Wikipedia:Broken by design

This Broken Chair is broken on purpose. You should not try to sit on it.

Some Wikipedia processes are broken by design. They make it hard to do something because we don't want you to do it.

If you're stuck in a process that seems broken by design, stop and consider whether you really need to use it. If there's another way to do what you want to do that's open to you, use it. If there isn't, ask yourself whether it's really in the project's interest. If the answer is "no", stop there and enjoy the hours of life you just got back. If the answer is "yes", you might be in Wikipedia's other type of process. If you find yourself trying to fix a process that seems broken, consider whether that's by design. If it's been that way forever and nobody else seems that bothered, that's a strong clue.

A common pattern in broken-by-design processes is that they involve you being forced to ask another editor to do something on your behalf. Because this is a volunteer project, and processing these requests tends to be tedious and unrewarding, there is little reason for other editors to respond to them in a timely or helpful manner. And when someone does respond, it is likely that they are motivated to participate in the process to stop you doing the thing you want to do, not to help you. For some specific examples of broken-by-design processes, see below.

Broken-by-design processes arise because it is very difficult to outright forbid something in an open project that has no firm rules. Instead, when the Wikipedia community decides that something is a bad idea, it can only add restrictions that make doing it more difficult. Another factor is that because we make decisions by consensus, there will inevitably be a minority that don't think that it's actually such a bad idea. This leads to compromises where a 'back door' is open left to do something that is otherwise discouraged – but the majority makes sure that door is as stiff and cumbersome as possible.

Examples

Articles for Creation

Articles for Creation (AfC) is the paradigmatic example of a broken-by-design process. It was created in 2005 following the decision to prevent unregistered editors creating articles, to allow unregistered editors to create articles. Subsequently other groups of editors that we don't actually want to create articles, including editors without an autoconfirmed account and editors with a conflict of interest, were required to create articles via AfC. It is unfortunately also where new editors end up if they use the Article Wizard, even though they don't actually have to use it and it's not the way experienced editors work.

AfC is perpetually backlogged (waiting months for a review is normal), makes it far harder to get an article published than just creating it directly, and outcomes vary widely from reviewer to reviewer. Articles from editors who are required to use AfC tend to be uninteresting and/or low quality—that's why we require them to use AfC–so we struggle to find enough volunteers to review them. Those reviewers that stick to it are either motivated by a desire to keep certain types of articles out, leading to unpredictable and steadily-inflating standards, or else are looking to rescue the few rough diamonds that end up there, a sentiment which unfortunately tends to be quickly eroded by wading through piles of rubbish to find them.

What to do instead

If you're not required to use AfC, then don't. If you've already created a draft, move it to the article namespace yourself and remove all the AfC-related templates. You're perfectly entitled to do so, even if the draft is already in the queue to be reviewed. If you haven't created the article yet, create it directly in the article namespace or as a userspace draft.

If you're technically prevented from creating a new article because you don't have a confirmed account, then get a confirmed account – it only takes 4 days and 10 edits, and your first article will benefit from the extra practice. If you have to use AfC because you have a conflict of interest, don't create an article. We don't want it.

Conflict of interest editing

Our conflict of interest guidelines advise editors with a conflict of interest to not edit Wikipedia. But because it's just a guideline, it has to reluctantly admit that if those editors insist on editing, they can do so as long as they disclose their conflict of interest and make an edit request for a volunteer editor to review (or, for new article, use AfC – see above). Unfortunately, editors with a conflict of interest often don't pay attention to that first part. Or if they do, we never hear from them: we only get edit requests from those thick-skulled or extrinsically motivated enough not to get the hint that their contribution isn't welcome.

Much like reviewing AfC submissions, processing COI edit requests is boring, thankless work. As a result, the edit request queue is perpetually backlogged and undermanned. The vast majority of requests are declined out-of-hand, because if the requester was capable of formulating a good edit, they probably wouldn't have to use the edit request system. And responders are rarely motivated to help work towards a better request, because they're busy and they don't get anything out of it.

What to do instead

If you have a conflict of interest with regard to what you want to edit, give up on editing that topic. The end result will probably be the same (remember, almost all edit requests are rejected), and you will save yourself and other editors a lot of time and hassle. To request corrections of clear factual errors on a subject you have a close connection to, it's usually more effective to skip the edit request system and just ask on the talk page (without the template), or to look in the history of the article for a major contributor and very politely ask them for help directly. If you don't have a conflict of interest, you don't need to make an edit request.