Board Thread:Administrative/@comment-5014364-20161112212922

Game of Thrones Wiki has grown very large, much larger than in the rought and tumble days of Seasons 2 and 3:
 * Around 3,400 articles and growing
 * Consistently above 1 million site visits per day during Season 6 itself
 * Consistently at least a quarter of a million site visits even now in the off-season of August to November.
 * 9 million (nine million) site visits in a single day after the Season 6 finale aired. The activity was so high for a week or two after that, such a flurry of activity, that even the Admins couldn't keep up.

In short, we've reached the point of larger wikis such as Memory Alpha or Wookieepedia, in which it is difficult for Admins and other veteran users to just "keep track" of what they personally consider "good" or "done" articles.

So I've worked out my own article ranking system - purely a triage system, you understand, a shorthand way for me to categorize "this is how done I consider this article". It's not like actual Wikipedia.org letter rankings which are confirmed, and I imagine different Admins would disagree, but it's just a shorthand way of saying "how done do we consider this, so we can focus limited work time on other things which need it more?"

The categories run down an alphabetical grading:

A+ = Fully researched information, writing prose thoroughly reviewed, and fully illustrated with all needed image files.

Very few articles will ever reach A+ status nor would we expect them to. Possibly dead characters like "Renly Baratheon" *might* approach this, but conversely, we didn't research Seasons 1 and 2 very well. A character introduced in Season 5 who dies in Season 6 might stand a better chance. Few articles will reach "A+" status, nor is this our objective - we need to focus on getting the "C" class into "B" class.

A = Fully researched information, writing prose thoroughly reviewed, no exhaustive attempt to create new screenshot images that might be needed. Might expend energy to make unique screenshots if important to the article (this doesn't count infobox image).

B = Not "fully researched" in terms of all possible information that exists (rewatching all the Blu-rays thoroughly, researching every possible interview or supplementary material in existence); but the material which IS already present is accurate; the prose flows well with one voice, not just a collection of disparate edits of poor grammar. It might forget some details we missed, but contains no outright false statements. In short, it's as best as it could possibly be *simply by sitting down and rewriting the information in the current article*, without expending the time to research it by rewatching episodes or re-reading articles. Needs at least a few images, if only taken from pre-existing screenshots borrowed from other group shots we already have (don't spend time screenshotting a new one).

C = Contains no outright factual inaccuracies but the prose might be poorly written, particularly in the straightfoward narrative recap sections. To gain "C" status, do a quick fact check for inaccuracies.

D = Contains outright inaccurate sections, or highly subjective parts even by our own standards (extreme fanon speculation presented as factual statements - distinct from speculation which is grounded, mature, and actually warns "it is possible he doesn't like him").

F = Offensively bad and requiring immediate attention; we don't want people to think this thing is factual wiki information in any way and it doesn't represent us at all. The absolute lowest of the low. In terms of "triage", this is stuff we need to drop what we're doing to at least fix up to "D" status.

"Stub" = if actually labelled a "stub", it isn't good or bad, it's just not even attempted yet. Unlike rank "F", a stub article might be "back burnered" for a while; it's not offending anyone by giving "wrong" information, it just has little or no contents - but we might not start it immediately because it would be a huge task.

Some highly trafficked articles need to be at "B" or "A" quality: things like "Jon Snow" or "Daenerys Targaryen". This is all fairly obvious.

But moving forward, given the amount of backlog we have, we need to focus on "at least fix D's and C's into B's" than on "how do we perfect an A into an A+?"

Just quickly re-reading articles to see if the prose flows organically, instead of looking like ten separate editors wrote ten separate sentences. This involves no research, just someone with fluent-level English skill reading to make sure it "flows" grammatically.

In order to write, one must first read. New editors often ask "where do we start": you start by reading and just fixing up the prose, heavy research comes later.

This is an entirely ad hoc system, it's just that from now on when I have "finished working on an article for the moment" you might see me say on the talk page "by my standards this article is now up to B-level quality"...meaning I didn't exhaustively research every episode a character appeared in, I just revised the article based on the text already in it. It's a way of saying "this needs images eventually to be A-level, but this is where I'm leaving off for now". 