The algorithm that can pick a bestseller

By Toby Manhire In Books, The Internaut

Print Share

The march of the algorithms continues. Now, apparently, they can pick which books will prosper simply by scanning the text.

Danielle Wiener-Bronner explains at The Wire:

Computer scientists have developed an algorithm they say can predict the commercial viability of a book, with an 84% success rate, based solely on the style in which the book is written.

 The Stony Brook University academics have identified the “predictive power of statistical stylometry”, they say, having studied the correlation between a book’s linguistic style and its sales.

The first, general finding: good writing is pretty much irrelevant.

More specifically, in Wiener-Bronner’s summary, for example:

Less successful [adventure] books rely on verbs that are explicitly descriptive of actions and emotions (e.g., “wanted”, “took”, “promised”, “cried”, “cheered” etc.) while more successful books favor verbs that describe thought-processing (e.g., “recognized”, “remembered”), and verbs that serve the purpose of quotes and reports (e.g. “say”).

It’s tempting to believe, mind you, that the authors of the study are pulling our legs. Consider their own prose style.

We conjecture that the conceptual complexity of highly successful literary work might require syntactic complexity that goes against readability.

In other algorithms-taking-over-the-world news, here’s one that lets you bet on sport without ever watching it.

See also: Tick here for book delivery by drone

Follow the Listener on Twitter or Facebook.

Post a Comment

You must be to post a comment.

Switch to our mobile site