These are over 4143 pages, so our user has perhaps edited 1,880 seperate article pages - call it 2,000.
Q. How many of those highly active users would we need for it to be statistically likely that one had, at some point, edited each of our 970,000 articles? (5 marks. Show working)
-- - Andrew Gray andrew.gray@dunelm.org.uk