Science Discovers Secret To Writing a Best Seller

[Edit, Since I discovered the paper and wrote this, I noticed that the interwebs have exploded with people linking to, and commenting on this story. Since this isn't actual news to most peoples at this point, please enjoy the picture I inserted.]

Seriously, that isn't link bait. It's an actual thing. I found this paper over the past weekend wherein the geniuses at Stonybrook University have finally cracked the code to writing a best seller.

They poured over the texts of thousands and thousands of books and looked for commonalities, not in plotting, or structure, but in prose. They then compared to a long list of underperforming books to see what they did differently, again, speaking strictly of prose.

And they found out a lot of stuff. Namely, when comparing just the prose of all those books, they can tell with an 84% reliability* of whether or not it was a best seller by just looking at the prose itself.

Now, since I wasted spent quite a bit of time looking at the paper, I think I can distill it down to a few simple points. However, I do encourage you to sit down and read through it if you can, it's written like most academic papers are, which is to say, dryly. But the information it contains is pretty interesting. And if you go over the first few pages carefully, you can probably see some flaws in their sampling methods.

This has nothing to do with my post. I just had this doodle, took a picture with my phone, hit the 'drama' button and... ta da!

Despite that, I think it's brilliant topic. I really hope I find more like it.

So, anyway, what are some takeaways?

1) Quit with the damned adverbs: That one might be obvious to any that spend much time reading those writing tips that are everywhere online. Turns out, science agrees. The folks that wrote the paper don't editorialize on why any of their findings might be indicators, but I know that I tend to not enjoy adverbs that much at all (especially in dialog tags).  But that's just me, a reader. As a writer I know that we have to use them sometimes, but I suppose it's like adding salt to your cake mix. A little can make it amazing, a lot can make it gross.

2) Write something that scores higher on the Flesch Index: What's that, you ask? Simply put, it's how readable the prose is, for example, the paper I linked to in the opening paragraph probably has a very low score on the Flesch Index, you want a higher score here. That isn't to say that people are dumb either, it can just mean that you write confusing sentences. So, you don't have to dumb it down, just don't make it harder to read than it should be.

3) Don't use Verbs: Wait, what? What the hell are they talking about? You know, after all the defending I do for Science, it irks me they go and tell me to quit using verbs. I mean, it's hard to send my hero on an epic quest to save the world when I can't have her, you know, actually do stuff. Which, of course, leads me to my final point.

4) Don't pay too much attention to that paper: I do think it's fun to look at, and I appreciate the effort to qualitatively show correlations in what makes works successful and what doesn't. Even if it is mostly unusable information. I hope they keep going. But at the same time, pointing out that successful books use words like 'and' isn't really the deep sort of insight I was hoping to get when I started reading.

*Not really, but it's hard to make a bold statement if I have to follow it up with a bunch of qualifiers. So, just go with the 84 number, okay?