Nerding out on Word Count per Lit Page - 3500

[AwkwardMD]

Doesn't that imply that Laurel read the text and manually placed the break?

Or at least skimmed it.

 

[ChloeTzang]

rotflmao. This strikes me as the quintessential Literotica geek thread. LOL. I mean, word counts - I love these posts! The analysis! The theorizing! The speculation!

This is all something I have never thought about. I think someone told me it was around 3500 words per Lit page way back in 2015 and I just accepted that and have worked on that for approximate story lengths ever since. I have concluded I am nowhere near geeky enough! LOL.

 

[NotWise]

rotflmao. This strikes me as the quintessential Literotica geek thread. LOL. I mean, word counts - I love these posts! The analysis! The theorizing! The speculation!

This is all something I have never thought about. I think someone told me it was around 3500 words per Lit page way back in 2015 and I just accepted that and have worked on that for approximate story lengths ever since. I have concluded I am nowhere near geeky enough! LOL.

At least the thread title is a fair warning.

 

[ChloeTzang]

At least the thread title is a fair warning.

LOL. It is....it made me laugh, it's so true. After reading thru all the posts I feel as if I am letting the side down. I am just not nerdy enough for this. LOL But I love it!!!!

I'm sure there's a plot bunny here for the geek day event.

 

[Voyeurkenneth]

I've also seen it where pages will break with a scene if there's a convenient scene break within +/- 500 words

I just benefitted from this so I'm apt to think this is true. But it's hard for me to believe this is being done on a conscious basis for 150 new stories a day or whatever that number is. Or maybe it is plausible--just tedious!

I just tested one of my stories against this 20,000k character limit per page. On 2 of the 3 full pages, the total characters by page (including spaces) did NOT exceed 20,000k (but did exceed 19,000). However, on those 2 pages, they could've accommodated an additional paragraph respectively and still have been below 20k. I used the 'Characters (with spaces)' counter on Microsoft Word.

 

[ElectricBlue]

I try to count only characters that appear on-screen, and it mostly works. HTML is either uncounted or (in the case of entities) counted as the character it represents.

One of the question marks for me is whether it counts the space that falls at the end of a line.

I believe it does, because a space is, in essence, a blank character.

 

[ElectricBlue]

Doesn't that imply that Laurel read the text and manually placed the break?

I tried twice to get my hanging sentence onto the page before, by losing some text. I put a note to the editor both times, saying what I was trying to do.

From that, I speculate that, once the data file has a page block allocated, you've got that page forever and there has to be something on it.

It makes sense to me that a database has a character limited size, and anything in it (including html code) contributes to the count. That's certainly the case for Blogger, which has a 4000 character comment limit, and another site I published on for a short while, which had a 10,000 character block for the primary content. That was tedious, because you had to load and link each page to the last yourself. The feedback was non-existent, so I stopped publishing there.

 

[TheWritingGroup]

... site I published on for a short while, which had a 10,000 character block for the primary content. That was tedious, because you had to load and link each page to the last yourself. The feedback was non-existent, so I stopped publishing there.

That describes this forum, EB.

-Annie

 

[LargoKitt]

For those who might find it helpful, I ran some stats to determine how many Lit words* are on one Lit page.

*Lit words: Number of words as calculated in the Submission box (which may be different than the word counter in Word or other processors)

The average I found: 3497 Lit words per Lit page.

^This average of 3497 was based on nearly 50 full pages from my own stories. If I was more of a math person, I'd be curious what the +/- margins are to see a comfortable range of how many words could fit on a Lit page.

Why might this matter?
-Sometimes clicking on the next page has flow implications for a reader. They might decide to take a break, using the next page as a pause point. I know I have.
-In those cases, it made me consider whether it's beneficial to tag dialogue on a subsequent page to avoid any potential confusion on a reader's part when they resume a story.
-The word count per page estimate can also be helpful if one wants to avoid that tiny amount of story text that sometimes spills onto a new page.
-Overall, your story is your story, and you can (and should!) write it however you want without regard to arbitrary word count considerations.

The ~3500 estimate matches what others have previously cited, but it wasn't ever clear to me if they meant "Lit words" or "Word words." There's also a 3750 estimate mentioned in AH. Obviously, one's "Lit word count per Lit page" will vary by one's style of writing: how blocky your paragraphs are, more dialogue-based, etc. Overall, I don't know if there's a hard "number of lines of text" limit per Lit page.

A couple of puzzles here: A Pages or Word document generally divides pages at about 350 words. So a Lit 'page' is a ten page print story. Another bit of ridiculousness, if you think about it, is that if you read on a laptop or pad you are perusing a long vertical scroll on a horizontal screen. How many 'pages' you see depends on your browser and how you set magnification. The 'silly' thing is that, since it's a scroll to begin with, there's no reason for pages at all. (Except maybe if you stop reading to please your succubus and then return to the story.) And then, if you read on a phone there's a whole different dynamic. (And maybe you are old school and print out your favorites...)

 

[ElectricBlue]

A couple of puzzles here: A Pages or Word document generally divides pages at about 350 words. So a Lit 'page' is a ten page print story. Another bit of ridiculousness, if you think about it, is that if you read on a laptop or pad you are perusing a long vertical scroll on a horizontal screen. How many 'pages' you see depends on your browser and how you set magnification. The 'silly' thing is that, since it's a scroll to begin with, there's no reason for pages at all. (Except maybe if you stop reading to please your succubus and then return to the story.) And then, if you read on a phone there's a whole different dynamic. (And maybe you are old school and print out your favorites...)

I suspect Lit started out in the late nineties using commercially available database software, and the page length is a legacy of that.

 

[alohadave]

A couple of puzzles here: A Pages or Word document generally divides pages at about 350 words. So a Lit 'page' is a ten page print story. Another bit of ridiculousness, if you think about it, is that if you read on a laptop or pad you are perusing a long vertical scroll on a horizontal screen. How many 'pages' you see depends on your browser and how you set magnification. The 'silly' thing is that, since it's a scroll to begin with, there's no reason for pages at all. (Except maybe if you stop reading to please your succubus and then return to the story.) And then, if you read on a phone there's a whole different dynamic. (And maybe you are old school and print out your favorites...)

If you have your word processor show you print layout...

Word and Google Docs have options for web view or non-paginated. I imagine it's a common feature in other word processors.

 

[IWroteThis]

First, yes, I know this is an old thread, but the forum software just recommended it to me . . . and I'm a number crunching nerd myself!

I suspect Lit started out in the late nineties using commercially available database software, and the page length is a legacy of that.

The TEXT field in MySQL held roughly 21,845 three-byte characters in the older utf8 implementation, so you're probably correct.

I just tested one of my stories against this 20,000k character limit per page. On 2 of the 3 full pages, the total characters by page (including spaces) did NOT exceed 20,000k (but did exceed 19,000). However, on those 2 pages, they could've accommodated an additional paragraph respectively and still have been below 20k. I used the 'Characters (with spaces)' counter on Microsoft Word.

I'm just guessing as to which story you used, but as you had a story published the day you made this comment, I went with that one.

If you view page source and search for "pageText:" you will then get what appears to be the saved copy of the story. The p tagged version that is displayed is farther down.

Looking at the first three pages of the story, I got the following character counts:
21082, 21029, 21224

Compare this to the character counts of the displayed story, which fits with what you said:
19616, 19705, 19957

I have a program that tries to predict where the page breaks will fall in my stories. That's mostly to avoid having small fragments orphaned on the last Lit page.

The program is fairly good at predicting breaks. It breaks pages before the paragraph that passes the 20000-character count. There are enough variations to know that the method isn't exact. The differences could be with how characters are counted, or with where the break is relative to paragraphs. It is, at any rate, very close. The method is hardly ever off by more than a paragraph, even if the paragraphs are very short.

The discrepancy comes from things like every paragraph break being replaced with "\r\n\r\n" which adds 6 characters. Every "<" is replaced with "\x3C" which adds 3 characters. Quotation marks are escaped, so there's an extra "\" in front of each one. Of course, the backslash itself would have to be escaped with another backslash as well. I'm sure there are more that I just didn't find in the sample I examined, but those should be the big ones.

I'm curious how accurate your program would be if you added in these substitutions and higher field size?

 

[BellaBestia]

Nerds lol

 

[NotWise]

I'm curious how accurate your program would be if you added in these substitutions and higher field size?

I've tried it, and the effect is too big.

Counting 20,000 displayed characters and placing a page break at the preceding paragraph break is exactly right most of the time. So, tags and non-printing characters aren't counted and html entities are counted as the character they represent.

In a word processor, 20,000 characters isn't exact but it's probably more accurate than a word count.

 

[IWroteThis]

I've tried it, and the effect is too big.

I have no clue what you mean by that. Could you explain it a bit?

Counting 20,000 displayed characters and placing a page break at the preceding paragraph break is exactly right most of the time. So, tags and non-printing characters aren't counted and html entities are counted as the character they represent.

Except, you had to add "most of the time" and the first sentence doesn't prove the second, even if you didn't need to add that phrase. Do you have any actual evidence to support your contention?

In a word processor, 20,000 characters isn't exact but it's probably more accurate than a word count.

What does word count have to do with anything I said? I referred exclusively to character counts, and all of the numbers I mentioned were actual character counts, not estimates from word counts.


EDIT: So, I remembered a certain writer who likes to write big paragraphs of straight narrative. I had to go through a lot of their stories, but finally found one that was more than one page. A character count on the displayed text from the first page came to 20593. Due to the minimal paragraph breaks and virtually no quotation marks, the pageText in the source code still only came to 20772.

 

[NotWise]

I have no clue what you mean by that. Could you explain it a bit?

There are too many of those non-displaying characters. If I count them they throw the character count off to far for the page breaks to be close where they actually fall.

Except, you had to add "most of the time" and the first sentence doesn't prove the second, even if you didn't need to add that phrase. Do you have any actual evidence to support your contention?

"Most of the time" because the page breaks I predict are sometime off by a paragraph. I've programmed the process and check all my stories for possible problems. There's a devil somewhere in the character count, but the difference between the way I'm counting and the way the site counts is very small. Something like counting and the paragraph breaks is far too large to account for the small differences.

What does word count have to do with anything I said? I referred exclusively to character counts, and all of the numbers I mentioned were actual character counts, not estimates from word counts.

Not arguing with you. Word count comes up because the usual advice on the AH is that a page is something around 7000 to 7500 words.

 

[IWroteThis]

There are too many of those non-displaying characters. If I count them they throw the character count off to far for the page breaks to be close where they actually fall.

As I pointed out, you would need to adjust your break points as well. Changing the way characters are counted will change the data you used to train your formula, so the formula would have to be retrained to match. Obviously, throwing the new character counts into the old formula is not going to give accurate results.

The question I pondered was, would the newly trained formula be more accurate on stories not used to train it than the current one is. Depending on the language used and your willingness to share, I wouldn't be opposed to making those changes myself so I could compare the results between the two.

For the record, I tried not to discount the 20K limit out of hand, as that would have been the same round number bias that I suspected led to it being postulated as the limit in the first place. However, as I was able to find a story that surpassed that character count on a single page, I can confirm it is not actually the limit.

"Most of the time" because the page breaks I predict are sometime off by a paragraph. I've programmed the process and check all my stories for possible problems. There's a devil somewhere in the character count, but the difference between the way I'm counting and the way the site counts is very small. Something like counting and the paragraph breaks is far too large to account for the small differences.

The key here is probably when you say "my stories". There's a high likelihood that you write at least somewhat consistently in regards to paragraph length, HTML tags, and quotation marks. Thus, you can write an incorrect formula that produces consistent results, as long as you maintain that consistency. However, such a formula would drop to working "most of the time" when your stories start to vary too much from that standard.

Not arguing with you. Word count comes up because the usual advice on the AH is that a page is something around 7000 to 7500 words.

Thanks for clarifying.

 

[pink_silk_glove]

^This average of 3497 was based on nearly 50 full pages from my own stories. If I was more of a math person, I'd be curious what the +/- margins are to see a comfortable range of how many words could fit on a Lit page.

I tend to write a lot of dialogue, so my number is considerably smaller at 3177 words per Lit page. That's an average of the "Word" word counts and Lit pages for all my stories.

Are you guys omitting the last pages of your stories?

 

[ronde]

Are you guys omitting the last pages of your stories?

I just have Word count the words and I use the page count at the bottom of the published story. I know the page count is not particularly accurate because any number of words can count as a page, but it's close enough for averages.

 

[RocketGrunt]

My stories are a sliver over 3500 words per page.

But looking at it in terms of characters:
One page is 19779 characters, but the next paragraph is 805 characters.
The next page is 20045 characters, the next paragraph is 1053 characters
Third page is 19903 characters, the next paragraph is 553 characters.

It looks to me that it truncates at a paragraph break such that the page is no more than a hundred or two more than 20,000 characters. At least, as measured by OpenOffice.

 

[IWroteThis]

My stories are a sliver over 3500 words per page.

But looking at it in terms of characters:
One page is 19779 characters, but the next paragraph is 805 characters.
The next page is 20045 characters, the next paragraph is 1053 characters
Third page is 19903 characters, the next paragraph is 553 characters.

It looks to me that it truncates at a paragraph break such that the page is no more than a hundred or two more than 20,000 characters. At least, as measured by OpenOffice.

The highest single-page character count I've managed to find is 20593, as displayed.

 

[pink_silk_glove]

I just have Word count the words and I use the page count at the bottom of the published story. I know the page count is not particularly accurate because any number of words can count as a page, but it's close enough for averages.

Not really, since the last page will almost never be full so it will artificially drag down your average quite a bit, unless you consistently write stories that are 80 lit pages long.

 

[8letters]

I've discussed this before. When a story reaches a certain number of lines, Literotica inserts a page break after the next paragraph. A story that's a Wall of Text, it will have a high number of words per page. If a story has lots of short paragraphs (typically lots of short dialog), it will have a low number of words per page.

Some examples:
I Did What I Did - 8846 words over 2 pages, so 4423 words per page. The first page is 4433 words.

Modified Adverts Ch. 05 - 16575 words over 6 pages, so 2763 words per page. The first page is 3220 words.

I average ~3400 words per full page.

 

[RocketGrunt]

I've discussed this before. When a story reaches a certain number of lines, Literotica inserts a page break after the next paragraph. A story that's a Wall of Text, it will have a high number of words per page. If a story has lots of short paragraphs (typically lots of short dialog), it will have a low number of words per page.

Some examples:
I Did What I Did - 8846 words over 2 pages, so 4423 words per page. The first page is 4433 words.

Modified Adverts Ch. 05 - 16575 words over 6 pages, so 2763 words per page. The first page is 3220 words.

I average ~3400 words per full page.

While that's a real outlier in terms of number of words, I Did What I Did has 20,340 characters to go with the 4433 words. Modified Adverts has 19,716 characters to go 3220 words.

The character target sounds a lot more likely than the number of lines target. When I copy those pages out onto my own word processor, Modified Adverts has so many more lines that it's 10 pages on my word processor instead of only 8 for I Did What I Did.

I think the thing that's making Modified Adverts come out at so many fewer words is that the author marks scene breaks with very large numbers of asterisks. And by "very large numbers" I mean two hundred and seventy-nine asterisks. In a row. Every single time. I think it's literally just holding down the asterisk key until the word processor the author was using stops registering it. And of course, copy-pasting it every time.

It's only one word, but with 279 characters for every scene break, it hits the character limit for the page much faster.

 

[NotWise]

For the record, I tried not to discount the 20K limit out of hand, as that would have been the same round number bias that I suspected led to it being postulated as the limit in the first place. However, as I was able to find a story that surpassed that character count on a single page, I can confirm it is not actually the limit.

I don't know where others got the 20K number. Possibly from me. I got it from @LupusDei. I asked on the forum years ago if anyone had studied the site's code or style sheets enough to say what determined the page length. He responded with something like "I saw something about 20,000 characters."

Not exactly carved in stone, but it was a starting point and it works out. I don't see much reason to experiment with other limits because I have no other constraints on what the limit might be, and the 20,000 character limit works pretty well, despite the devil in the details.