nonlinear scale for visualizing story ratings?

[theWollstonecraftWoman]

Views are not reads. You are creating preferential treatment for popular categories and a negitive handicap for those that aren't. How is that better?

There must be a formula to Combine the two averages to create one rating however maybe it should give more of a percentage of up for current rating. Then use a smaller average by views. Like a 70% rating and 30% views= x for overall rating.

 

[MillieDynamite]

You can't give weight to views. As Mary said, views do not = reads. Why should a person get credit for someone looking at the story's opening and saying, what a piece of shit, I'm not reading this."? Absolutely not right to give any weight to views. If you go that route, which the site won't change so this is purely a waist of time, you need to put a penalty on those with low votes to views ratios. So say 70% knocked off the score if only one in 2000 vote compared to one in 100 for other stories.

The way you have laid it out, there could be a thirty percent increase in the score based on how many people don't read but a small bit, it and those who do read it all hate it. A 2.3 score could receive a 30% bump up that way.

There must be a formula to Combine the two averages to create one rating however maybe it should give more of a percentage of up for current rating. Then use a smaller average by views. Like a 70% rating and 30% views= x for overall rating.

 

[MillieDynamite]

I need to get back to writing rather than wasting time here. Sorry, but bye-bye for now.

 

[ElectricBlue]

I didn’t realize there was a difference from views to actual reads. How does the sit know if you read the whole thing?

I suspect the site has a way of knowing, or at least maybe knowing, but it's not shared with authors.

Chaptered stories give us a bit of a clue, by knowing how many Views the last chapter got. I reckon that's an approximation to actual Reads, because why would you go into a last chapter unless you intended to read it?

Machine language doesn’t understand. If they give it a poor rating it will reflect on that side of the equation. If reads vs fully read is tracked then let’s put a deduction on the opened but not read stories maybe?

My guess, based on my chaptered stories, is maybe one in five, one in six, who open a story, finish it. I see no reason why stand-alone stories would do any better.

 

[SimonDoom]

That's mind-boggling. So the coveted red H is, overall, a sign of mediocrity? But I'm not constantly reading 4.5+ stories and thinking, meh, about average. And certainly it's not been my experience that every other story I see has the red H.

Do you have any explanation for why my perception disagrees so strongly with apparent reality? E.g., are there subsets of stories that inflate the average? Forgiving categories, highly-rated series, etc.?

I can think of two reasons.

Number 1 is the prevalence of chaptered stories. 8Letters' master chart, I take it, includes ALL published stories during that period, treating "Sex Ch. 88" the same way as a standalone story. Late chapters significantly skew numbers upward because of the attrition factor, i.e., only those who like the story are still reading it and voting on it. If you're not tuning in to other authors' late chapters, then you might notice how numerous these stories are and their effect.

I'd be curious to see the difference in percentile for standalone v. later chapter stories, and also how many later chapter stories there are compared to standalone stories.

Number 2 is the talismanic hold of the 4.5 and the red H. It's taken on a significance that is, based on this data, a complete joke. That's why I'd recommend scrapping the current system and basing the red H on a percentile system, adjusting for category. If 8Letters can do this in his spare time then presumably it wouldn't be difficult for the site to automate it. A story at, say, the 75 or 80 percentile would get a red H going forward. Loving Wives authors would love that, but authors in many other categories would be bummed because they'd have some of their red Hs taken away. But one can't seriously claim it deserves special recognition if it's the 55th percentile.

 

[SimonDoom]

How does the sit know if you read the whole thing?

I don't believe it does. Even if it has a way of monitoring whether you have clicked on the last page of a story, it doesn't know for certain that you've actually read it. You might just have clicked the last page to read others' comments. As far as I can tell, there is no way to know how many people have actually read the story.

From my stories I know that, on average, out of 90 who "view" the story by clicking on it, 1 person votes on the story. I can infer from that that the view:actual read ratio is somewhere between 90:1 and 1:1. What is it? Sheer speculation, as far as I can guess, unless there are statistical methods to use that I don't know anything about (I barely know statistics, so this could easily be so).

I suspect the view:read ratio is higher than many people realize, which might seem discouraging because it means far fewer people are actually reading your story than you would think based on the view numbers. I think it might be 10:1. I know in my case that I browse the first pages of many, many stories without ever reading them, so 10:1, or even higher, is probably right for me.

If one believes, otherwise,that the view:read ratio is lower, say 5:1, then it means that the read:vote ratio is much higher. For instance, given my 90:1 ratio, and assuming I'm testing a fairly new story that probably has not been re-read a lot, then if 1 out of 5 viewers read the story to completion, then only 1 out of 18 people who read it vote on it. Is that reasonable to assume? I have no idea. I think we're just guessing. But that ratio seems high to me. The lower the ratio, the closer your "actual read" number is to the vote number rather than the view number.

My suspicion is the view:read ratio is quite a bit higher for standalone and first chapter stories than many people believe, and the ratio drops, but not as much as you would think it would, as the story gets into later chapters. The reason I think that is that the view:vote ratio doesn't change as much as you go through chapters as you would think it would. That suggests to me that even for late chapters for whatever reason many people click on the chapters but don't read them.

 

[SimonDoom]

Chaptered stories give us a bit of a clue, by knowing how many Views the last chapter got. I reckon that's an approximation to actual Reads, because why would you go into a last chapter unless you intended to read it?

I think a plausible explanation is this: Your Sex chapter 27 shows up on a new story feed, either the overall feed or the category feed. Many, probably most, of the potential readers who view the feed have never heard of your story before, but it looks interesting, so they click on it to get more information about it before clicking on the first chapter, or as a way of possibly navigating to the first chapter. That's a fast way to do it, right? Click on chapter 27, check it out briefly, maybe look at the first paragraph, then click on your name, and go to the first chapter. I think you never go wrong if you assume Internet users are ignorant and lazy.

We're both just speculating, and we've discussed it before. I'd be curious to get someone with more mathematical background to weigh in. The flaw in your hypothesis that late chaptered views approximate actual reads is that, if that's so, one would expect view:vote ratios to change dramatically as chapters progress. But they don't. They change some, but not by a factor of five or six. If one assumes, as I think is logical, that actual read:vote ratios don't change a lot over time, and in fact if anything should get lower, then your assumption doesn't make sense. The assumption doesn't fit with what we're actually seeing in the view:vote data. The data is much more consistent with the hypothesis that, for whatever reason, maybe the one I've guessed but maybe some other, many people continue to click on late chapters but don't actually read them.

 

[joy_of_cooking]

I think everyone here is saying "read" to mean "read to end"?

I prefer to take views at face value. If someone reads the first sentence and bails, that's a read in my book. I had my chance and blew it.

Also, I think the views:vote ratio varies enormously across categories. @SimonDoom is quoting 10 views per vote. That's near what I saw for a 750 word romance story, but my full-length BDSM and Fetish stories are more like 100 views per vote.

Wild guess, but it makes sense considering the diversity of content in some categories. Romance is a particularly narrow one, I hear, and I wrote to that formula. Fetish is a very, very big tent, so it makes sense that there are a lot of people who come in, take a look around, and say, "yeah, wrong fetish."

But the "yeah, wrong fetish" people don't dissuade me from equating views with reads. I've had a bunch of comments along the lines of "I'm not into X but I liked your X story!" So I know sufficiently compelling writing can overcome a fetish mismatch, and I take "wrong fetish" as "wrong fetish, and not written well enough to keep me around despite the wrong fetish."

 

[joy_of_cooking]

(waits on tenterhooks for @intim8 to come in with his big dataset to show that actually Fetish stories average 10 views per vote just like Romance and mine are just unusually bad for keeping readers interested.)

 

[AwkwardlySet]

I think everyone here is saying "read" to mean "read to end"?

Nope. It means that every time you click on the story it counts as a view. The website doesn't store the information about who has already seen the story, so every new visit counts as a new view. A reader who clicked on your story just to check its length and tags still counts as a view. You can test this by opening one of your stories in a new tab. Check the number of views in your control panel and then refresh the tab where your story is opened(the first page of the story only). You will see the increment in the view counter every time you hit refresh. It's easy to imagine the same people revisiting the story again and again, to read some titillating paragraph or something like that. So, as you can imagine, it's hard to estimate the true number of reads, even including those unfinished ones. My rough guess is that the number of individual reads after, say, a several months period is reads = views/3 or something like that, and that the ratio only becomes worse with the progression of time.

 

[8letters]

Number 1 is the prevalence of chaptered stories. 8Letters' master chart, I take it, includes ALL published stories during that period, treating "Sex Ch. 88" the same way as a standalone story. Late chapters significantly skew numbers upward because of the attrition factor, i.e., only those who like the story are still reading it and voting on it. If you're not tuning in to other authors' late chapters, then you might notice how numerous these stories are and their effect.
:
Number 2 is the talismanic hold of the 4.5 and the red H. It's taken on a significance that is, based on this data, a complete joke. That's why I'd recommend scrapping the current system and basing the red H on a percentile system, adjusting for category. If 8Letters can do this in his spare time then presumably it wouldn't be difficult for the site to automate it. A story at, say, the 75 or 80 percentile would get a red H going forward. Loving Wives authors would love that, but authors in many other categories would be bummed because they'd have some of their red Hs taken away. But one can't seriously claim it deserves special recognition if it's the 55th percentile.

First off, I want to give credit to @Bramblethorn of presenting to me the idea of ranking stories in a category by percentile. They did that in a thread discussing how there could be an overall ranking of authors.

Secondly, I think SimonDoom's idea has a major flaw. Let's say that only the top 25 percentile of stories should get a red H. Because late chapters skew significantly higher, the majority if not the vast majority of the stories with red H's will be late chapter stories. I can see authors and readers being seriously unhappy about that. And don't say, "Make a separate percentile scale for stand-alone stories and chapter stories". The primary way of determining if a story is a chapter or not is the story's title, and that is very, very dirty data. Literotica gets it wrong all the time. Some authors such as Silkstockingslover intentionally don't use chapter numbering because it hurts their view count. If it impacts the likelihood of getting a red H or not, that will give authors even more incentive to disguise that their chapters are indeed chapters.

 

[joy_of_cooking]

Nope. It means that every time you click on the story it counts as a view.

I understand. I'm saying that when people distinguish between views and reads, that's because they consider "read" to mean "read to completion." I'm arguing that we should do away with the views vs reads distinction and accept any view as a read.

 

[joy_of_cooking]

Well, there's some nuance here because probably some views are nonhuman page loads. E.g., by Google indexing the page or something. Let's ignore that for now.

 

[Bramblethorn]

You can't give weight to views. As Mary said, views do not = reads. Why should a person get credit for someone looking at the story's opening and saying, what a piece of shit, I'm not reading this."? Absolutely not right to give any weight to views.

Other issue with views is that they go on increasing over time. A story posted in 2000 will probably have more views than a similar story posted in 2024 because it's been around longer; that doesn't make it better.

 

[Bramblethorn]

And don't say, "Make a separate percentile scale for stand-alone stories and chapter stories". The primary way of determining if a story is a chapter or not is the story's title, and that is very, very dirty data. Literotica gets it wrong all the time. Some authors such as Silkstockingslover intentionally don't use chapter numbering because it hurts their view count.

Maybe you mentioned this already, but for series that already have at least two parts, wouldn't the series navigation links be more reliable than the title?

Here's one of SSL's stories. Title is '"Mom-Son": Anal Virginity' which is not clearly a series: https://www.literotica.com/s/mom-son-anal-virginity

But you can tell it's a series by looking at the "Story Info" box. Here's what it looks like for that story:



On other stories that aren't part of a series, that middle icon is missing:





If you click on that middle icon, it'll show a little bit of info about the series. That doesn't require reloading the page; it's already in the page source code, like so:

<p>Part 3 of the 3 part series</p></div><div class="bn_au"><a href="https://www.literotica.com/series/se/564329" class="bn_av">&quot;Mom-Son&quot;</a></div><div class="bn_an"> ...

That gives you not only the fact that it's part of a series, but a unique identifier for the series. If you go to https://www.literotica.com/series/se/564329 you can then see the full series in order.

There's also this sidebar on the last page:



Of course this only works if the chapters have been defined as a series. An author could just post a bunch of stories without chapter numbers and not flag them as a series, and then there'd be no way to recognise them as such short of actually interpreting the text. But that kind of thing would probably greatly reduce the view/score effects of a series anyway.

 

[8letters]

Maybe you mentioned this already, but for series that already have at least two parts, wouldn't the series navigation links be more reliable than the title?

Here's one of SSL's stories. Title is '"Mom-Son": Anal Virginity' which is not clearly a series: https://www.literotica.com/s/mom-son-anal-virginity

Yes, I'm pulling that information. Authors don't use that feature consistently. You picked an interesting example:

SSL included three of the stories into the series but left out the last chapter.

Of course this only works if the chapters have been defined as a series. An author could just post a bunch of stories without chapter numbers and not flag them as a series, and then there'd be no way to recognise them as such short of actually interpreting the text. But that kind of thing would probably greatly reduce the view/score effects of a series anyway.

I don't know. Here are some recent stories by one author:

The stories in the red box seem to share the same characters - Jack, Terry, Lauren, and Lucy. But there's nothing in the titles to indicate that they are a series. They aren't defined as a series by the author. Again, from when I categorized stories into chapters and stand-alone before, chapter one stories had a third less views than the comparable stand-alone story, and the views dropped off rapidly from there. By making the series entries look like stand-alone stories, the author is probably getting much higher views than if she had labelled the chapters "Ch 01", "Ch 02", etc.