[WikiEN-l] Experiment on new pages

Mon Dec 5 15:45:45 UTC 2005

Delirium wrote:

> I suppose I'm not really sure what this will buy us.  It only takes 
> away the class of problems: "page doesn't exist, anon creats a bad 
> one".  These are actually among the easiest to spot.  Much harder are: 
> "page exists as a stub, anon adds a sentence to it that is bad".  
> These are harder to spot, and this does not address that; in fact, it 
> will encourage troublesome anons to just find existing articles to put 
> misinformation into, rather than creating new ones.  I think, on the 
> whole, that would be *more* of a problem.

To clarify (I should've put this in my initial message), I think the 
fundamental problem is un-vetted anonymous edits (or edits in general, 
but most problematic ones are anonymous), and taking away the class of 
edits that occur at page titles that don't yet exist isn't much of a 
gain, because there are over 100,000 obscure stubs on the English 
Wikipedia that provide a vast array of extant page titles to choose 
from.  In this case it just happened that the page didn't yet exist, but 
it could just as easily have happened that someone created an obscure 
1-line stub, and then the anonymous user came by later to edit the 
stub---it would still be an obscure article that nobody would find, 
because the person isn't all that notable, and the edits would have a 
pretty good chance of remaining there.

So, more interesting would be to address the fundamental problem---that 
is, mark revisions that have been reviewed by [n] people.

-Mark