What is the ideal chapter length? (2023)
66 points
7 months ago
| 15 comments
| penguin.co.uk
| HN
sambeau
7 months ago
[-]
I am writing children's books. One mantra I try to keep to is to make my chapters ~15 minutes long when read aloud by a parent at bed time. This gives them the ability to be consistent in their bedtime routine.

This way children don't get short-changed and adults don't find they have committed to something they can't get out of. It also allows them to easily give in and do another chapter — after all, it's only 15 minutes.

30 mins reading out loud is about 4,500 words, so 15 minute half-chapters are 2,250 words.

A 2,250 word chapter is also very manageable by a child.

reply
sien
7 months ago
[-]
I read mostly on the Kindle and appreciate shorter chapters just for general reading.

If you read before bed it's handy to know that a chapter has, say 8 minutes to go.

reply
ComputerGuru
7 months ago
[-]
As a parent that’s grappled with this: thank you!
reply
sebastiennight
7 months ago
[-]
When I was re-writing one of my books in 2016, I spent the whole summer asking successful entrepreneurs (my reader audience) what was their favorite business book, and why.

It turned out that the "favorite book" answer was fairly varied (aside from a couple of big-name bestsellers), but the "why" answer almost always revolved around:

- "bite-size chapters"

- "really quick read"

- "short and to the point"

- "one idea, one example for proof, and that's it, go to next chapter"

My book went from ~120 pages to ~150 during that rewrite, but also went from ~10 chapters to ~80.

Not only did the new version outsell the original by a 10x factor, but I'd estimate the completion rate also shot up considerably (based on reader comments about the final chapters).

reply
qingcharles
7 months ago
[-]
James Patterson's books are generally the worst I've ever read, but his short 2-3 page chapters are surely one of the reasons for his success. They're like popping M&M's -- you're thinking, "I can just do one more."
reply
treetalker
7 months ago
[-]
Renowned author Dan Brown, who walks with his feet while not writing renowned books as the author of them, has renowned books that (quite successfully) fall in this same category of books with short chapters with short sentences that read quickly because they are all short, not to mention renowned.

https://onehundredpages.wordpress.com/2013/06/12/dont-make-f...

reply
GeoAtreides
7 months ago
[-]
happy for you you got that bag, but oh my god do I weep for mankind

tiktok and its ilk surely did a number on our collective attention spans (and along with it the capability for nuanced and complex reasoning)

reply
SllX
7 months ago
[-]
Nah, most business books or “ideas” books really are too long. Something I used to do when I was interested in a book is I would go to the local book store, skim the introduction and chapter listings to get a feel for the book, find the chapter that had the same name of the book and just read that. Never walked away feeling like I didn’t get the author’s main idea, and usually most of the rest of the book is trying to pre-sell it and examples of where to apply their idea. Maybe some anecdotes about how it helped them in their life or how other people do it too.
reply
jll29
7 months ago
[-]
I was hoping for some statistics about avg. (mean, mode) chapter length, and changes over time, given that this post comes from Penguin, who have plenty of data at their fingertips.

Still, an interesting blog post.

reply
persnickety
7 months ago
[-]
Ironic that they prefer to tell us stories rather than show the facts.
reply
passion__desire
7 months ago
[-]
Humanities department generally / traditionally haven't believed in insights that come from mathematics.
reply
keiferski
7 months ago
[-]
Depends on what you mean by traditionally. This division is largely a consequence of the battle between scientific-technical education and classics/Latin-Greek education in the 19th century. Prior to that, there was definitely a ton of overlap between humanities and mathematics - consider Leibniz as a prime example.

https://en.wikipedia.org/wiki/The_Two_Cultures

reply
passion__desire
7 months ago
[-]
Conside the "Thesis, Antithesis, Synthesis" theory. It is nothing but Synthesis = λ × Antithesis + (1 - λ) × Thesis
reply
beezlebroxxxxxx
7 months ago
[-]
I wouldn't say "believe" is the most accurate way to describe it.

They're just very skeptical of the insights from statistical analysis of works in the mediums because there usually aren't very good large data sources. Cinemetrics.lv was amazing before the domain was stolen and had a large community of film academics using it. In literature, a big problem has been the way that publishers do not allow access to any of their data, some public access programs are starting to try to overcome that though.

In general, though, most humanities academics aren't interested in statistical analysis or its insights, not because they don't believe in them, but because the insights generally aren't that useful or likely so general as to be not that useful for very specific research programs.

reply
passion__desire
7 months ago
[-]
Computer has become accessible only last 1-2 decades. (Python) Libraries which allow for easy computation of stats have been popular for far less than that. (Wide availability of) People experienced with such methods having been around for even less. So I can understand why outsiders won't bother themselves with using mathematics to find insights.
reply
sandspar
7 months ago
[-]
>Humanities aren't interested in math because it isn't useful for them.

Or they just don't understand it. ~115 IQ field vs ~130 IQ field.

reply
bccdee
7 months ago
[-]
Mathematics can only give you insight into things that can be quantified. When people try do things like (say) perform a statistical analysis of the tone of a narrative, they have to either go through and numerically quantify the tone of each passage (time-consuming & of dubious value) or feed the whole thing through sentiment analysis software (practically useless). You'll learn much more from a close subjective reading of a few interesting passages.
reply
carlmr
7 months ago
[-]
But we're talking about chapter lengths here, something that is trivially measurable.
reply
bccdee
7 months ago
[-]
We're talking about the ideal chapter length. Yes, chapter length is easy to quantify. But what about chapter idealness? If we want to discover the ideal chapter length using statistics, we need a way to quantify how effective (or "ideal") a chapter is, and we need ways to control for all the variables that might skew the ideal chapter length (genre, style, etc).

Of course, these ways don't exist. "What is an effective chapter" isn't even a question with a single answer—it depends on the author's goals (to inform, to entertain, to provoke a very specific feeling, etc). So anyone who tried to pull out a "chapter idealness in relation to word count" graph in an academic context would be laughed out of the room, and justifiably so.

reply
watwut
7 months ago
[-]
The term "ideal chapter length" is not objective measurable term.
reply
passion__desire
7 months ago
[-]
Math has much more complex tools to deal with uncertainties. Error bars, distributions, grouping based on various criteria. learn more about them.
reply
watwut
7 months ago
[-]
None of them helps to clarify what "ideal chapter length" means. This is not about uncertainty in terms of probabilities, this is about the term meaning different things to different people.

This is exactly what is the problem - none of the stuff you suggest has remotely anything to do with what the actual question is.

reply
passion__desire
7 months ago
[-]
But somehow using more words to explain other "words" will solve the problem.
reply
bccdee
7 months ago
[-]
Yes. This process is also known as "thinking about it," and it often proves useful when data-driven approaches fail.
reply
passion__desire
7 months ago
[-]
but somehow it is unbaised and "pure".
reply
watwut
7 months ago
[-]
No one said it is pure. Also, no one said "unbiased". We say it is less misleading then the alternative which amounts to pretending certainty while misusing the math.
reply
bccdee
7 months ago
[-]
No. Just useful.
reply
watwut
7 months ago
[-]
A lot of bad science consist on applying math on subjective things with unclear definitions (ideal chapter) and then insisting that this is useful. You then end up with ridiculousness like a guy from IT insisting that widely popular and simultaneously critically acclaimed book is actually bad, because it does not conform to whatever random statistics he misapplied.
reply
carlmr
7 months ago
[-]
This is just Goodhart's law in action: When a measure becomes a target, it ceases to be a good measure..

The measure still might be interesting, but it shouldn't be used as a target to guide decisions about what is or isn't a good book.

reply
passion__desire
7 months ago
[-]
So according to you, subjective analysis of subjective things is ok because humans are in charge, knowing fully well how many biases we carry around?
reply
watwut
7 months ago
[-]
No. And what you wrote suggests that you do not actually know what "humanities departments" or in this case literature departments actually do.

Also, misapplied statistic is not "objective" despite containing numbers.

reply
passion__desire
7 months ago
[-]
misapplied literary analysis is also exactly that. what is your point?
reply
beAbU
7 months ago
[-]
Long enough that when I'm reading at night and I'm getting sleepy, that the chapter end is always a reasonable amount of pages away.
reply
jb1991
7 months ago
[-]
The larger question for me is the long-term effect of rapidly vanishing attention spans in our population. The short-form format across tweets, tiktok, much on youtube, instagram, SMS vs. voicemail -- even email is often considered "too long" for newer generations -- what is the likely consequence of this? It seems dangerous in the long-term, but maybe not. Is it just a temporary cultural shift, or is it resulting in a permanent shift to humanity's ability to concentrate?
reply
rvba
7 months ago
[-]
There are much more things to consume. There were of course books and newspapers in the past (in fact even more press and higher quality than now), but I think for average person there was much less of everything - you had one book, one newspaper, few tv channels to choose from.

So people want the essence. Also the quality of modern writing is often poor - you dont learn much from articles, since the people who write them must pump few per day

reply
jb1991
7 months ago
[-]
That's an interesting observation -- perhaps the proliferation of quantity is encouraging audiences to care less about engagement time so they can maximize exploration of all that is available.

The ability to speed up the playback speed of Netflix content would be an example of this.

reply
lumb63
7 months ago
[-]
I worry about the same thing. I will say, however, that I worry more about this in the context of news articles, conversation, and other shorter-duration activities. Books are a big commitment, and it is highly disappointing to find out someone stretched a 50 page book into 250 pages to get it published. Sometimes it helps to instill the idea, or provides useful elaboration, but sometimes it is filler with no discernible benefit to the reader. Our time is important and as such, I think there ought to be a referendum against wasting it. That is not to say all content should be short, but that it should be no longer than it needs to be, and that information density should be as high as is tolerable.
reply
iambateman
7 months ago
[-]
This is largely an issue of trust and priority, not attention span, right?

The same people who fly through TikTok videos also regularly sit through a two hour movie or read the entire Harry Potter series.

I think that people can pay attention if they want to, and the cultural shift is _what_ we expect to give attention to.

reply
adonovan
7 months ago
[-]
I too used to believe that, but I've since met too many children who have told me they don't like going to the movies because you have to pay attention for two continuous hours during which you aren't allowed to use your phone.
reply
reidjs
7 months ago
[-]
Perhaps it’s the movies fault for not being interactive or engaging enough
reply
forgotacc240419
7 months ago
[-]
Or for being unnecessarily long; 90 minute films seem scarce nowadays even for kids
reply
wongarsu
7 months ago
[-]
If we take fanfiction as indication of how younger people produce and consume content then the effect is that chapters get shorter while books get longer.

Of course take that with a grain of salt because *waves wagely at the statement full of issues*. But it tracks with other trends in social media. Many people will be able to tell you more about their favorite tik-toker than their favorite long-form-content producer. The length of the snippets doesn't dictate the depth of the story

You could even draw a comparison to the rising popularity of series over movies. People now prefer 6 45 minute segments over one 3 hour installment.

reply
camel-cdr
7 months ago
[-]
> If we take fanfiction as indication of how younger people produce and consume content then the effect is that chapters get shorter while books get longer.

Except, that assumption might not be true.

To set a baseline I looked at LOR statistics:

    62 chapters
    1255 pages
    381103 words
    7759 words/chapter
    20 page/chapter
Then I looked at the 100 most bookmarked fanfics on AO3 in one of the most popular fandoms (Harry Potter):

    218401 average word count
    33 average chapters
    9435 average words/chapter
Now 15 of the 100 fanfics are one-shots with only one chapter, if we exclude those we get the following numbers:

    253386 average word count excluding one-shots
    39 average chapters excluding one-shots
    7544 average words/chapter excluding one-shots

For reproduction, I used the following for scraping and processed the data further in vim:

    lynx -dump 'link to ao3 filter page' | grep -A2 'Words\|Chapters' | tr -d '\n,-' | sed 's/\[[0-9]*\]//g;s/  */ /g;s/Words/\nWords/g'
reply
watwut
7 months ago
[-]
On the other hand, you have people binge watching long series. Which is something that would not exist a generation ago. A movie used to be 2 hours. Now it is something unserious, serious stories have one mega story over 20 hours of a watching time.
reply
carlmr
7 months ago
[-]
I think this is because of two things.

1. the commitment for one more 45-minute episode is a lot less than a 3h movie, even if you then spend 6h bingeing it because you're hooked.

2. Netflix allowed binge watching which wasn't possible before, except after the series ended and you got the complete DVD collection.

reply
watwut
7 months ago
[-]
Yeah, but people intentionally set themselves to binge watch. They wait till whole series is out and plan to watch multiple episodes at the same time.

I agree it is possible only due to technologically advancement.

reply
Conscat
7 months ago
[-]
Slightly tangential, the longest single chapter of anything I've read so far is Fallout: Equestria chapter 37, at 51.5k words. Fallout: Equestria - Project Horizons is the longest single novel, at 1.8m words, but in terms of chapter count (which seems to be what the article means by "length"), I doubt either of those come close, since they merely have 50-odd chapters. Counting the number of chapters seems like a strange measurement to me, in light of that.
reply
captn3m0
7 months ago
[-]
A Memory of Light (last book in Wheel of Time) has a single chapter at 79000 words. Quote from the author about the chapter:

> The chapter is just shy of 79,000 words. It contains (by my quick count) 72 scenes–but only 31 distinct viewpoints, as numerous ones repeat.

reply
trescenzi
7 months ago
[-]
I need to actually finish these. Robert Jordan passing and, at the time, me having no clue who Brandon Sanderson was, meant I stalled out. But now that I know how incredible Sanderson is, and the TV show seems like it’s actually solid I’ve recently been toying with the idea of reading them all through.

I love the insanity of that quote. Sanderson has an awesome ability to write things like a chapter with a dozen viewpoints that still flow so incredibly well.

reply
gamblor956
7 months ago
[-]
Most WoT fans skip books 4-14 when they reread them because nothing happens except a bunch of women smooth their dresses and tug on their hair.

You'll understand what I mean by this once you've read the books...

Sanderson also does a horrible job of wrapping up the series. I won't spoil it but there's a 50/50 chance that you will regret investing time in reading the series once you get to the end. This isn't entirely Sanderson's fault; he only wrote the final 3 novels based on the outline he was given by Jordan's widow. Unfortunately, Jordan did not wrap up all of the story lines in the outline and Sanderson makes not attempt to do so.

reply
ryanar
7 months ago
[-]
I read through the wheel of time last year and enjoyed it. It is a commitment, I think the series slows down a ton and then Sanderson comes in and it picks up, and it was amazing to see all the building of plots see their conclusions. So it is well worth the journey.
reply
taneq
7 months ago
[-]
I waited eagerly for Crossroads of Twilight, read it, found that it STILL didn't address the events at the end of book 8, at ALL, closed it and never went back. I've heard the new ones started advancing the story again but I just... couldn't.
reply
over_bridge
7 months ago
[-]
Skip Crossroads of Twilight. Read the summary on Wikipedia. It's a gimmick that covers the same day and events from lots of different perspectives. None of them are good. That book serves as a slap in the face for the author who made a course correction for book 11 and got it back on track. Then Sanderson naturally has a faster pace than Jordan and brings it home strong. 11-14 are great again for those who survive 7-10.
reply
BeFlatXIII
7 months ago
[-]
You could fit an entire novel in that chapter.
reply
082349872349872
7 months ago
[-]

  Why does clover keep on growing
  Why are there flies in my hay
  Don't they know it's the end of the world
  It ended when you clopped away

    -- "Horsefly" Davis (as heard on DCR)
reply
fuzzfactor
7 months ago
[-]
I like it best when a chapter takes between 10 minutes and a half hour to read through.

Textbooks you're surely going to focus on one chapter at at time and you don't want it to take very long to re-read, which most students will likely be doing, especially when you may have hours of problems to solve for a single chapter. You want to be able to re-read without a lengthy commitment compared to the problem-solving.

Regular non-fiction may not be as universally re-read but it would be more often than with fiction.

With fiction or storytelling a half hour is a good max so there will always be a stopping point within reach, with two chapters not being out of the question lots of times. People may not re-read a fiction chapter as they go along very much, but why not make it easier anyway.

Then again people probably like a chapter that takes whole hour or more but is such a page-turner that it seems like a half-hour or less :)

reply
thih9
7 months ago
[-]
> there is no magic button when it comes to chapter length: the 'right' one is a blend for each novel being written
reply
dehrmann
7 months ago
[-]
> There is no magic button when it comes to chapter length: the 'right' one is a blend for each novel being written. There's no point in worrying about the length of your piece of string if the string itself isn't useful or compelling.
reply
Yawrehto
7 months ago
[-]
1. I wonder how much of the reason people seem to like shorter chapters is because shorter-chapter books tend to have quicker pacing. Would there be some way of controlling for pacing? 2. The title is a bit misleading. I went in assuming it would be some sort of analysis - average chapter length vs number of copies sold, for instance. It's certainly a lovely read and decent history, but it doesn't answer the central question. 3. The username of the poster (samclemens) is a fitting username for a post about books with a particular focus on older 'classics'.
reply
amateurCoder5
7 months ago
[-]
Chapter 11 in the book Gremlins was just (spoiler) "Pete forgot."
reply
camel-cdr
7 months ago
[-]
For me 10k-15k words is roughly the optimum, where it isn't to short, but you still have ampel oportunity to stop at a chapter break.

Although for serial works it's generally the longer the better.

reply
dark__paladin
7 months ago
[-]
For me personally 5k-10k / chapter and 20-30 chapters per book is optimal.
reply
ProllyInfamous
7 months ago
[-]
I'm 2/3rd of the way through Love in Time of Cholera and it's only chapter four (@200+pages, so far)... otherwise, an absolutely beautiful story of unrequited/parasocial love.

By far the worst chapter structure (and "book") is The Sound and The Fury. It seems like an exercise in-the-second-person, gone wrong. Sensorial overload... and isn't the point of a good story <TO MAKE SENSE>?

reply
asplake
7 months ago
[-]
(In fiction)
reply
jayski
7 months ago
[-]
20
reply
gandalfgreybeer
7 months ago
[-]
Assuming all chapters are the same length, I think 10-20 is the sweet spot where I can randomly pick up the book
reply
novosel
7 months ago
[-]
I bid you 21 good Sir.
reply