The front page · 12:27am Feb 29th, 2016
Fimfiction's emphasis on the frontpage as the primary means of story propagation is the biggest shortcoming for what is obviously meant to be a social fanfiction site. All that nonsense about the feature box and popular stories columns, which have obvious self-sustaining amplifying effects. This heavily biases visibility towards ultra-long-runners, (which keep slowly rolling until they finally hit it off) comedic oneshots, (which amass followers, which get a new story into the feature box right off the bat) and very short daily updating stories, (which get into the feature box because they are exposed to new readers daily) which is, in my opinion, neither fair, nor a good thing. (It's perfectly possible for something to be a good thing without being fair.) Who would even expect that you would need time slots for updates like in television...
I would follow people for their reading choices, if they could designate bookshelves, updates from which are exposed to their followers, so that new entries in those would show up in my feed. But there is no such option, and if there is, it's cleverly hidden. On a properly social site, this would be the primary means for people to find stuff to read, there would be people specializing in reading and recommending stuff in this manner. Not the groups -- which tend to be either voluntary serve-all-comers affairs or very narrowly defined by a genre -- but personal tastes of people you find yourself agreeing with. Most of the more engaging stories I remember reading were found by trawling the followed lists of the authors I liked.
Oh well.
P.S. Yeah, I'm a bit annoyed that, apparently, because I posted two singular chapters in a row, (which sounds like it's a bug, but I see no other reason) Aporia disappeared from updated stories column entirely today. Wouldn't you be? 
I'll be the first to agree that the promotion system on this site is seriously flawed. I suspect it's an effort to control the enormous supply of stories, lest the entire site be overrun with self-promotion posts. I've been quite irritated myself with the unusual promotion restrictions. As someone who puts quality above all else, and who has precisely one story to tell (though I have since expanded it to 'one continuity', I'm likewise in a bind because of the current system.
Your observations are correct in that the stories which make the feature box seem to be the categories you provided. In particular, it irritates me when an excessively popular author like Pen Stroke can write the lowest common denominator of quality of a story and instantly achieve the feature box. Also... a 1200:46 like ratio is more than a little ridiculous for that story.
I recently added a chapter to my own story and while it did sit on the front page for a portion of the day, it gained precisely one new like for the effort. It certainly does seem that "success" on fimfiction comes in other packages than one would expect. Even a poorly written clop story can make the feature box. The actual writing quality seems to have little effect. It would be nice if there were a better way to get a good story in front of all the people who would enjoy it.
3817933
Titanium Dragon says there is some correlation between votes and story quality, which mysteriously breaks down at the extreme ends of the scale. He backs it up with some statistics and goes on to determine if views correlate with story quality. The result is similar: There is correlation, but it's not perfect.
And I expect the reason that it does break down is the absence of sideways propagation, further muddled by the self-reinforcing effects of the feature box and similar mechanisms.
3817992
True - to have zero correlation would be rather strange. But there are many factors which influence popularity on fimfiction, not the least of which are stigma and target audience.
The "average" reader here has fairly low standards, objectively. To the point where they may not even be able to distinguish quality above their own preferred range. It's a similar phenomenon to professional critics and how they can detect nuances in quality which the common person cannot. If a person's range of quality goes from 1 to 5, then they may not be able to tell the difference between a 6 and a 10. In fact, they might view the 10 as worse because it employs certain techniques which are devoid of storytelling mechanics that they are accustomed to and have grown to like. An example would be the use of overly-descriptive or irrelevant MLP or pop-culture references littering a story "just for the sake of it". Some people probably feel tickled when they noticed these, but if they do not serve the story in some way, then they shouldn't be there. A lot of low-end people also enjoy melodrama, and while intense reactions are appropriate contextually, typically, melodramatic character behavior is considered a literary mistake.
As for stigma, well, there are a few key areas within the fandom which are considered poor (dare I say taboo) choices to write about. I got lucky enough to stumble across several myself. Exploring certain themes also has niche groups of fans. Like if your story is dark and violent, a lot of people will enjoy it, but a whole lot of people won't because MLP isn't supposed to be dark or violent. *cough*
As for the heat mechanics of the front page, well, as you seem to have alluded to, there's a momentum of sorts which contributes to it. If a story is doing very well, it will continue to do very well for a while. If it slips to the other side of the curve, however, it will rapidly fall off the heat scale. And if it ends up getting a certain amount of dislikes right off the bat *cough* for no reason, then it may never actually recover regardless of quality.
Well, that would be easy enough to test. All one would have to do is to find a story with a high degree of quality that is doing poorly on the site to disprove the theory.
Granted, however, statistically, there are always exceptions. I suppose it's possible that the general rule holds.
3818014
I have a public bookshelf for those. :) While there is no accounting for taste, it probably contains stories that really deserve to be better known objectively. But I still imagine the general rule holds, and aberrations are artifacts of exposure more so than they are artifacts of the audience.
In the end, I'm not writing my own story to be popular. I'm writing it to say something worthwhile and to find my audience. If there are no such people on Fimfiction, well...
...too bad for Fimfiction, I say. :)
3818036
Very true. If one writes to be popular, there is a formula for that already in place. It wouldn't be difficult. But if one has a labor of love, it doesn't matter so much what its score is. "Artifacts of exposure" does sound about right. Exposing a story to all the right people is something which this site seems to have no mechanism for. If you don't follow the predefined formula for story success, and yet your story would be liked by 10,000 people who come to fimfiction, there's actually no way to get it in front of them that I am aware of. Groups are okay, but that can bite you as well because if you put your story in a niche group and the denizens of said group don't think your story represents their group properly, you can end up getting burned for it.
Interestingly, I've found that one thing which works well over time is forum participation. If you engage in discussions within the fimfiction forums (in any topic), people will start noticing your story. I've found that going to the Writer's Group and starting interesting threads on complex literary principles will gain a fair number of new fans even without asking for them.
Now, I have to go tackle that celestial mechanics blog you put out. This is gonna be interesting.
3818058
Actually, it works in real time too: posting appears to correlate with views with a delay of about an hour. Which is evidence for the theory that sideways propagation, i.e. a mechanism for recommending stories conveniently -- like, subscribing to a bookshelf -- is actually very much in invisible demand. People are looking for stories to read by means other than search and rankings, and Fimfiction would do better to be more social. Groups, even when explicitly meant to be used as recommendation engines, are inherently clumsy...