For the other 20% it works poorly. These are cases where we end up discussion what Wikipedia is, or how the article space should be organized. Most of these conversations end up deadlocked, and the outcome is that the articles are kept. The many "List of," "Slogan,"
and "biography" discussions are examples of this. The quality of the
decision making was low, and the number of Wikipedians made upset in
the
process was high.
Yes. But I think this shows that the biggest problem is not with the deletion policy but with the deletion criteria policy (i.e. What Wikipedia is Not).
Most of the contentious items on VfD are not *really* arguments about the particular article, but about the general criteria for including or deleting certain types of articles. For example, the debate over whether a certain elementary school should be deleted is not a debate over how famous that school is - everyone acknowledges that it's not any more famous than any other school. But some people feel that only famous/important/notable subjects should be included in Wikipedia, so they vote for deletion; others think that any true fact should be included, so they vote to keep it. So the vote gets contentious. Then the next day (or next hour) another item gets listed, and the same debate happens again, without ever getting resolved.
Now imagine that we came to an "official" decision that only "notable" subjects belonged on Wikipedia.* Now the only debate would be whether this particular elementary school was notable or not, which would probably not be much of a debate.
I think we need a better process for turning the recurring debates over individual articles into policy decisions. We have a standard system for deciding on a particular article (that's what we're debating now), but to my knowledge there's no standard system for deciding on policies for entire classes of articles, which I think would have a much bigger positive impact.
--Alex (axlrosen)
* Although what would really be needed would be a general idea of *how* notable the subject should be. Of course this would be impossible to define precisely, but there are a number of ways I can think of that we could pin it down fairly well. Certainly much better than now, where there's not even a policy that says whether or not the subject needs to be notable at all. Also, we'd need to settle the whole "list" question as soon as possible, and more...