It's all down to technicalities, isn't it?
Vernon39, I doubt training for deletionists would help much. First, I suspect Wikipedia's deletion industry attracts a personality type that would resist such training (i.e. bureaucratic, officious, unempathetic, vengeful - even if only a small percentage are like this, it's enough to negate any training you give the other 95%), and second, by the time a new user has invested hours of editing under the encouragement of the software to keep editing, with no corrective early feedback about the acceptability of the content, the damage has been done. Our system is a honey trap for new users. Nobody should attempt to start a new article until they are at least aware that Wikipedia has a thriving deletion industry that eats newbies for lunch (making a mockery of en:WP:BITE in the process).
The Main Page shows a count of Wikipedia's articles. Why does it not also give the count of deleted articles? Why do we try to hide the truth about that, and let the parade of new victims keep marching blindly to their doom?
A possible way forward might be to implement some sort of interactive chat feature, such that when a new user tries to create their very first new article, an experienced user immediately opens an interactive chat session with the new user to find out what sort of article they want to create. Many new users will want to create new articles on topics that are only marginally notable, or lacking in reliable sources, and so on. It would be better to give the bad news early than let the new user sink hours of work into writing an article that gets speedily deleted. Early intervention is usually cheaper than late intervention.
I second Teratornis' suggestion regarding mentoring a new user in the creation of their first article.
Another thought is to be able to classify an article as, for example, "beta quality", ie it is not of sufficient quality to qualify as a mainstream Wikipedia article (eg. it doesn't have sufficient referencing) and yet is still of merit. In fact, all new articles could initially be classified as "beta quality", and then promoted to a mainstream article once reviewed? That would definitely provide the encouragement for new users to create new articles (they will at least exist as "beta quality" ones) whilst at the same time providing a strong incentive to produce quality work (the acceptance of the article into the mainstream). Alexandrews 13:16, 11 March 2011 (UTC)
So, who gets to move a "beta" article to "mainstream"? And what are the standards? I kind of like the idea, but I have also seen what has happened on Citizendium, where they have "live", "developed" and "approved" articles, and only about 10% of "developed" articles have reached "approved" status. If we quarantine new articles in a beta status, we will need a process to review and approve those articles. It would be unfair to leave beta articles sitting un-reviewed for long periods (or else the status would lose its meaning). A possible model for beta article review is the DYK process, but that deals with only a small percentage of new articles, and has recently changed its rules to require editors who have submitted more than 5 articles to DYK to review a submission for every new article they nominate. While such a model for moving beta articles to "mainstream" might be possible, would you want the decision to be made by one editor? Or should we require a discussion, ala AfD, for each beta article. And what happens to beta articles that are not approved for "mainstream"? Are they then deleted? That would be equivalent to sending every new article to AfD. If we treat the beta status as an opportunity to improve an article, how long do we wait for it to reach "mainstream" quality? And ho many times and how often would an article be reviewed before determining it would not reach "mainstream" status?
Yes, I would be happy for one editor to make the decision. As for beta articles, if they do not satisfy the criteria for immediate deletion then they are free to stay as beta articles until such time as they are good enough to become mainstream. The criteria for being a beta article is that the material is of merit, but is not of sufficient encyclopaedic quality to become mainstream. BTW, beta articles should have this emblazoned (maybe as a page background, a bit like "duplicate", and a big banner at the top?), so that users viewing it are fully aware that the article has not met the requirements to be mainstream.
Remember: the purpose of this discussion is to encourage new contributions from new contributors.