Data, big or small, will never write a good novel.
In so much of our daily lives, there are simple problems with simple solutions, or simple problems to which a few general rules can be applied to solve them. Consider a window that won’t close.
The rule “don’t force it” will prevent you from breaking the frame or hinges, and might also lead you to look for the blockage, free it, and close the window normally.
Then there are simple practical ways of helping yourself around things you find problematic.
For instance, I have trouble with short term memory, so I write a lot of things down. My Google Calendar is packed with notes and appointments, and I keep my phone near me at all times, mostly for timekeeping and remembering things.
Think of all the methodologies you had to learn in school and college. All the situations where there was a right way to do stuff… or several right ways.
If you know what science is – you’ve read your Karl Popper, you understand the principles of philosophy of science, then you’ll understand exactly why there are scare quotes around “the” in the title.
A lot of people who think they understand science will tell you “science is a methodology, not a set of rules or knowledge; science is not about what we know, but about how we know it.” This is true, of course.
But it it’s a misleading representation. Too many people who have never done science, think that it’s about acquiring knowledge by applying The Scientific Method. As if this were something repeatable.
You keep hitting a problem with The Scientific Method, and eventually it yields a solution.
But science is more abstracted than that.
The recursion is intentional. Because the scientific approach is to discover, invent or design the optimal method for each problem. Each problem, each area of knowledge, requires its own method. The scientist’s job is to invent the method that matches the problem.
The rigorous scientist discards the method as soon as it has yielded results, and looks for a new method to see if that will yield the same results.
You could be forgiven for thinking that the title you choose for your book will make a difference to how many people read it, because it will.
So you might also be forgiven for thinking what an awful lot, possibly most, other people think:
If you could only discover the underlying rules behind the titles of successful books, you could apply those rules to create a successful title.
I’m going to have a damn good try at showing you why that’s impossible.
You’re thinking, ‘a great title doesn’t make a great novel’ – the title might be awesome, but the book has to meet the expectations it creates, has to deliver on the promise; has to live up to the title, otherwise it’s a one way trip to Refund City.
But… if you have written a great book, then if you can apply the Ten Rules for Killer Titles, then you can find a title that will do your book justice; that will attract the readers who will discover the book, and you’re made, whereas if your title sucks, then it doesn’t matter how good the book is, because no one is going to discover it.
Of course not. But you saw that coming. Actually, it’s nearly true. You can certainly discourage readers with a bad title. Can you encourage them with a good one? Kinda.
Big Data isn’t anything new. But for the last few years, we’ve really, truly, had the processing power for it. It’s yielded some amazing results. Read the Wikipedia article, though. All of it.
Through Big Data, Google is a very effective search engine. Through Big Data, you can gather a lot of information about the topic you want to investigate, and apply statistical analysis to discover trends, correlations, discrepancies, and these can lead to discoveries, to new knowledge.
And since we want – that is to say I want, and I’m assuming you do, since you’re reading this – to become better booksellers as well as better writers, then Big Data must be able to tell us something about what makes a good book title.
My excuse is that I’m expected to write these compelling articles, that are at once obvious clickbait and deep, valuable content – knowledge and insight that you can apply immediately to improving your work. That is a lesson in marketing right there.
So I’m excusing myself for having spent some time analysing the data. I have a friend and author who is also an API whizz, who has acquired data for me on sales (estimated), ranking (provided by the retailer) and title. I have applied various analyses to this, by genre.
Yes, I am kidding.
But yes, I did the analysis. The best selling titles in Heroic Fantasy had 2, 3 or 4 words (I counted groups of digits as single words, so the title 1, 2, 3! is three words, but the title 123! is one word.).
I’m not giving the absolute numbers because the difference was not statistically significant.
It will come as no surprise that books with no words in the title had no sales. Consequently the Bell Curve was asymmetrical, but sales don’t fall sharply until the number of words exceed 12. I was quite surprised by that.
One word titles sold less well than 6 and 8 word titles but better than 7, 9 and more.
Reason suggests that this analysis has some sense to it, because it uses a measure that is only quantitative (objective). You could conduct this analysis, for instance, independent of language.
No, I am not kidding.
But yes, I’ve done the analysis, and so have others. I particularly like this analysis by Tor.com. My analysis used a much bigger sample than theirs, but the results were the same.
I’ve also done the same analysis on SF books, and this time I included subtitles and series name if it appeared on the cover, because I’m also interested in how redundant all that (Book 4 in the Arch Ark Arc series) is… I excluded the number that the book was in the series, so the only numbers are ones that appear in the main title.
Here are the top 15 (excluding the, in, and, of):
If you keep the subtitles and series names in there, then the most common words are Chronicle, Book, Novel and Series. I only point that out, because in SF, everyone seems to like to write a chronicle. I’ve written one.
For the next time I decide to do some unnecessary and fruitless statistical analysis on book titles, can we agree on a couple of things?
I get this is one of a series, or one in a world, or a universe you created, but when ‘Berth of Darkness’ is book one of the Dark Universe series, don’t call it:
The Dark Universe Book One: Berth of Darkness,
Berth of Darkness (Dark Universe Book One)
This is partly because it would make the statistical analysis easier. But spare a thought for the reader, too – and think like a salesman.
Readers who like your Dark Universe series will want to read all the books. Once there are 5 or more books in the series, they want to be able to identify as easily as possible which ones they don’t have.
They’ll search Amazon for “Dark Universe” and scan down the list. If the first words of every item are “Dark Universe” they have to read the whole title of every book to see if they’ve got it, whereas if it starts with the book title, they can tell at a glance.
The Dark Universe Book Two: The Dark is Everywhere
The Dark Universe Book Three: Dark Truth
The Dark Universe Book Four: A Lie in the Dark
The Dark Universe Book Five: Penury of Light
The Dark Universe Book Six: Dark Messiah
Berth of Darkness (Dark Universe Book One)
The Dark is Everywhere (Dark Universe Book Two)
Dark Truth (Dark Universe Book Three)
A Lie in the Dark (Dark Universe Book Four)
Penury of Light (Dark Universe Book Five)
Dark Messiah (Dark Universe Book Six)
(Please note this series does not exist. At least, I sincerely hope it doesn’t. In fact, if anyone reading this specializes in making horrible clichés a reality, you’re welcome to it.)
If you have to do the colon thing – and please don’t – can everyone please agree on a format? Either:
The Dark Universe: Berth of Darkness
Berth of Darkness: A Tale from The Dark Universe
Obviously, the latter is better.
Please (Arthur C. Clarke I’m looking at you here) don’t put a comma in the actual title.
Not only is this cliché mindblowingly pretentious, it’s also utterly pointless. Yes I know someone did this recently and got some big indy success.
I guarantee they did it out of insecurity coupled with the desire to evoke echoes of certain whimsical or experimental writers of the mid twentieth century.
Your book is going to be in the fiction section unless you’re unlucky or have so little tech savvy that you can’t get over the very low bar that Amazon wisely set in the KDP interface.
Anyone buying your book from the bookstore is going to find it in the fiction section and I promise you, bookstore owners and librarians will put it in the right place. They check.
This excellent Guardian Article has a choice remark to make about this strange practice.
There were 14 of these in my sample of the top selling 1000 SF titles, which isn’t as bad as I was expecting.
In Fantasy, if you analyze the titles including the series names, you get much the same outcome as you do with SF, but Book, Series and Chronicle are joined by the equally inevitable Saga and the ubiquitous Trilogy.
In Fantasy, the first word after these series words is, predictably, Dragon, closely followed by War and Blood.
You fantasy writers should be duly embarrassed by the fact that the next two most common words appearing in titles and subtitles in Fantasy are: Novel and Fantasy.
My raw data takes many thousands of the top ranked titles. So arguably, what I’ve been doing is analysing the words that appear in the most successful books.
The conclusion seems to be that the most successful book in either SF or Fantasy is a series. In SF, the first book in the series would be called:
End of the World (Alien Time War book 1)
… and in Fantasy:
Dragon Blood War: A Fantasy Novel (Book One in the Series Trilogy Chronicles Saga)
What’s the main thing you notice about those titles?
They look like everything else on those genre lists.
The result of applying a big data analysis to successful book titles to try to work out the rules for writing a successful title is a title that fails to stand out from the others.
And this is exactly what I’m expecting. I’m expecting it because I’ve approached this problem from two directions.
As a literary editor, I understand what creativity is, and how it works, and I’m aiming to show you exactly what creativity is, through this slightly silly exercise of word counting. Counting word use enables you to see what is happening. Specifically, it reveals trends.
Trends are what they sound like they are. In Fantasy, at the moment, everyone is still writing about Dragon Wars. And Blood. In SF everyone is still writing about Time Wars. And Aliens.
Now I don’t much expect the subject matter to evolve anytime soon. Those are proven favorites among readers. But it does look like there’s a trend for tediously and unimaginatively titling books as if they’re tins of paint, or baby food.
Dragon War is a book about war, with dragons. So much so obvious. Dark Magic Mage King is about… well, you get the picture.
If you look at the title trends in new books that are selling well, the titles all seem pretty samey.
However, look at the titles of the breakout hits of the last 100 years, there are quite a few oddities.
In #6 of Scott Berkun’s excellent little summary of the problem, he lists a few of them, and makes the best possible point about them: the titles we remember are the ones that are titles of good books… but also, are easy to remember, because the don’t have what the Guardian article calls the Samey Virus.
On the other hand, a few of the great books of the twentieth century have truly dull, samey titles that hardly set them apart from the pulp they either rose above or were already set above: Sons and Lovers, or A Passage to India.
But those were authors who didn’t worry too much about their titles because they knew the content was worth reading. And by “knew” I mean they knew it was true objectively, because they possessed the necessary literary education and experience to be able to judge.
The samey titles we see today are all coming from the wrong place. That place is: this is the sort of title that sells well, so this is the sort of title I should use.
But think about that for a moment, and go and read this article on the BBC News Website.
If it was really possible to work out how to make a killer title by analysing the titles of successful books using statistical techniques to develop algorithms with which to create new titles, then sooner or later, AI would be able to write novels.
And it will be able to, but not creative ones, for exactly the same reason that getting your title from the wrong place will mean your title is indistinguishable from thousands of others: it is inferring and then applying rules.
The result is therefore a kind of average, like a face average: it’s bland and anonymous; it could be anyone’s.
Creativity is the capacity to break the trend. To produce something that fulfills all the other needs of a title, which according to Scott Berkun (and I don’t disagree with him) are:
To be able to achieve all this, and not sound like every other book in your genre, requires creativity. Because only creativity can invent an alternative way to achieve these goals; a way that isn’t by the numbers; a way that isn’t based on satisfying criteria.
You can do this because you’re a human being, and you use a brain. Brains are messy. Brains use association. But crucially every part of your brain is used, in myriad different configurations, for many different purposes.
Which means you can make connections between ideas in a way that an algorithm, or even an AI (that isn’t based on some sort of highly plastic chaotic network), just can’t. You can make irrational associations; find, and indeed force connexions between ideas that really ought not to be there.
A lot of people think of creativity a something “going wrong” in the brain but actually, it’s an example of the brain’s necessary disorder going exactly as specified. Above all else, the brain is a shortcut machine.
Biological thought is embarrassingly slow, and the brain is an expert in short circuiting itself in order to save on resources.
So what is creativity? Creativity is thinking in leaps. Linking ideas that are not habitually linked.
If you insist. Here’s an example of a creative SF title:
The Time/Cost/Quality War
… and here goes with Fantasy:
The Accountant’s Apprentice
In these two examples, I’m satirizing the current trends to make a point. You will remember these – especially if their associations resonate with you.
But actually, if I had written a book in either of these genres, I’d look for something a little more creative. A title that evokes the genre but in no way suggests what the content might be. So for SF:
The Only Pace
… and for fantasy:
Olive Token in the Pod
This is a phrase invented by Terry Pratchett as part of the culture around the story of one of his earliest books, The Dark Side of the Sun.
With a typical mixture of mastery, wit, and a nod to popular culture, this title screams SF, but is incongruous, even impossible, in its meaning.
One of the features of the story is frequent mention of fictional philosopher and explorer Charles Sub-Lunar, and this phrase, the poet and the mad computer describes him. But it also describes the process of creative thinking.
The mad computer because your brain’s very disorder is what makes it such an effective thinking apparatus, and the poet because it is through mastery of meaning that you can both decipher and create great writing.
Here’s my advice for creating a great title.
Science is about finding stuff out, by applying a suitable means to a problem. Statistical analysis is not a suitable means of finding out what makes a successful title.
Although it can reveal patterns in current successful titles, it can’t tell what direction trends in titles is going to take. It can also reveal patterns in culture, thinking, behavior, even desire.
But empirical analysis is not a suitable means for understanding how a creative literary process works. If science means anything, it means knowing when to use analysis, and when not to.
The suitable means for creating a great title is the same as for creating a great story. It is the practical application of creative thinking.
And since creativity relies on the brain’s capacity to exploit disorder to find more efficient solutions, the results of creative thinking differ from person to person. In short:
Use your creativity and your titles will be unique to your way of thinking. Use analysis and your titles will be the same as everyone else’s.
That can’t be a good thing.