Wikipedia:Preliminary Deletion


 * Based on Managed Deletion and this mailing list post.

Introduction
The purpose of this page is to propose a third category of deletion candidates, beyond the speedy delete and yet not to the level of votes for deletion. These will be Preliminary Deletions, to be organised on a page tentatively titled Purgatory (thanks to Jens Ropers ).

The Case for Change
A number of alarming new developments have emphasized a need for a change in speedy delete and VfD categories.

A) Cause: "You mean this isn't a speedy delete?"
 * Inevitably, VfD gets hit with articles that are obvious deletes. They rack up ten to fifteen "delete" votes with no "keep" votes but that of the author, and someone will say, "This is really a speedy delete candidate."  In some cases, that person is correct.  In most cases, that person is not correct because the article is an obvious delete but not a speedy delete.

B) Cause: VfD is too long
 * Every few days, someone on the Village Pump will argue that VfD is too long. VfD reaches over 40kb routinely, even when there is not a contentious debate on it. Indeed, it is so long that some people do not list articles there, and many more do not go to vote there.  This proposal will not reduce the lengthy debates.  Nor will this proposal eliminate the vital deliberative nature of VfD, but it is designed to cut down on the clutter and make it a truly deliberative page.

C) Cause: VfD is overwhelmed
 * On December 21, 2004, there were seventy-one (yes, 71) new nominations to VfD for the single day. Of these seventy-one, approximately forty-five received nearly unanimous "delete" or "redirect" votes. Another dozen were speedy deleted before the conclusion of voting. At best, only a dozen articles had a lack of consensus or consensus to keep. The nominators to VfD were acting appropriately, and yet it was extremely difficult for voters to consider all of the candidates; about a dozen articles had less than five votes. The sheer volume of nominations can make it virtually impossible to reason together and consider articles carefully, case by case, which is what VfD is for. More statistics can be found at User:Johnleemk/VfD statistics.

4) Cause: Inappropriate Speedy Deletion tags
 * Things languish on Candidates for Speedy Deletion because administrators look at the articles and see that the tagged item is undesirable, but, simply put, not fitting the narrow definitions of a speedy delete.

5) Cause: VFD can't be centralised forever
 * VFD has been overburdened with pages for a very long time. Some proposals such as Categorized Deletion would make voting much simpler; inclusionists who think all schools should be kept would merely leap to the separate page assigned to schools. However, the fact still remains that many pages are simply obvious deletes, and thus, would quickly add up, creating a bottleneck on certain pages (i.e., the Categorized Deletion page for biographies).

Core proposal
Proposed: That there be a new category of deletion entitled Preliminary Deletion. Articles listed on the Purgatory page will remain there for 72 hours.

Note: Part of this procedure was originally under a different optional section, but as this proved to be the most popular in the first round of voting, it was merged with the proposal directly.

Procedure:
 * 1) Article is nominated for Preliminary Deletion by adding a preliminary deletion template to the top of the article's page and by recording the nomination on the preliminary deletion candidate page (page to be developed). Or in other words, it's just like VfD, except on a different page. Anyone may make this nomination.
 * 2) Voting will go on following the procedures of VFD.
 * 3) If an article receives two or more "keep" votes, making up at least 10% of the total vote, it will be referred to VFD where deliberation will proceed under current regulations. If it receives fewer than three "delete" votes, it will also be referred to VFD. The creator of a listed page may not vote, but is permitted to comment.
 * 4) If the article fails the threshold above, it will be deleted, and the discussion archived as per standard VFD policy.
 * 5) Articles referred to VfD will be organised under a separate section for articles formerly listed on purgatory. Articles formerly from purgatory may be removed from VFD after two days instead of the customary five.
 * 6) An article which survives Preliminary Deletion shall never be relisted here. It may still be relisted on VFD, if the rules of VFD allow that.

If you didn't bother reading all that, what it means is that this will be just like VFD, except the discussion is over after three days, and the moment a single article gets two or more keep votes not from its creator, which are at least 10% of the total vote, it goes to VFD. Yes, this proposal can be boiled down to one sentence!

Alternative 1
Jens Ropers has suggested that instead of VFD-style voting, the only permissible vote be keep; if nobody votes to keep, the article is deleted, thus not necessitating delete votes. A possible caveat would be not being able to judge how many people actually support deletion, a foundation of current deletion policy.

Alternative 2
Anthony di Pierro has proposed that once an article has undergone a major rewrite, all votes be nullified should the article's listing move on to VFD. This is to avoid a mass number of delete votes overwhelming the keep votes for an article that has greatly changed.

Criteria
Preliminary deletion will be appropriate for any article which fits the following:
 * 1) Clear advertising with no amelioration or mitigation.
 * 2) Obvious vanity articles and homages to private enthusiasms.
 * 3) Articles which innately cannot be NPOV.An example of this would be "list of John Kerry flip-flops", which lasted seven days on VfD.
 * 4) Jokes that are clearly such and where no valid content could be put in instead.
 * 5) Anything else that falls under what Wikipedia is not and could never conceivably be rewritten to be of encyclopedic value.

Preliminary deletion is explicitly not for any pages in the user namespace (userpages, talkpages, or user subpages).

Note: Purgatory is strictly NOT a replacement for VFD, and should never be the default place for requesting deletion of an article; if unsure, list the article on VFD. You see, if you doubt whether the article falls under any of the above criteria, it probably doesn't. Therefore, people are summarily allowed to move requests that they feel don't meet the above criteria for listing to VFD.

Answers to common objections/questions
1) Why not expand speedy deletions instead?
 * Managed Deletion failed for a reason. There's a reason why nobody's drafted a proposal for expanding speedy deletion crtieria — nobody but a few deletionists (or centrists leaning towards the deletionist side) want it.


 * The largest complaint about Managed Deletion was that it placed too much power in admins' hands. A good part of the community distrusts three admins to handle a deletion, so our alternative is to let one admin decide? That makes even less sense.


 * There's another compelling reason not to expand speedy deletion criteria. It's one thing to delete "ioshgohgoaghoeg". It's another to delete a paragraph or two of readable text; these are borderline cases which some admins delete, but some admins don't. Why do we have VFD? VFD is a place where borderline articles are placed when people don't know what to do with them.


 * For example, take a poorly written article on some rather obscure subject, say, a 1920s Bulgarian actor well known within his home country only, for pioneering filmmaking there. Google probably won't yield too many results on him. It may look like vanity. So following our current system, an editor places it on VFD, which basically advertises to Wikipedia: "Hello, I'm an article which is so confusing, nobody knows what I'm about or whether I should even be here. Can somebody help sort me out?" Anyone who knows the actor can easily describe how he is encyclopedic and should be kept. Speedying full-fledged prose destroys this process, and as such, is probably not too feasible.


 * Being stubborn for the sake of your principles is not very pragmatic. Let's face it &mdash; speedy deletion criteria aren't going to be expanded sufficiently to stem the tide of obvious deletes any time soon. Are we going to let VFD wallow in its slothfulness while we waste our time trying to persuade people who won't ever accept this in the first place?

2) This proposal is too complex and too bureaucratic / contains instruction creep!
 * Well, see, there's this thing called a tradeoff. You can't have your cake and eat it at the same time. Either you give people/admins free rein in deletion, (in other words, Managed Deletion or expanding speedy deletion criteria), or you have to control them somehow. To control them, well, you need some sort of complication. But is this complicated? The whole premise can be fitted into one sentence.. If you read nothing but the pertinent sentence describing this proposal, you know enough about it already to understand how it works. Try doing the same for Managed Deletion. This proposal is really quite simple, in the end. So it's your choice, because there are few, if any other, options: Either expand speedy deletion or expand the bureaucracy.

3) We have too many avenues for solving problems already.
 * How many pages do we have for dispute resolution? RFC, RFM, RFAr: three. How many do we have for creating new policies? The policy thinktank, current surveys, RFC: three, at least, in all likelihood, more. How many do we have for plain discussion? Seven; six pump subpages, and the main page itself. How many do we have for improving articles? Cleanup, Peer review, Pages needing attention: Three, again. Now, which of these gets the most usage? The village pump. Yet, even RFC and RFAr have both implemented subpage systems. The listings on Cleanup, Peer Review and Pages Needing Attention are far less verbose than discussion. Yet we have only two pages dealing with article deletion, which is used as much as the pump. Yes, we have subpages. But they're all crowded into one page, and they are merely listings, not lists which can be read separately like the pump subpages. How long can we last? How long before dial-up users can no longer vote on VFD? If anything, the avenues for deletion are way below the mark for the demand they get.

4) Why not centralise deletion on one page?
 * Simply because it won't scale. The average dial-up user won't be able to use VFD much longer at the rate we're expanding; more articles are being listed all the time. We can't expect VFD to handle this forever. For more, you can also see the answer to question three.

5) But isn't having different types of deletions confusing?
 * New pages can be defined into three categories: Obvious tests; pages with topics that are not encyclopedic or simply unacceptable (such as Vanity pages); and pages covering encyclopedic/acceptable topics. VFD handles the latter; speedy deletion handles the first. But the second is lumped into the third. which is quite confusing. It's annoying to have obvious deletes wasting time on VFD for a full five days when another entry could be using that space instead.

6) This problem's not that bad, is it?
 * Take it from someone who got burned out from tackling WP:VFD/Old. VFD's been overloaded for quite a while. Try loading it; at the rate we're expanding, it won't be long before the page goes into the hundreds of kilobytes (including templates). Are you suggesting this problem is non-existent?

7) Why allow only two options?
 * Preliminary Deletion is for handling a specific category of pages. If it's determined that a page listed doesn't fall under the category Preliminary Deletion tackles, then the listing should no longer fall under Preliminary Deletion. It's as simple as that.

8) This won't scale.
 * This is hopefully meant to be done in tandem with other operations to reduce VFD's size such as Categorized Deletion. Hopefully nobody thinks this is the solution to VFD's problems, because if they are, they'll be heartily disappointed. Individual solutions by themselves will alleviate VFD's load slightly, but the whole is greater than their sum. Besides, VFD isn't scaling either. Preliminary Deletion will reduce VFD's load by perhaps 10% to 30%.

9) Why not replace speedy deletion with this?
 * See the answer to question 5. There's a difference between obvious tests and prose that has meaning.

Round 1
Voting closed at 00:00 UTC November 5 12 2004.

The first question posed was: "Should Preliminary Deletion become official Wikipedia policy?"

The possible choices were:
 * Yes
 * Yes, but only permit keep votes (see #Alternative)
 * Yes, but add precautions to prevent exploitation of the process (see #A corollary for handling a possible caveat)
 * Yes, but only permit keep votes (see #Alternative) and add precautions to prevent exploitation of the process (see #A corollary for handling a possible caveat)
 * No

If consensus was not reached, votes for choices 2 until 4 would be added to the total for the first choice.

The second question to be posed was: "Should a month-long trial of the policy outlined at Preliminary Deletion be held before a permanent implementation?"

''Note: The trial's implementation would be based on the outcome of the first question; if pure "yes" or "no" wins a consensus, then the implementation will default to the pure policy without the alternatives; otherwise, the implementation will be based on the choice which wins consensus. Should a trial be held, a second poll will be held at the end of the trial run to judge whether the implementation shall be permanent.''

The possible choices were:
 * Yes, a trial would convince me of this policy's workability
 * No, it works fine/No, it's so useless, a trial wouldn't prove anything else

Round 2
Voting opened at /Vote2 from 00:01 9 December UTC and close at 11:59 23 December UTC.

The vote failed in almost all senses of the word.

The first question posed was: "Should Preliminary Deletion become official Wikipedia policy?"

The possible choices were:
 * Yes
 * Yes, but only permit keep votes (see #Alternative)
 * Yes, but add an incentive to rewrite articles (see #Incentive_for_improvement)
 * Yes, but only permit keep votes (see #Alternative) and add an incentive to rewrite articles (see #Incentive_for_improvement)
 * No

If a minimum voting tally was not reached, the votes for choices two until four would be added to the tally for the first choice.

The second question to be posed was: "Should a two-week trial of the policy outlined at Preliminary Deletion be held before a permanent implementation?"

''Note: The trial's implementation would be based on the outcome of the first question; if pure "yes" or "no" wins a consensus, then the implementation will default to the pure policy without the alternatives; otherwise, the implementation will be based on the choice which wins consensus. Should a trial be held, a second poll will be held at the end of the trial run to judge whether the implementation shall be permanent.''

The possible choices were:
 * Yes
 * Yes, but only if the first question passes
 * No

Round 3
Round 3 is likely to be held some time in late 2005 or later, but will definitely not be held any earlier than June 2005 without some sort of extraordinary circumstance.

The proposed structure of voting will be something akin to the following:

"Which parts of Preliminary Deletion should become official Wikipedia policy?"

Possible choices:
 * 1) Core proposal
 * 2) Core proposal + Alternative 1
 * 3) Core proposal + Alternative 2
 * 4) Core proposal + Alternative 1 + Alternative 2
 * 5) None of them

Voters will be permitted to vote more than once.