On Writing a BookAbout writing a book
Do you want to make a bestseller? You use an accounting algorithms that uses cash.
It is the million-dollar issue that has been under the minds of editors and creators for decades: What makes a best-seller? Trying to type could certainly be worthwhile - the world's highest paying writer, JK Rowling, made $95 million (£70 million) last year, and the world's top 10 most highly paying creators have made more than $310 million between them, according to Forbes.
Few playwrights will see that kind of thing. Authorhors' Licensing and Collecting Society says that since 2005 professionals have seen a 29% fall in revenues in real life, with the top 5% making 42% of the UK authors' earnings. A book's advertising budgets are usually linked to advances, with prominent contributors increasing more than their equity in publishers' advertising spending, regardless of the standard of their work.
Finding a book to make a living with is not simple. Best-selling writers such as James Patterson, Danielle Steel and Stieg Larsson have repeatedly had texts turned down, while self-publications such as Fifty Shades of Grey have surprised the book world. "Everybody in publishers is always looking for the next bestseller," says Jodie Archer - and she should know.
"At the same moment, I wanted to know why the whole wide globe seemed to study the same book. How does the public opinion view the type of book the reader wants to see? How do novelists write this unfamiliar business hit in their stories? "Instead of asking writer and reader, Archer chose to ask a computer.
"Whether a book is released as literature fairy tale, romanticism, sci-fi, criminality or any other kind of book, there are some hidden characteristics of the bestselling cathedral in scripts and these are recognizable by a computer algorithm," she says. She and Matthew Jockers, an authority on text-miners, have developed an algorithms that can tell if a script reaches the New York Times best-seller lists with 80% precision.
For some writers such as Rowling, Patterson or EL James of Fifty Shades, the algorithms were more than 90% certain that their scripts would be best sellers. "I was bewildered when we first ran it," says Archer. Fortunately, she and Jockers have chosen to exchange what they have gained from their algorithms in their book The Bestseller Code 2016.
"This shows how literature, as a computer reads it, has a big impact on how humans will react," says Archer. Most of the winning writers know that, of course - and other insights seem just as evident. Thus, for example, the algorithms found that bestseller reading professionals like short phrases, voice-controlled stories and less learned words than reading fictional literature.
She also foretold that writers who had worked in the journalistic field had the best opportunity to make a best-selling comeback. "This type of workout will help you type for a successful marketing environment and means that your styles are often available and colloquial," says Archer. It also recognized the issues debated in a book by evaluating the text for often related word groups.
She came to the conclusion that there were 500 themes (such as "Love" or "Lawyers and Law") that were important for classifying a book as a best-seller, and various combination of these themes gave the reader a feeling for the topic of the book. Then Archer and Jockers found out that accomplished writers dedicated 30% of their works to just one or two Themes.
"Lesser skilled writer have a tendency to incorporate too many themes into their first designs, so their works seem to have no focus," says Archer. Arker says many people have asked them how they can correct their scripts. She and Jockers set up an agent later this year to allow the author to use their algorithms.
They give authors a review of how the algorithms have evaluated the probability that their manuscripts will be on the NYT best-seller listing for an unpublished charge, and give them enough free rein to find out how they can increase their odds.