Popularity math: story sidebars and the relative strength of OC clamor. · 12:41pm Sep 4th, 2014
Ten is more than zero. However, zero is higher than one, which beats three, which totally devastates two and brings that poor digit to the point where it can no longer speak at all. Then the ranking goes four, eleven, two again -- twice -- poor thing got concussed and wandered off -- and finally finishes off with seven. And that is the order of numerical dominance, but only when counting by negatives.
I can't figure the Top Ten sidebar out.
You've seen it, if you've peeked to the right on any author's story page. Some mysterious piece of code sorts out an author's stories and says 'based on the positive to negative vote ratio, these are the most popular works.' The top four get their short descriptions in the sidebar: everything else is Title Only. Right now, my top four are:
100% Move = 50% Fire
Pinkie Pie vs. The Soufflé
Naked Lunch
A Total Eclipse Of The Fun
(Luna's Lottery Lunacy was in that upper tier for a long time, but then someone added a second downvote.)
Total page views seem to have nothing to do with it. Popular Story and Feature Box aren't factors: #2 never saw either. One downvote may effectively negate about eighty upvotes, which feels like an oddly executive decision for the downvote to make. (I could probably get some Congressional representatives to sue the sidebar.) It's just some piece of code saying 'Of all the horse apples you've smeared across the site, these are what we feel to be the least smelly, and so there is a chance of having people go towards this particular stench next.'
Weird.
[/hr]
On the whole, I haven't done a lot with original characters. Some antagonists, a few background ponies. Ponyville's population has filled out, but only a little. We've met a few Guards (mostly Lunar) and know a bit about the palace kitchen staff. It took nearly thirty 'verse stories before Mr. Flankington actually said something. And on the whole, the response to most OCs has been what I'd expect: dead silence. Because it can be hard for an OC to catch on. Most people may be here for the Mane Cast: 'Tell me about these characters I know, not some idiot stranger I don't care about!' And there's also a certain stigma to OCs, perhaps because of how they sometimes get used. Overpowered self-inserts who only exist to make the world refocus around them... oh, those are out there. Perhaps not in the quantity which so many fear, but enough to provide a moment of hesitation before checking out any personal creation. Just in case.
The largest local OC depot was the ponies of the CDA, and they didn't really draw much initial interest. Hatred, yes -- until their versions in the main 'verse were put on display. Once they became the staff of the Immigration Department... well, fine, says a reluctant Dear Reader: maybe you could do another piece set here. Sometime. But don't push your luck.
And then came Ratchette.
Ratchette... got a reaction.
...okay, eight people. But considering that nopony else has really even drawn so much as an extended blink, having eight people actually saying things along the lines of Who Is This, Can We Get More Of Her, Is It Possible To Borrow Her For A While, and one I'd Ship That came as something of a shock.
Is it what might have been seen as a built-in misfit status with more than a touch of shy outcast and woobie thrown in for a pony who's in exactly the wrong profession for her entire species, because a pegasus is not supposed to have a mark related to devices? Are people just seeing an entire backstory opening up for a young mare who's trying to repair and improve things which she can never actually create? Is it the social uncertainty and fear which almost rewinds back to S1 Twilight and starts that reel going again?
Was it because I said she's pretty?
Along with being Pretty Weird?
There's a little curiosity about Ratchette. And I wasn't going to do that much with her, not for intensive focus -- until those eight reactions came in. And then... well, I could do something. A little day-in-the-life story. It's got a tentative title: Mechanical Aptitude. It shows how she fits into Ponyville, or doesn't. Some Mane Cast members show up, but she would be the focus character. And it would be...
...OC-focused.
Arguable story view death sentence, although I'd like to think I could count on eight.
(Naked Lunch doesn't count. People came for the butcher shop.)
I'm not going to ask what people feel makes for a good OC and I suspect the topic of what represents a bad one has been slightly overdone. But I do want to ask the following:
Have you ever avoided a story just because it was visibly OC-focused? (I realize descriptions may play a major part in that, especially with any short summary which starts out 'Meet my OC!')
For the writers: did you ever take an idea away from an OC and give it to a Mane Cast member or known pony because you were afraid it wouldn't catch on with the OC as the star? Does anyone deliberately avoid creating OCs entirely, filling out their Equestria with named ponies only?
In your opinion, how many auto-downvotes is one OC worth?
But Ratchette got a reaction. A small one... but it was there.
Maybe eight views wouldn't be so bad.
If nothing else, it maxes me out at eight downvotes.
OCs are great, I read a number of OC only stories. It's not about OC, it's about story and character. I know there are people who only come for the Mane6, but lots and lots come for the pony.
More than anything, Ratchette probably got her comments from being something of an attention bait.
If a character is introduced as miserable and slightly damaged, you bet your sweet ass she's gonna get some attention. There's a reason emo people exist, you know :V
That is not to say that Ratchette is a bad OC or that her feelings are unjustified, just that you shouldn't bother yourself to much with the attention she's gotten; "Number of comments" is by no means an indication of the quality of a character.
On to he next topic; I think you're really blowing the OC-hate thing way out of proportions here.
Yes, people will be more likely to pay less attention to an all-OC story, but I've rarely seen anyone actually attack a story for being just about OCs.
And hell, there are plenty of OC-focused stories out there that are petty great
What can I say? I like intellectual pairings. At the very least, Twilight would have somepony with whom she can talk in technobabble. But, as Twilight's Escort Service showed, she definitely wasn't yet ready for that kind of friendship.
In any case, I would certainly read something starring Ratchette. She's a fascinating character in her incongruity. And, if you don't think she'll be enough of a draw, you can always include an intriguing piece of magitech to serve as a butcher shop analogue.
Or... you could – and really should – work on some of your unfinished stories instead, like Orange Is The New Blue (hasn't updated since April 25th), A Mark Of Appeal (no new chapters since March), or Triptych itself, which has been stalled for just shy of a year now. (Last chapter posted Sept. 13th, 2013.)
If there's one thing I would criticize about your writing, this is it. You have an distressing tendency to start off strong when you have a new idea, but all too often you seem unable or unwilling to maintain that focus and actually follow through on finishing what you start instead of constantly chasing whatever Shiny New Idea randomly flits through your head.
Whatever idea you have for Ratchette... just write yourself a few notes, then set it aside and come back to it later, after you've finished one of your unfinished projects. The idea will still be there.
I don't avoid OC stories, I avoid stories about [Author name anagram], a human who is summoned to equestria because only he can defeat the evil of the week and then he marries Celestia and becomes a double alicorn.
Don't judge based on character tags, judge based on obviousness of Gary Stu's in the description.
2427101
...okay, Surprise Update Every Continuing Story Day is now cancelled. (I'm too depressed to write.)
I laugh because my immediate reaction to this blog was to apologize for not leaving better comments. I confess when I notice you've written a story I dive right in. The sad/dark/tragedy tags are typically the only things that make me hesitate.
As for OC-hate I know in the early fandom days, ocs got a lot of undeserved stick, but I'm not sure they still do.
Either way, I confess I'm hopelessly in love with the ocs and the worlds you create with them. Please don't ever think they are unloved just because they're not as well known.
Here's my advice. Free, so you won't feel overcharged. (well, maybe).
Write what you like to write.
That's it. Don't write to the crowd, write what you really like and enjoy what crowd comes to you. They're nicer people anyway. I wrote Genealogy because I wanted to, not because I thought people would flock to an OC 'love' story (and they didn't, but that's OK). I wrote Monster in the Twilight because I had this "What if" idea stuck in my head that wrote the full first draft before I pushed a single key (and then was nearly drowned in the massive love for it, which was/still is a lot of fun). And I wrote the 'Tutor' series because I was irritated by the preponderance of 'Hi, we just met, I love you, let's bang' Romance tagged stories, and wanted to write something more touching and thought-provoking (and funny, because nopony could romance Twilight without getting into hilarious trouble)
Your top ten is figured out not just via vote ratio, but comments and views as well. How they all relate to each other is a mystery, so just go with it.
Nope. I came to fanfiction with a long history of having read vast quantities of non-fanfiction, and non-fanfiction is more or less 100% made up of OCs.
An OC can be a bad character, but that's on the author and not the concept. And frankly, if someone cannot write an OC well they probably aren't going to do too much better with a canon character.
Silly you, thinking downvoters actually read what they downvote.
I don't mind OCs. A bunch of my stories feature them.
The only sort of an OC-focused story I reflexively avoid is the one where the OC is a Human in Equestria... and not always even that, since I read and enjoyed your CDA stories.
But yes, OCs are fine! As great as the canon characters in MLP:FiM are, it's really limiting if all the fics must involve them. There's the whole of Equestria, and the world beyond that, to work with!
The thing about OCs is that they need to be written differently than existing cast. With an already existing character, they already have a personality, voice, attitude, etc. With An OC you need to give them these things, and it had to be done subtly and gradually to avoid show vs. tell issues.
Personally, I prefer a well written OC to the main cast.
2427158
Sarcasm and two bucks will buy you a cup of coffee. (Except at Starbucks.
)
Look, I understand how easy it is to get caught up in a New Idea, or just a New Thing in general, because the New is bright and shiny in its still-sealed packaging, while the Old Ideas and Old Things are just... well, old, and not-so-shiny-and-exciting anymore; their packaging has already been opened, with a few worn spots and fingerprints and a bit of dust on them, and taking them back out of their boxes doesn't offer the same sense of Excitement or Novelty that the New Thing offers when you break the seal on its box for the first time.
Me, I have other hobbies outside of fanfic, and I've been through the same problem with those; i kept taking on too many projects because they seemed interesting, only to discover that a year later, I still had too many projects, and none of them had ever actually gotten done because I was spreading myself too thin across too many different things.
In the end, there are only so many hours in the day. Every hour I spent writing a novel-length story was an hour in which this half-restored arcade cabinet got no closer to being properly restored. Every hour I spent organizing and cataloguing my record collection was an hour in which the novel didn't get worked on and the arcade cabinet didn't get restored. Every hour I spent doing "X", was an hour in which "Y" and "Z" didn't get done, and every new project just added another letter to the list of unfinished projects and helped insure none of them would ever get done.
There comes a point at which we all have to say to ourselves, "Self, we can't take on any more new projects right now, because we already have more stuff on our plate than we have hours in the day to work on them, and any time we spend on the new ones will have to come at the expense of the ones we're already working on and will only help insure that those unfinished projects will stay unfinished for that much longer."
It's a matter of setting priorities. Okay, sure, you've got eight readers who are invested enough in Ratchette to openly express interest in seeing more of her. Let's say there's maybe a couple dozen more who found her interesting, but didn't leave comments to that effect. Now ask yourself: how does that stack up against:
• the 290 readers who are invested in finding out what happens to Scootaloo in Scootalift;
• the 141 readers who waiting to see Diamond Tiara get her comeuppance in A Confederacy of Dunce Caps;
• the 539 readers who want to know who's running around Ponyville making a polychromatic mess of things, and why, in Orange is the New Blue;
• the 447 readers who are invested in Celestia's fate in A Horse Called Sunbutt;
• the 97 readers who want to see how Rainbow Dash gets out of her current predicament in Stupid Direction Face;
and if you want to talk OCs, how about
• the 589 readers who are invested in your OC, Joyous Release, in A Mark of Appeal, or
• the 318 readers invested in the fate of Her, the broken not-quite-an-alicorn, in Tryptich?
(And that's just going by the actual "favorites" count; that doesn't include readers who have the stories on their "read later" list or just bookmark them in their browsers manually, since the site doesn't show stats that way.)
You've already got a significant audience for the stories you've already begun.
635 followers brought in by your multi-part epics.
Weighed against eight people expressing interest in a bit character OC.
You don't need to chase popularity or votes, dude. You've already got them.
2427335
So who's being sarcastic? That was legitimately depressing.
ETA: While we're at it, the freeze on Quantum Castaways is my fault. I exchanged letters with Dust Traveler and he hasn't posted anything new since.
I'm serious. I'm pretty sure I was the final straw. I'm not sure how, but The Evidence Speaks. And Convicts.
2427335
Geeze, man, this is ridiculously harsh.
The writer will write what the writer wants to write. The writer has no obligation to do anything else. The writer isn't your bitch.
I judge an OC story by the same merits as a cast-iron canon story: By the merits of the story. You want to expand on this Incantion Technology specialist (and incidentally, on the technology of Equestria
), go right ahead!
2427368
I'm not trying to be harsh -- well, okay, maybe a little, but what I'm really trying to do is give him a proverbial kick in the butt to get him out of this notion he's somehow acquired (and which seems to run through more than a few of his comments, both in his blog and in the stories themselves) that the stories he writes aren't actually worth anything, or that no one actually cares if he posts anything or not.
(Maybe you don't mean it that way, 2427342, but that's how your continuous self-deprecation comes off sometimes. When was the last time you had anything unequivocally positive to say about yourself or your stories, without immediately tacking a gloomy negative onto it afterward? Do you really think that all those followers and favorite-ers clicked the "follow" and "favorite" buttons by accident, and that any minute now we're all going to wake up and realize our mistake? Because, again, that's kind of how you make it sound.)
Sure, Estee should write what he wants to write -- but that's just it. Is a story about Ratchette what he really wants to write? Or is he just considering it because eight people commented on the character, and now he thinks he has to go down that bunny trail to make them happy, even at the expense of all his other stories -- stories which, presumably, he also wanted to write, and which have a far larger proven audience already? Because that's kind of what "maybe eight views wouldn't be so bad" comes off sounding like; it makes it sound like he thinks he needs those eight views, when even Stupid Direction Face -- his least popular story, judging by the vote totals -- has a proven audience over ten times larger than that. Which, again, makes it seem like he has no confidence in the worth of any of the stories he's already started, despite several hundred favorites and upvotes proving otherwise.
I suppose I'm also trying to make him see that constantly chasing the New and Shiny Ideas, at the expense of his other stories, is going to be self-defeating in the long run -- because the longer those stories go unfinished, the more likely it is that he'll wind up acquiring a reputation as an author who never finishes what he starts... and the more likely it is that the self-deprecating image will become reality, and no one really will care if he posts anything or not, because people won't want to start reading stories that they know will never be finished. (That's what killed a lot of fanzines and anthology comics; the readership got tired of constantly being hooked by stories and comics that just suddenly stopped updating right when they were getting interesting.)
Not even Isaac Asimov could write every idea that came into his head the moment he thought of it. Even Jules Verne, Charles Dickens, Mark Twain, and William Shakespeare had to set some ideas aside for later and finish up the ones they were already working on first. The need to prioritize has probably been the bane of storytellers all the way back to the days when Ug the caveman suddenly thought of this really great idea for a new story about the Buffalo Spirit and the Lightning God while he was still in the middle of telling the story of The Fire Spirit Below The Mountain.
But if they didn't prioritize, we wouldn't have any finished stories at all.
And that would be a shame.
2427368 2427888
This. This. Both of these, in equal measure.
But really, I came here to talk about featurebox math.
2427187 is probably right, but there's one factor that affects the down/up in the sidebar: it's based on a confidence interval rather than the explicit ratio itself. (Source: Discussion here on fimfic a long time back, IIRC from knighty himself, probably in the comments section of yet another news post about the featurebox)
The idea being: let's say you have a story with 10,000 upvotes and 45 downvotes. This is a pretty huge sample of readers, a significant fraction of FIMFiction. So you can be pretty confident of your up/down ratio being really close to just mashing those two numbers together.
Now take a look at that exact same story earlier on, when it only had 1,000 upvotes. Let's say at that point it had 5 downvotes. If you didn't know the "real" number of downvotes at 10,000 ups, and you had to make a guess of what it would be — based only on your 1000-vote sample — then "10000 up:50 down" would be a pretty good guess, right? But when you're scaling up from a smaller data pool like that, statistics tells us that you probably won't get exactly 50 downvotes. Statistically, you easily might end up with 45 or 55, because the data you've got right now is only sampling 1/10 of the people that the larger, more accurate data set does.
A "confidence interval" is basically about saying, "This story will most likely have between X and Y downvotes if it got read enough to accumulate 10,000 upvotes," with a certain degree of confidence (usually 95%, as in "I'm giving numbers wide enough that 95% of my guesses will be correct if we go back and measure later").
Now take the exact same story, except you take your first sample when it has only 100 upvotes. Based on our special omniscient knowledge of 10,000 readers, we know it "should" have 0.5 downvotes at that point — except that's impossible, because someone either downvotes it or not, so your 100-vote story will either end up with 0 or 1 downvotes. (You might even get really unlucky in your sample and have, say, 3.)
This is why confidence intervals are so important. "100 upvotes to 0 downvotes" is infinitely good because you can't divide by 0, but by turning it into a confidence interval, we can say "Yeah, it'll probably have somewhere between 0 and 100 downvotes if we find 10,000 readers."
Why is that important? Because everything I've seen about the sidebar measurements is consistent with placing stories by measuring the up/down ratios at the pessimistic end of their confidence interval rather than at their midpoint.
Your "+10000/-45" story has a tiny confidence interval because you're basically sampling everyone already. So at the 10k mark you know it has 45 downvotes.
Your "+1000/-5" story has a small confidence interval. Pessimistically, at the 10k mark, you can predict it will have 55 (or fewer) downvotes, because you leave some room for error when you're scaling up.
Your "+100/-0" story has a huge confidence interval because you don't have much data. Pessimistically, at the 10k mark, you can predict it will have 100 (or fewer) downvotes, because you're scaling up zero but you have to leave a lot of room for error.
(And if you have a +10/-0 story, you're basically throwing darts; the margin of error overwhelms your real data.)
So in your sidebar list, they would be ranked by the bold numbers above, putting them in this order:
10000 / -45
1000 / -5
100 / -0
Even though the 1000 story has the "worst" ratio, and even though the 100 story has the "best".
I don't think I've ever dodged a story purely due to OCs. If the description gave me the impression it was just one of THOSE fics, then I'd probably skip it, but I like well-written stories about OCs. It expands the world.
Ratchette, I think, had just a lot of room for growth, both in terms of backstory and future features. The relatively little information we got about her was enough to set off the imagination.
2427888
Okay, let's make this simple.
If there isn't one new chapter posted on all continuing stories before September 13th, I will revoke all submissions.
There. Done.
2428158
I kinda hate to say this, but I don't think it's a great idea to be using confidence intervals in any way to rank stories. I know, I know, most people don't have my stats background. And I'm not disputing your statement that we do. I seem to remember that we do something fairly crazy. But... this is just not the type of problem you use confidence intervals for. Sure, it's a parameter estimation problem, and technically confidence intervals are the "right way" to approach that, but... well, not on this scale. Not when you've got tens (or hundreds) of thousands of stories you want to rank-order. No, at that point you just figure out a way to give everybody a point estimate and you run with it. And that's almost painfully easy to do, from a computational standpoint.
tl;dr confidence intervals would waste a lot of memory for no good reason.
The only time I avoid ocs is when the entire story is exclusively populated by oc characters.
2428458
2427888
All right, loves, let's not get our collective knickers in a twist here. Everypony back to their respective corners and cool down for a bit, shall we?
2427070 Yeah, I think Estee nailed it when the author called her a woobie. There's a certain character type that is always going to be popular. Remember how way back in the beginning Princess Luna got all those fics when she was super gothy and just wanted friends?
I think as a control, Estee should really write this short story, and I agree with 2427095 that if you add some cool new magitech worldbuilding that will draw people in no matter what, but you should end the story with Rachette becoming popular and confident, and at the end have ANOTHER OC appear that is even more adorably awkard and listens to "Death Cab for Cutie Mark" or something. Then see how many readers demand you write more about that other OC. I bet it will be double digits.
2427888
2428458
Just quickly on this. Estee is not writing a bunch of stand alone stories. It's an entire interconnected universe. Quite frankly, it's bigger than any one story. Personally, I don't mind the scattershot approach. Each new story adds to the others. Pinkie' slack of feel came up in Tryptich, but the full emotional exploration was in Lazy River. Twilight being asked to bless a newborn foal? We get to see how Celestia copes (or doesn't) after thousands of years in Blessing.
If you look on the various novels as chapters, should Terry Pratchett have written all the 'Guards' books in order, followed by the 'Death' books, and then all the 'Witches' books?
The creative process is a fickle one, and often linked to motivation. None of them have been switched to 'hiatus' or 'canceled.' The updates will come when they come...
On to the actual question of the blog...
To be honest, a story based totally around OCs has got to be either a really good premise, really well written, or from an author I like. It's not that I actively avoid OC stories, there are some great ones out there. It can be a sort of filter, however.
And you can add me to the list of people that would like to see a Rachette story.
2428458 Seriously don't listen to him. There are a lot of stories that yes I sure want to know more about what happens, especially that one pony who's name I can't remember, which is very appropriate considering his talent. I had issues with never finishing anything. It's not like I couldn't get a good ways in, most of the time we're talking at least 10k works, possibly over 100k, but then I'd just stop. I'd just burn out, play video games until some new idea came to mind and never look back. Though I never consciously made such a decision.
Then I came into the pony fandom, decided I guess I was wring fanfiction and started writing. Several ideas in a row actually. Now I'm up to 14 finished stories. Now there is such a thing as focus, but but sometimes you just gotta write what you're gotta write. It might even help you go back and finish other things later. It's very annoying for readers, but far better than if you never finish anything at all. Obviously you should try to finish things, and at least reading where you were last every so often is a really good idea, sometimes that can spark a great idea and get you back in there, but you in no way have to write a certain way.
My take on OCs is both odd and hypcritical. I almost never read anything about OCs because it just usually sounds boring, but on the other hand I almost always have if not OCs background characters running around. I'm also fond of writing in roles for the blind bag ponies. And then to top it off I like most OCs in stories, especially yours, so if it's a story by some author I really like I'll read it even if it sounds boring.
I don't avoid fanfiction with OCs -- hey, the more popular two of my three stories have them -- but if it seems like they're the focus... the author generally has to have gotten my attention with them as a non-main character first. I can only think of one exception, and that's because she's an actress who plays Daring Do in public shows.