Chapter Length and Interactive Fiction

Does anyone have any thoughts about chapter length in interactive fiction? I know it’d vary by genre and author, of course, but I’m not even sure what the general parameters would be—if any. I’m more thinking of how many words the player sees on a single play-through, here, not so much how many words are actually in the scene file.

With linear novel-length fiction, I often hear ‘between 2000-5000 words’, or ‘about two scenes’ per chapter. Of course with ChoiceScript, that raises the question of ‘what’s the difference between a scene and a vignette’ (and possibly a few other questions, too).

I’ve been thinking over this as I consider story flow and pacing, and how that relates to chapter/vignette length. Does overall length of the piece affect ideal chapter length? Does it have more to do with how story events play out? What makes chapters too short or long? Is it more about number of choices than amount of prose? Et cetera.

Any musings welcome!


I don’t think overall length should have much to do with chapter length. Really, I see chapters as more of an indication that the flow of the story is breaking off into a new direction, if that makes sense. Maybe it happens after something is resolved, or a decision is made, a new problem comes up, time shift, etc. A longer story would probably just mean more chapters, rather than longer ones.

I think the rule is to let your content decide. The examples I listed are super important to determining the flow of a story, and if you stick to that rule it might help you figure out what things are dragged on too long or maybe get resolved too quickly. Obviously not all chapters need to be equal, but you can get a pretty good idea of what the norm in your story is.

In the end, chapter shifts are really nothing more than an indication to the reader that “hey, something is about to change”. If you stick to that, your chapters will naturally fall into place.


This is holding true for me. Even the 2000-5000 words per play through seems to be holding up as well.

With regards to “vignette” - A short story or anecdote that presents a scene or tableau, or paints a picture:

I think a singular scene is not enough “meat” for the reader to feel sated; the two scenes seem deep enough to draw the reader in but not too much to digest at a time.

Exactly - some stories are more verbose then others, a hard, fast rule seems to go against the nature of a CoG game.

With all that said, there should be a flow and pace to the choices presented as well. We’ve discussed this before and I think you have a handle on this in practice; you are able to point out where you “feel as if I am just hitting the next-page button”



I am far from an expert here, and I might be tainted by non-fiction notion of what a chapter is supposed to be, but this seems the best explanation and guideline for the issue at hand.

I have been dabbling around with a small project of my own, and the first chapter has something like 30k words in total, but only something like 7000 per playthrough. So I would say that 2000-5000 words seems like a good number to keep in mind.

This seems indeed like a bigger problem. How many choices should a chapter have? It depends on the story, of course, but if you have a solid balance between choices and number of words, you can have longer chapters and no one will complain (well, someone will always complain, but you get the idea).

Do you know about the “Rule of Cool”? Mutatis mutandis and it is basically the same thing, not necessarily in the sense of “how cools is it if I kill 20 zombies with a pencil?”, but in the sense this is a really interesting/good/cool scene or chapter, I really don’t mind if the author keeps me in it for a little while longer.

If your chapter isn’t that interesting (not too interesting chapter development, not interesting action pieces, etc.) you should keep it short and move on to better parts of your story… If it is interesting and you are sure (with the input of your beta testers) that people will like a given chapter, you can make it longer and the reader will enjoy it more.

But hey, I know nothing of this, I’m only talking as a reader. Authors will probably give you better input.


Chapters can be used in different ways. Lots of short ones can help give your work more of a compulsive “page-turner” feel – I think some standard advice for authors of airport potboilers is to keep your chapters short and end each one on a cliffhanger. It can also make the game feel longer as the chapter number increases. I remember my enjoyment when the first Heroes Rise game kept pulling out new chapters just when I was expecting (based on the length of the average CoG game to that point) to have to wind things up!

In my own game, I’ve only got four chapters (plus a prologue) and I use them more like acts in a play. They’re each quite a bit longer than the usual CoG chapter, and bring together lots of material that fits with a particular phase of your uprising.

One reason I do this is that my readers have the option to avoid long stretches of the game by making certain choices – “let’s just kill those strangers,” for example, or “let’s kill/drive away/alienate character X.” If I had frequent short chapters, some of them might end up only 2-3 choices long in some readthroughs, and that would feel a bit silly to me. Games that use lots of short chapters often tend to be a bit more railroady, which keeps the chapters from being too short – again, Heroes Rise leaps to mind.

As those who’ve playtested will know, I’ve also used a different approach to making the game feel longer, which is to make it really long.

We’ll see how it’s received. :slight_smile:


I think chapters are like the phases of a game like - character development and background , introduction of plot / problem , getting into it , betrayal / something unexpected , boss level , ending.

Maybe you would like to do some fine adjustments to it.


This is actually something I’ve been putting a lot of thought, and a lot of back-and-forth thoughts at that, into with my current WIP.

Since it’s the story is pulled in different cases- each one with its own inner story that ties into the overarching plot- my first thought was that each chapter would be a case. That made the most sense, as it would be the “resolution” of the mini-story within the larger plot…

But then I began to doubt that thought-process based on the amount of words/choices per chapter that would mean. As people have pointed out here, 2,000-5,000 words seems like a good amount per chapter… But, take the current case, it already has around 20,000 words (according to both RandomTest and my own testing) of reading, and is still in the beginnings of said case. So if I were to be following my initial thought process then I wouldn’t just be breaking this general rule, I would basically be taking a sledgehammer to it and then burning its remains.

So then I think I should break it up more. Okay. But where? In the middle of a case? That seems awkward too since the chapters wouldn’t be organized by what’s happening but just by word length… and it wouldn’t fit very well into the story itself. It wouldn’t work- it’d feel stilted and strange.

So that won’t work. Do I compromise? Do I create larger “arcs” (or in this case, “cases”) and break up the arks within themselves by “chapters”. What do I do when I reach the end of an “arc”? Do I hit the reset button and make it “Chapter 1, Arc 2”? Or, say there were 10 chapters in one arc, make it “Chapter 11, Arc 2”? Both of those seem… odd, to me. Since in one, it’s not like you’re starting an entirely new story, as everything is still connected to the main plot… but at the same time, you are starting a new story in that you’re working on a new case with its own inner plot, so just adding another chapter onto the end feels… wrong… as well.

So, I think to answer any of your questions, would greatly depend on the type of book you’re writing. There isn’t any “one size fits all” example since each story is going to have certain stopping points that feel natural, and trying to shove a stopping point in there that isn’t natural wouldn’t be doing any good.

In the end, I think the most important idea to follow would be what @seabean mentioned,

Do whatever will feel most natural to your story. Whether that’s 2,000 words or 20,000, the most important thing is that it feels natural.


Nah, you just need to change the perspective. Just change your definition of a chapter.

I’ll personally call your chapter, as a PLOT/ARC.
And the CHAPTER as an act.

Breaking into a museum, stealing certain jewel, and survive waves of S.W.A.T chasing after you on a copter are 3 different CHAPTER, but it is one PLOT.

Keep in mind that that’s the way I define my “chapter,” and YMMV. :thinking: :confused:

P.S. Now when I think about it, Chapter, Act, and Arc should be 3 different things… hmm hmm :thinking:
Ah! IDK! :exploding_head: :face_with_head_bandage:

I’ll seclude myself on a nearby cave and meditate until the inspiration comes down upon me.


True…which is partly why I was looking for ideas as to what is a ‘good’ chapter length for interactive fiction. Maybe it’s my background in theatre, and musical theatre in particular, but it helps me to have some kind of framework in mind when I start setting up events. Three hour theatrical performance in two acts with an intermission is going to have completely different pacing to a variety show, so I’d go into it with different ideas on what I want the framework to be, and then I’d fill in the story.

Same thing with games; a long comic historical fantasy is bound to need very different pacing to a conventional novel about encounters at a cafe. So deciding what sort of pacing is typical, or good, is a helpful starting point for me as I decide where I want to drop or add plot points. (I’m starting to gather that this might be an unusual approach. But at the moment, it’s helping me to make a game and not a sprawling and too-random story, so I’m running with it).

So a sort of sound wave peak? Valley to a mountain, and then level out before the next chapter? That would make sense narratively.

Yes, very true. One of my favourite pieces of advice for authors went something like, ‘If it’s boring to you, then you can be absolutely sure it’ll be boring to the reader. Drop that section, write something that excites you, and then find a fun way to tie it together.’

Or to quote Margaret Atwood, “Hold the reader’s attention. (This is likely to work better if you can hold your own.) But you don’t know who the reader is, so it’s like shooting fish with a slingshot in the dark. What ­fascinates A will bore the pants off B.”

True, but you’ve also broken it up in various ways, right? Seasons for one section, weeks for another—diverse milestones.

I do confess to thinking, especially for modern readers, a chapter that is (or could be) thirty thousand words without having any milestones whatsoever becomes rather interminable. It’s like listening to someone speak in a monotone; most people are used to places to take a breath.

I did a bit of research with these ideas in mind, to see if there are any documented time-related trends with chapter lengths, cos I was curious. One of the things I came across was this very interesting article about chapter lengths in fantasy fiction (a notoriously long-winded genre, even). And it has numbers, and even charts! Some of it was fascinating, especially the spread of longer chapters throughout a book; they tend to form a gradual hump in the middle, or sometimes two humps in thirds.

I’ve also heard people say chapter length should be like paragraph length; varied, to keep the reader’s attention and not go into that dreaded mental monotone territory. I like the idea of having them vaguely even for a game, for my own peace of mind planning, but I’m tempted to experiment.

Regarding instincts and what feels natural…I’m an editor and musician first and a writer second (or fifth, or something), and well, sometimes instincts serve one beautifully, and sometimes not. I find I have to shove my instincts in another room and lock the door when I’m trying to sort out my own pacing. My poor alpha testers can confirm this. XD