Chapter lengths and fic pacing

[Arakawa]

Inspired by Dracos' musings in the other thread:

Fanfic writers, in general, are not concise with their word usage.  Many stories get to 100k or more words, just to introduce the opening cast, or about a 400 page novel.  I re-read https://www.fanfiction.net/s/6942921/1/He-Who-Fights-Monsters this sunday.  93k words, start to finish.  Character introductions, several scenes, many violent murders, family scenes, and a conclusive ending setting the serial killer up as the only role he could fit: one who teaches monsters.

I agree very much, but that makes me ashamed because I've had serious problems with this myself. This led to thinking. I want to break down the ingredients that go into excellent/good/bad/terrible fic pacing, and extract some lessons in what to aim for.

Chapter Length. First, I looked at resources that give word counts of major epics, as well as my own vague memories, and kind of grouped it by chunk/chapter/episode/update length in my mind. Think of these as weight classes for fiction instead of boxing, because it seems to me like chapter length totally changes the dynamic of how your story will work.

10,000-12,000 words: advanced Fanfic. Seemed to occur sometimes in this forum, but kind of rare elsewhere. Done well, this leads to a very immersive fic, with lots of room for both relevant characterization and fun scenery-chewing. Done poorly, it drags on like a Wagner Opera. In retrospect, it was probably way too daunting for a beginning writer like me. If you factor in the fact that fics are serialized and don't pop into this world fully-formed, you also have to write like your life and sanity depend on it. Otherwise, you'll get bogged down and discouraged before reaching any of the story beats that inspired you to start the effort in the first place. Speaking from experience here.

7-8,000 words. Feels like a pretty common length for a fanfic, though I may be misremembering. It seems easy to converge here if you don't have time to generate consistent 12,000 word updates, but have a tendency to overwrite (as many fic writers do). Or you can start from the 3,000 word weight class (below) and add deeper characterization and detail. Interestingly, He Who Fights Monsters averages around this class.

~6,000 words. This is the territory of Lord of the Rings, put here for comparison. This metric is frequently broken by longer chapters where the action overflows 6,000 words, but in general a lot of stuff gets done in LoTR and Tolkien is fine with getting it done more briskly than many fanfics. Writing the story out by hand, twice, probably helped keep the word count down. Hard data and pretty pictures here: http://lotrproject.com/statistics/books/chapters.

3-4,000 words. Another common length for fic. This seems easy for writing good but basic fics, and much harder for something like a novel where you want insightful / thought-provoking / in-depth interactions. Not only do you have to come up with the insight and characterization, but you have to fit it into the plot/action in a way that doesn't bloat the word count. If it bloats the word count, you'll quickly find yourself in the 7-8,000 word category.

1-2,000 words. This seems a pretty doable recipe for getting something of Dungeon Keeper Ami caliber -- not a lot of characterization, or characters come ready made from a base canon and aren't developed much further, whereas neat story advancements have to happen with clockwork regularity. (Although it's still possible to destroy the pacing by devoting 10 chapters to the same long scene, like DK:A did with its recent Dwarf Stomp arc.) Can also think of it as "3,000 words with further subtraction of characterization".

Less than 1,000 words. Usually a sign of something being written for purely temporary meme value.

Why chunk/chapter length? Overall length determines how far the story goes, but chapter length seems pretty well tied to how fast it gets there. There are some extreme cases of bad plotting (where chapters and chapters go by without anything meaningful happening) but in most cases people don't have trouble ensuring one, maybe two significant happenings per chapter. In turn, that dictates how complex the plot can be for a given length of fic.

Arc Length. The next pacing category is how many chunks/chapters/episodes go to make up an arc/book/season/story. Even the way you name arcs in your story leads to different thinking. If it's an arc, there will, and should, be lots of loose threads left to continue the story. If it's a 'season' or a 'book' that seems to imply the payoff has to be complete and satisfying enough that, should the story end after finishing the 'season' or 'book', the audience won't feel too cheated. Any loose threads have to be optional sequel hooks. (Remember the famous quip about how calling LoTR a 'trilogy' is a complete misnomer, since the 'books' fail to stand alone and it's actually one book that was published in three chunks of dead-tree due to the limitations of dead-tree book format.)

At the arc level, what mostly governs the pacing is story beats and how many chunks it takes to deliver them effectively. There's a thing called Beat Sheets http://beatsheetcalculator.com/ which serves to explain why Hollywood keeps making the same screenplay over and over again. (Because it's easy!) Trying to cram every story into Blake Snyder's beat sheet is nonsense (e.g. how do you do 'The Bad Guys Close In' in a story without any designated Bad Guys?) but it's a good illustration of ingredients needed for a story that feels like it's going somewhere -- there has to be varying tension (climax and anticlimax) and a bunch of 'reversals' (which implies false victories and false defeats, whether real or metaphorical). Since any character that is not a corpse has motivations, and those motivations typically drive the action of the story (even if the action consists of having a tea party), and these motivations take time and effort to get satisfied, victory and defeat are semi-appropriate concepts for plotting any character interaction.

I made the following simplified beat sheet as an exercise:

• Introduction Introduce the story -- how this happens is pretty unique to the story.
• Debate Introduce the conflict -- this could be a thematic conflict, or a plot conflict. Either way it hits on the characters' motivations, which causes them to start bouncing off each other. But the situation is not clear, and the tension hasn't been amped up yet. Hence, leisurely debate about what to do.
• Parallel Stories Since different characters want to do different things, the story typically has to branch out in initially unrelated directions.
• Midpoint And Reversal Some character reaches a false victory, then it gets torn to shreds. For a plot-driven story, this typically ramps up the tension as they scramble to adjust plans.
• Nadir Some character reaches a false defeat where it looks like their motivation has been totally thwarted. The story might have to stay there for a good chapter for it to really sink in, with the character reacting to the consequences of their seeming defeat, rather than seeing any way to avert the defeat.
• Climax and Finale Sudden realization about how to avert the false defeat and then a sudden burst of action leading to real victory. Finale unfolds the consequences of the victory, then the arc wraps up.

It's kind of stupid, but makes me realize that it's not that hard to throw together a self-contained framework for things to happen in. That lets you establish a minimum number of chapters, which (after picking your chapter weight category) gives a clear idea of how much word count it will take.

It also makes me realize that the 'beat sheets' (and the stakes of what victory and defeat means) have to vary significantly from arc to arc to keep the overall story from feeling repetitive. Without variation in arcs and story beats, the story will feel like it's going nowhere. Rather like the animated series "Cliff Hanger, Hanging off a Cliff" (see https://www.youtube.com/results?search_query=cliff+hanger+between+the+lions).

Overall Story Length Something like https://www.sffchronicles.com/threads/539481/ is interesting, but not overly useful except for setting a hard upper limit on one's efforts, because most of those series are not known for their good pacing. Something like "if the projected length of your story is longer than Game of Thrones, you can probably find a way to reach your planned payoffs more efficiently".

It becomes obvious that many fic writers who fail to do these back of the napkin calculations, end up trying to write Game of Thrones scale fics, and then get bogged down because while GRR Martin is getting fat paid for his opus, the typical fic writer isn't.

[Arakawa]

Minor factual correction to what I wrote: DK:A is actually in the 3-4,000 word weight class, with occasional longer chapters. What's very odd is that the typical chapter feels much shorter than that.

[Anastasia]

Oh yeah, most fanfic writers are dreadful and suffer from tremendous word bloat. Fanfic authors in general rarely conserve words and often spend them like water, stacking up huge word counts compared to most other literature. To be fair, fanfic authors do lack editors who have to cram a story into an actual book, as well as using a medium that allows unlimited wordage.

I do think chapter length is important, but far less important than arc or total story length. Chapter size is one thing, any can work so long as you're using it effectively as part of a tightly wound and paced arc, rather than a meandering trip that lacks planning and direction. You're so right about planning not being used, as well as discipline to stick to that planning. It can be easy to meander, stretch things out, let the story go in its own direction and so forth and so on. All of these eat up words, weaken pacing and stress the structure of a story. This is so critical and you can tell most fanfic authors are bad about this.

Incidentally, I consider DK:A an outlier regardless, both by succeeding in the Addventure format as well as in spite of making the occasional odd choice. Not to mention that it makes good OCs, which is genuinely rare in fanfiction.

[Arakawa]

Incidentally, why decent OCs in fanfiction are considered rare and extremely hard to do is slightly puzzling to me. I think a bit too much stigma against it developed because in the hands of bad writers OCs tend to (a) be made of cardboard and (b) get forced into an illogically important place in the existing story interactions, causing egregiously implausible behaviours in the canon characters. More thoughtful authors who might have been capable of introducing decent OCs that just do their job for the story and get out of the way, instead opted to avoid the OC mechanism entirely.

If you think about it, it doesn't take all that much to make the DK:A OCs work. E.g. for Jered and Cathy all I remember is that Jered makes crude jokes and gets elbowed in the ribs by Cathy. They're surface mercenaries, wary of working for a Keeper but open minded enough to stick with Ami in spite of that. Basic characterization out of the way, they can participate in the action and develop gradually based on the incidents they get involved in.

By the way, I'm also curious how pacing works for running long RPGs. It seems to be subject to most of the same constraints: a single session can only be so long without stressing out the DM and players; players want something to happen every session; and players want to get to a large, satisfying payoff for the overall campaign / story arc in a reasonable amount of time.

(Unfortunately my only RPG experience was a play-by-post run by Dracos which quickly bogged down, so I can't really comment on this aspect from experience. The format was rather curious: it seems like Q was going to bounce the characters from unrelated scenario to unrelated scenario but eventually things would start connecting somehow. Never got that far, and I didn't worry about it since I was having fun playing a character with no long term thinking abilities.)

[Muphrid]

This is an issue I've pondered at great length and gotten nowhere conclusive, but I think it's always worth discussing because it helps one weigh the pros and cons of various structures.

First, what is a scene? I think each author has a typical idea of what a scene accomplishes. Sometimes they may have briefer or longer scenes, but the bulk of the story is made up of "typical" scenes. This probably varies even for different points in an author's career, or for different works (even if they were written at the same time). But in general, I feel that a scene is the smallest discrete unit of moving the story forward: it should represent a change in someone's understanding of the story--whether that be a character's understanding or merely the reader's. It has to establish the situation and then show what's about the current understanding is changing or has changed.

If a scene is <500 words or so, then the point of change must be rather simple and straightforward to get across. A story dominated by scenes such as these probably comes across as intellectually unengaging.

If a scene is 500-2000 words, there's room to develop more complex points, but it's also relatively brisk in terms of pacing.

If a scene is 2000-4000 words, then things are slowing down a bit.

If a scene is >4000 words, we're really taking our time to get points across, or the scene may be relatively dense in terms of several points that are all changing understanding of the story. A reader might start having trouble keeping all the things that are changing straight.

Because scene divisions are influenced by considerations such as how many different times or places were involved, I think there's considerable room for fluctuation here. And I think it's worth debating what a scene is: is a scene merely what comes between explicit breaks, or do transitions delineate scenes that are not visibly and explicitly broken?



Now, what is a chapter? To me, a collection of scenes. More than the typical length of a chapter, I like to think of the typical number of scenes a chapter is composed of.

If a chapter is one scene, I think that works best to emphasize that the reader should consider what was learned in the course of the scene carefully and digest it. This naturally fits longer scene structures. Used with shorter scenes, it may seem irritating or unnecessarily pace-breaking.

If a chapter is two or three scenes, then you have the ability to string together those distinct changes in understanding into a broader point. There's a feeling of movement or progression that shouldn't be too hard to follow.

If a chapter is four or five scenes, then you're starting to use a more complex progression, and indeed the point might be for people to follow that progression rather than the individual scenes' changes.

If a chapter is six or more scenes, then we're really starting to take a while to tie everything together. Maybe things shouldn't be tied together but held on to and filed away for later purposes.



And finally, what is an arc? I'm not going to answer this question in terms of lengths. I've been doing a lot of this type of thing recently--similar to the beat sheet but tweaked to fit the progression of things that I happened upon on my own. I call it the three ordeal structure, consisting of

• A call to action, in which the protagonist is roused from the status quo, and an opporutunity to act presents itself, culminating in the protatgonist's decision to accept the call;
• A first ordeal, or uncovering the flaw, in which the protagonist's initial actions lead to a temporary success or victory, but in doing so, some flaw or shortcoming is uncovered, and usually, only the protagonist is truly aware of it;
• A second ordeal, or overreaction to the flaw, in which the protagonist attempts to address the flaw in an incomplete, short-sighted, or otherwise non-constructive manner, and their attempts to do so undermine the broader effort, leading to a serious defeat;
• A third ordeal, or cleansing of the flaw, in which the protagonist takes the time to address the issue properly, and in doing so finds success within and without
• A conclusion, in which the protagonist enjoys the fruits of victory and passes the lesson learned to someone else

This is clearly specific to the types of stories I've been telling lately; it would not do for a tragedy, and I wonder whether it would apply to a comedy.

When I was using this for The Second Coming, I could more or less map it like so: in each arc, the call would be one chapter; the first ordeal and second ordeals would be about two chapters each; the third ordeal and the conclusion would take up two chapters (the conclusion being very much shorter). Lately, I've been trying to use this structure instead at the arc level, which required me to subdivide each part to be used at the beat level. I eventually settled on the following beat structure:

• An initial setup that establishes the situation, including an inciting event;
• An expansion from that setup, which deals with the protgonist's emotional and intelectual reaction to the situation, as well as their initial efforts to change it;
• A primary action, in which the protagonist acts and typically fails to change the situation;
• A reaction, in which the protagonist contemplates that failure and considers a new course of action;
• A secondary action, in which the protagonist succeeds, changing the situation to move into the next arc

And it's this generic structure of beats that combines with the overall goals of the three ordeal structure to generate something that is more tailored to a given arc. For instance, in the original piece I'm working on called The Parallel Murders, the second arc (First Ordeal arc) looks like this:

• Setup: Dr. Hazel Baker, having used a device found in her mother's effects, ends up transported to 1999, 18 years in the past. She tries to make contact with her family in that time period, only to be pursued by federal agents and arrested--by her own mother, who was an FBI agent at the time.
• Expansion: Hazel is questioned by her mother about the device and about a murder case that her mother is investigating, as Hazel's use of the technology is consistent with things the suspect has done. Hazel realizes that this is (or was) her mother's job: using some sort of time travel technology to gain evidence about crimes and stop them in the present. Hazel is released when it becomes clear she doesn't have anything to do with the case, but since it relates to a series of attacks back in 2017, Hazel convinces her mother in 1999 to work together.
• Action I: Hazel and her mother pursue a known accomplice of the suspect, but they're stymied when the suspect outwits them, killing the accomplice before he can be taken in.
• Reaction: Hazel, feeling adrift, spends time trying to reconnect with her mother (with whom she'd had a tense relationship) and actively tries to intervene with her past self, interrupting a training session between young Hazel and her karate teacher, whom Hazel recognizes in hindsight as a bad influence. It is for the sake of these relationships--for the secret of why her mother had the device (which she shouldn't have had) at all, for the case that haunted her mother for 18 years--that Hazel regains the drive to continue forward and find the killer.
• Action II: Hazel and her mother work together on a sting to catch the killer when they attempt to kill their intended victim at a body shop. Hazel quickly realizes she's in over her head here--her enthusiasm for solving the case is much greater than her competence as an investigator or agent. Though the intended victim is saved, the danger posed by the suspect prompts Hazel and her mother to hatch a new plan: to go back to 2017 and get information there that Hazel didn't bring with her. Hazel agrees to this, believing they will be at an advantage over the suspect--who has resurfaced in 2017--because of their combined knowledge from two time periods.

You may see where I'm trying to go with this: Hazel realizes she's in over her head in 1999, and agreeing to go back to 2017 is part of her trying to reassert control over the situation. The second ordeal then goes into how she can't really repair her relationship with her dead mother through her mother's 1999 counterpart, how her knowledge of things in the present doesn't really give her an advantage over the suspect in 2017, and in general that her problems--internal and external--aren't things she can solve by looking into the past and future for easy answers.



I think at this point it bears discussing what targets I have in mind for this story. I had in mind that it would be somewhere around 80k words, with the call to action and each of the ordeals taking up around 20k, and maybe a few words left for the ending. Perhaps using the word arc is the wrong word for these subdivisions of the story. Maybe each of these should be considered a "beat" and the smaller-scale beats I've discussed are like "sub-beats". I definitely think this varies based on the overall size and scale of a story. For The Parallel Murders, the smaller beats I've spoken of could be chapters by themselves, but I think it suits me to split them among three chapters: one for the initial setup, one for the expansion into the primary action, and then one for the reaction and the secondary action. This probably plays out as a "short-longer-longer" structure around 4k-8k-8k.

But, that's just what I'm doing now. I didn't have nearly as well-formed ideas before, and some of those pieces turned out all right. It makes one wonder how much of this is structure for the sake of structure.

[Dracos]

A lot of good thoughts above.

It's worth noting that sometimes chapters are short because they are just a daily writing exercise, often one or two scenes at most.  This can sometimes lead to fics that are 50 chapters, but only 30-40k words.

Personally I prefer longer, but it is also a nature of the medium that can allow that kind of more regular interaction in a fiction as well.