User:North8000/How to fix Wikipedia

How to fix Wikipedia - Introduction

Wikipedia is a huge success story but also has some problems. This page identifies fixable ones, describes their causes and describes how to significantly improve them. So these criteria apply:

Problems are ones that have a significant negative impact on Wikipedia and which are fixable
Discussed causes are generally limited to ones that can be fixed. So for example, "gravity" would not be listed as a cause of plane crashes

For brevity this page uses brief wording which is inherently less accurate. For example, saying "fix" instead of the more accurate "significantly improve the situation". Please interpret everything in that context.

Knowledge itself automatically improves things or is necessary for acceptance of effective changes

I have pretty extensive experience and involvement in the areas covered by this page and I think also good and active at analysis of the situation. I'm hoping that folks reading this will assign some credibility based on that and so this does not include arguments for most items. But there are some areas of Wikipedia that are very difficult to learn. Many times a better understanding of Wikipedia operates and other realities will inherently lead to improvements or explain why described measure will lead to them. So this plan also contains contains explanatory content in those areas.

WP:Onus

Analysis

Structure and context

WP:Onus is a redirect which goes to the "Verifiabity does not guarantee inclusion" heading / section in the WP:Verifiability policy. Taken literally, it says little because it merely rules out the absurd argument that merely meeting wp:verifiability is sufficient to categorically force inclusion in any place that somebody wants. The common way to interpret such statement would be that merely meeting wp:verifiability is not a strong argument (or even an argument) for inclusion, or inclusion in any particular place.

The text in that section that seems "onus" related is "The responsibility for achieving consensus for inclusion is on those seeking to include disputed content". This sentence does not withstand a rigorous logical analysis but one common interpretation is that when there is a disagreement, the person(s) arguing for inclusion needs to meet a higher bar / win a consensus in order for the material to be included.

Intent

There was little discussion about the addition of this sentence, so the intent would be that expressed in the addition plus that of the others who allowed it to be kept, which would be based on the addition and it's edit summary. Importantly, the context / location is important. It's in the verifiability policy and in a section about verifiability not guaranteeing inclusion. The edit summary for the addition was brief, in general be about being a safety against verifiability causing indiscriminate inclusion.

What has gone wrong

One thing that gives a sentence or essay fame and immortality is a shortcut which is a simple word or phrase which can be easily using in sentences during debates. This has that, to the extent that a term used in the shortcut ("Onus") is famous and heavily used even though it is not in the policy. And it is widely used to tip the balance against the inclusion of material which the user does not want in the article. While the intent clearly relates to new material, it is also used to tip the balance against restoration of deleted long-standing material. Note that (yes, I know it's not a vote) a "consensus" is a sort of super-majority of arguments, strength of arguments and votes. So it let's those seeking to exclude or remove the material say (in essence) "unless you win with a supermajority, the material stays out".

This is also often in conflict with WP:Consensus policy, which unlike WP:verifiability, is the policy that covers how such debates/decisions are handled.

How to fix the new page patrol backlog

Analysis

NPP has a huge backlog of articles that need review. One obvious contributor to a solution is to get more reviews done, but the real work thee lies in "how?" This applies both at the "simple math" level" and also at a more mathematically accurate & complex level where the current backlog is a state that is determined by a feedback loop. The difference in the latter is recognition that backlog is merely a symptom of what is needed which is more reviewing capacity.

The main (fixable) cause of the problem is that it is unnecessarily difficult to do NPP reviews. "Unnecessarily difficult" in turn impacts all of these areas:

Harder to get folks to become NPP reviewers
Harder for NPP reviewers to become fluent, which is a mathematical necessity, thus lowering the number of folks that do so.
Demoralizing, deterring, demotivating folks who want to know that they are "doing it right" which a nearly impossible feat to achieve in any reasonable amount of time
Subjecting NPP'ers to criticism for not accomplishing those nearly-impossible areas. This matters not only for severe instances, but even milder instances can cause us to lose NPP activity.
More and more difficult work lowers the throughput of article reviewers
Wikipedians tend to do more in areas that they find to be pleasant and less in areas which they find to be unpleasant. So they would tend to make them do less NPP reviewing or make them stop doing NPP reviewing

Wikipedia has a process which reviews screens new articles to decide whether or not an article on that topic is allowed to exist in Wikipedia. New Page Patrol's core mission is to keep that process functional and in place by doing reviews for that criteria. Without this Wikipedia would be flooded into uselessness by billions of advertisements, resume/CV's and other unsuitable articles and "articles".

Keeping this process functioning is the only job that only NPP can do. Every other article problem:

Is NOT unique to new articles
Is something that the other zillion editors can also handle

How to fix it

Change NPP guidance and expectations to say that the job is only to review it for "should this article / an article on this topic be allowed to exist?" aspects. Reviewers are free and thanked to review for other items, but it is not expected. Modify all guidance including training programs, flow charts etc. to reflect this.
Keep up the good work with backlog drives and recognition for work in them. Keep them at their current frequency
Start recognition for consistent reviewing; I have been running this low key at the co-ordination page and it has been successful even in that limited form. The main goal is to help keep experienced reviewers from "drifting off"
State and promulgate the idea that for NPP reviewers (as opposed to help keep deletionists at bay) the recommendation under wp:before is to just do a quick web search. For sourcing-related AFD's the subject is to be whether suitable sources exist, NOT questioning the "conduct" of the NPP'er regarding having done wp:before. And such a shift in focus is itself mis-conduct
Experienced reviewers who accomplished a larger amount of reviews have learned a great deal. Most decisions revolve around wp:notability including the combination of of other factors that go into decisions made in the name of wp:notability. (roughly outlined in Wikipedia:How Wikipedia notability works ) . They've learned a overall view of how the community operates regarding this, and have the confidence (or thick skin) to make judgement calls based on that criteria. We need more of these. Create advanced training (created by reviewers who have achieved ongoing higher review numbers) that provides guidance and confidence towards this.
State and promulgate within NPP the expectation that, for articles dependent on sources for wp:notability, the main task and responsibility of article creation is finding and including sourcing suitable for that purpose. The fact that wp:notability can be judged by assessing sources not present in the article does not negate this.
State and promulgate that under all of the above the main reason for draftifying by NPP is for the creator to find and add suitable sources for cases of sourcing-dependent notability. Or to let it fade out if they do not exist. And that this is an important, well accepted use for draftifying by NPP. Promote the idea that NPP is exempted from the three month deadline for draftifying, and until that is achieved, focus and organize on NPP being actively doing this within the 3 month window. We should a practical, not overly strict standard of GNG compliance when this is done. Edge cases should not get draftified; doing so would cause many problems. Promulgate this as friendly mentoring to avoid more unpleasant things like AFD.
Number 1 is already the official standard of AFC (likelyhood of surviving AFD) but half of the AFC reviewers are not following it and are declining articles for other quality issues. As a result AFC is an overly random and difficult place which many avoid. Through discussions and mutual reviewing participation help get this fixed so that AFC will get used more, providing guidance and vetting so that fewer articles that would fail NPP get to the NPP que and that articles which need something (most commonly GNG sources) to be in article space get that feedback and guidance earlier.

A foundation... how Wikipedia works

It's presumed that folks reading this are already an experienced Wikipedian. This does not repeat the main building blocks off who Wikipedia works such as policies, guidelines, key essays and other venues within Wikipedia. It instead covers aspect that are not clearly defined and generally not understood.

This essay Wikipedia:How editing decisions are made describes the main unknown aspects of how Wikipedia operates. In short, rather than be dictated or guided by tidy rules, everything is guided by the sum of multiple considerations. Rules are deliberately vaguely written to allow for this and find a way to handle that they often overlap and sometimes even conflict.

List articles (including those by titles that don't include "list")

Analysis

List articles also include various other types of articles that don't have the word "list" in them. Other examples include set-index articles, chronology articles, discography articles or other lists of works (e.g. books, movies) relating to a particular creator or performer. The guidance that Wikipedia gives on the existence of these is minimal, weak and scattered.

Wikipedia says that Synthesis is not allowed, but in reality sometimes it is allowed, particularly when it is not disputed. For example, summary style writing which is universal to Wikipedia is technically synthesis. And when that synthesis is summarized in the lead, it is synthesis of synthesis. List articles are generally synthesis. The editor can create a subject which no source has written about as a whole and then collect factoids that fall under that synthesized subject and create an article. While in some cases there may be sources which cover the topic as a whole, there is no wiki-requirement for such.

Solution

The solution is to create a special notability guideline for these. This requires recognizing how Wikipedia notability actually operates. It is the name given to most "can an article on this topic be allowed to exist?" decisions. The SNG would includes these criteria:

Are there sources which cover this list per se? (this is also a measure of degree of usefulness, and also discriminates against compound-criteria editor creations) So "list of the highest mountains in the world" would easily pass this criteria. "List of red-haired NFL quarterbacks" would fail it.
To what degree is it enclyclopedic? / How strongly does it comply with WP:not?

Political bias in English Wikipedia

Neutrality is difficult to define much less achieve. By the terminology and standards of US politics, the English Wikipedia is left-biased on topics related to US politics. And this is the most significant bias controversy of the English Wikipedia. Three reasons for striving for better are:

The more severe cases negatively impact the informativeness of articles
It negatively impacts Wikipedia's acceptance, image and support. The bulk of Wikipedia builds it's acceptance, image and support. Bias works against building theses things.
Neutrality is a (worthy) objective of Wikipedia

A main cause is editors seeking to further their real world objectives, biases and "sides" in a a real wold contest of clash by biasing articles. It's important to recognize this cause even if it's not possible to change it. And then move on to changable things that can reduce it's affects.

Fixes

(just a quick start)

Get rid of the binary concept of a source being "wp:reliable" where they get the unconditional keys to the city for wikilawyers and those not in that club are unconditionally deprecated. . That club is determined by trappings which are those of legacy media and for not getting voted out / deprecated. And go more with actual reliability which is (context-specific) expertise and reliability with respect to the text which cited it (which is a wp:ver context)
WP:weight was intended to apply to "two sides of an issue" coverage but has been hijacked by wikilawyering (in tandem with the wp:rs issue) to exclude coverage on all "I don't like it" items even if they are are not "two sides of an issue" type situations. Fix that.