User:Hut 8.5/RfA and content

This page explains my views on requirements for content creation at requests for adminship (RfA).

The problem

A large portion of people commenting at RfA today expect candidates to have a very extensive history of content contributions. This requirement goes far beyond requiring that the candidate write a few articles. Good articles are enough to satisfy some people's demands, but not all, and even candidates who have written multiple GAs sometimes get opposed for lack of content contributions. Though RfA candidates have been attracting opposition for this reason for some time, I think it has got worse in recent years and that it is a major contributor to the decline in RfA promotions.

The underlying argument of such oppose rationales is that it is essential that any prospective administrator have an extensive record of content contributions. This argument ignores much about the nature of administrative work. Simply put, there is no administrative job which involves writing content. Furthermore, there isn't a great deal of overlap between the skills required to create high quality content and those required to do administrative work, and what overlap there is consists of general skills that are not specific to content creation. We have excellent administrators with poor content creation records, and outstanding content contributors who would make poor administrators. There are a few administrative tasks which involve assessing the quality of high-quality content, such as reviewing pages to appear on WP:DYK or WP:ITN, but these are distinctly minority pursuits.

I'm not arguing here that administrators can be entirely ignorant of content creation. An administrator who is ignorant of the core policies of WP:NPOV, WP:V, WP:OR and WP:BLP is going to run into problems. So will an admin who does not know, at least in principle, how to take a poor quality article and improve it somewhat. On the other hand that's a low bar to meet, and certainly far lower than the standard currently applied at RfA. Nor am I saying that administrators cannot or do not write content. I'm just saying that when they do they are acting as editors, not administrators.

Arguments

This section lists some arguments that are sometimes put forward to defend requiring extensive content creation experience, together with responses. All of these arguments are attempts to provide an objective reason why it is unacceptable to have an admin without extensive content creation experience. Without an objective justification such a requirement is arbitrary and unnecessary.

One point that can be made in response to several of these arguments is that we do not expect potential administrators to have experience in every area of admin work. The number of such areas is very large and few (if any) people have experience of all of them, so if we did require this then almost nobody would pass RfA. What we do expect is that candidates have sufficient common sense to know their limitations and stay out of admin areas where they are not qualified to help. This means that "the candidate isn't qualified to do X" isn't a valid reason to oppose an RfA unless either (a) all or almost all admins do X, or (b) the candidate has indicated an interest in getting involved with X.

• Content creation teaches editors how to interact with other editors / how to comply with policies and guidelines.

Lack of interaction with other editors is a fair reason to oppose an RfA in its own right. Why equate it with content creation? After all, content creation isn't the only way of getting experience interacting with other editors. This is a collaborative project, and almost all areas of it involve some degree of interaction with other people. In addition content creation doesn't necessarily teach editors how to interact with each other.

Something similar is true for complying with policies and guidelines, as almost all aspect of the project are covered by policies and guidelines, and editing almost anywhere will teach some degree of compliance with them. In addition the policies and guidelines that content creators need to enforce are, by and large, not the same policies and guidelines that are enforced by administrators.

• If you're going to delete articles you need to know how to create them.

This argument says experience of writing and improving articles enables someone to make better decisions about when to delete them. The problem with it is that the kind of articles which are sent through the deletion process are not at all like the articles which RfA commenters expect to see. Articles which are sent through the deletion process and stand any chance of being deleted are, by definition, very poor quality - so bad that we're questioning whether the article should be here at all. It is conceivable that experience working with very poor quality articles may help admins make better judgements about deletion. For instance experience adding sources to unreferenced articles may help someone make decisions about closing AfD debates, where the issue is usually whether sufficient sources exist to write an article. However this is not the kind of article work which many RfA commenters are looking for. They want to see an individual article brought up to a high standard, and cleanup work probably won't go down very well.

There is a much better method of deciding whether someone is fit to delete an article. Look for articles they have tagged for deletion as a non-admin. If they consistently make good decisions about such tagging then they can be trusted with a delete button.

• Administrators shouldn't have power over people who write very good content unless they have written some very good content themselves.

Administrators do not have power over content creators. They have the power to block people, but to enforce policies laid down by the community, not because of some personal whim. About the only area where admins have any other authority over people is in arbitration enforcement, as in some topics the Arbitration Committee has ruled that uninvolved admins can apply sanctions to people. However participating in AE is very rare for an administrator, and objecting that an admin isn't qualified for it falls under the "candidate isn't qualified to do X" argument discussed earlier.

• Administrators shouldn't be blocking established content creators unless they have written high quality content themselves in order to ensure that they understand what content creators are going through.

Why stop at content creation, exactly? Why don't people oppose RfAs with rationales like "candidate has no experience enforcing the non-free content criteria, doesn't understand what people who clean out fair use images have to go through, and so shouldn't have the power to block them"? Or substituting any other area of editing? Simple: you don't have to have the same experience as an editor in order to decide whether it is appropriate to block them.

It is rare for established content creators to be blocked, and in any case it is far from clear that experience of creating high-quality content would necessarily improve a candidate's judgement as to whether to block someone for, say, 3RR, as whatever content creation experience they have may not involve dealing with edit warriors in the first place. (This assumes that the candidate actually has any interest in enforcing 3RR, of course.)

• Everyone is here to build an encyclopedia.

This is true, but it doesn't support high content standards at RfA unless it is misrepresented. Arguments along these lines implicitly assume that content creation is the main or most important aspect of building an encyclopedia, which isn't true. Content creators are extremely valuable and the encyclopedia would be in serious trouble without them. On the other hand I could say the same for RC patrollers, bot operators, gnomes or people who write templates. If the project is to survive and flourish it needs all these types of editors and more. On the other hand RC patrollers and template editors will find the admin tools are far more useful in their work here than content creators will.

Consequences

RfA is broken. People have been saying this for a long time, of course, but passing statistics now bear it out. The promotion rate has declined from several hundred a year to just 52 in 2011 and about 20 in 2012 if current trends continue. Unless these statistics improve the number of active admins is going to decline, and the admins we have left will mostly be people who passed RfA a long time ago. This will result in larger admin backlogs, not only because of the reduced number of admins but also because the most active admins are those who were promoted recently. There are three basic reasons why the promotion rate has declined: a decrease in the number of user registrations, the nastiness of the process towards candidates, and raised standards.

Part of the raise in standards is an expectation that admins have more content contributions, which is particularly dangerous when combined with another common RfA qualification. People expect (rightly, in my view) that potential admins have some experience in some admin-related work, such as tagging pages for deletion, or reporting people for blocking. This ensures that a new admin does actually know how to use at least some of their tools properly. However it also means that to pass RfA a candidate must have both extensive content creation experience and work in admin-type areas. Such people are quite rare, and most editors mainly stick to one or the other.

What this means is that the most successful RfA candidates these days tend to be people who primarily write content but who also dabble in admin-type work. This is enough to satisfy both standards. Such candidates are less likely to use the admin tools though, because they mainly work in content creation. Even if these people did start doing more admin work it wouldn't necessarily be a good thing because it would detract from content creation. People who mainly work in admin-type areas, by contrast, have a much harder time at RfA because they usually don't have enough content contributions. This leads to larger backlogs, because these people have to request that admins perform tasks instead of doing them themselves.

The consequence of raised content standards, therefore, is that the people who have the most experience of admin-type work are the ones who are least likely to become admins. That's not good.

Counterarguments

This essay sometimes gets cited in support of the view that admins should have substantial content contributions. The following arguments are cited in support of this:

• We are here to write an encyclopedia. Everything else, including adminship and indeed this essay, plays second fiddle. Well, not quite. We're here to improve and maintain the encyclopedia. That involves more than writing and improving content. Also valuable towards this end are preventing people from damaging content, getting rid of unsuitable content and creating an environment in which people can write good content. The admin tools are much more helpful in these areas. It is certainly true that content creation is valuable, but it doesn't follow from this that admins (or potential admins) should write content.
• All good managers have some experience "in the field". I could argue that this isn't necessarily true, but fundamentally it's a bad analogy. Admins aren't managers. They don't assign work to editors, supervise the work of editors, offer performance feedback to editors, or most of the various things managers typically do. Indeed if an admin tried to do those things I expect they would quickly be told to piss off.
• An argument that writing content shows you can communicate well. The argument is that by writing good prose in articles you show you can effectively communicate your admin actions to other people. Of course you are writing for very different audiences in those two situations and they aren't comparable as a result. RfA candidates should have communication skills, of course, but there are many ways they can demonstrate that other than writing content. Most areas of Wikipedia include a substantial degree of communication with other editors.
• Administrators are frequently called in to settle content disputes. No they aren't. Admins don't settle content disputes. They can intervene to prevent disruptive editing in content disputes but they aren't supposed to take a position on the underlying dispute, unless it's something particularly serious such as a BLP violation. An admin who gets involved in a content dispute isn't able to dictate what the article should say. Admins can mediate content disputes or close RfCs arising from content disputes, but so can other kinds of editors.
• An argument that people should feel refreshed by writing content as it takes them away from "noticeboards, Arbcom cases, reverting vandalism". This probably isn't true for everyone, as people enjoy different activities. Furthermore this isn't a great characterisation of the sorts of things that admins typically do. Disputes which end up at ANI may be some of the most high profile parts of adminship, but they aren't representative. Take a look at the most recent blocks and see how many would be likely to be debated at those venues. It's probably none.
• An argument that writing content helps you spot copyvios. I don't think this is particularly true. By far the most effective way to learn to spot copyvios is to spend time spotting and cleaning up copyvios. That doesn't involve any content creation at all. Copyright violations are almost always written by new, inexperienced editors, as the experienced ones usually know what they're doing. However inexperienced editors rarely write high quality articles anyway. Experience in cleaning up those kinds of poor articles may help you spot copyvios, but writing high quality content of your own probably won't.