Paul Doiron here—

The Web is a wacky and wonderful thing. Spend enough time browsing, and you’re bound to come across something weird. A while back, I stumbled across Lulu.com’s Title Scorer, an app that uses algorithms to determine whether the proposed title of your novel will increase or decrease the chances of it being a national bestseller. Here is the dare it presents the hapless author who stumbles upon the site:

PUT YOUR TITLE TO THE TEST

Want to know if you’ve got a killer title for your novel? Now, for the first time in literary history, you can put your title to the scientific test and find out whether it has what it takes for bestseller success.

Are you brave enough to put your title to the test?

Who was this Web site to question my bravery? I typed in the title of my new book and received this surprising feedback:

The title bad little falls has a 69.0% chance of being a bestselling title!

Nearly 70 percent! That was almost a sure thing. I was tempted to call my literary agent on the spot, and say, “See, I knew we were right to push for Bad Little Falls! We’re practically guaranteed to hit the List with this one. Heck, we should go to Vegas with the dice rolling our way like this.”

Flush with excitement, I read Lulu’s explanation of how the Titlescorer works:

The Lulu Titlescorer has been developed exclusively for Lulu by statisticians who studied the titles of 50 years’ worth of top bestsellers and identified which title attributes separated the bestsellers from the rest.

We commissioned a research team to analyse the title of every novel to have topped the hardback fiction section of the New York Times Bestseller List during the half-century from 1955 to 2004 and then compare them with the titles of a control group of less successful novels by the same authors.

The team, lead by British statistician Dr. Atai Winkler, then used the data gathered from a total of some 700 titles to create this “Lulu Titlescorer” a program able to predict the chances that any given title would produce a New York Times No. 1 bestseller.

I was feeling very cocky, I have to tell you. Titles are difficult for me, and receiving the blessing of Lulu’s statistically super-powered, all-knowing computer sent a boost of confidence shooting through my circulatory system.

Now I don’t mean to boast, but my previous book had been a bestseller, albeit not on the Times list. It had landed, for a single week, on the esteemed American Booksellers Association Indie Bestseller List, right along with Ann Patchett and Tom Clancy. Granted, it was #15 of fifteen books, but still, I can now call myself a bestselling author for the rest of my life without my nose doing the Pinocchio thing.

Had Lulu predicted my good fortune, I wondered. I typed in the name of my last book and received this reply:

The title trespasser has a 35.9% chance of being a bestselling title!

So I had gotten lucky with that one! Little better than a one in three chance, and I had scored. Perhaps if I had known about Lulu Titlescorer sooner I could have experimented with better options and landed at #14 or even #13.  I might even have hung on the list for two weeks.

In light of this information, I began to reconsider Bad Little Falls, wondering what I could do to improve my odds even more. Alas, no tweaks I tried seemed to elevate my chances above that 69 percent threshold.

So what was the secret? Obvious Ann Patchett had figured this game out. Her book State of Wonder has floated on a magnolia cloud atop the bestseller list when Trespasser hit bookstores. I tried it in the Titlescorer. According to the computer the name had only a 22.9 percent chance of being a smash.

Hmm, I thought. What about Clancy? He was a man who knew from titles. But Against All Enemies received only a 26.3 percent chance when I typed it in.

I was beginning to lose faith in Lulu. What if she wasn’t infallible? With my authorial future riding on Bad Little Falls, I needed some reassurance. How about Stephen King! He had given his most recent blockbuster an unconventional title—the date of JFK’s assassination. It was a risky gamble, even for such an established author. Who gives a novel a date for a title? No doubt, his publishers had had cold feet. Surely Lulu, however, with her awesome predictive powers and cold analytical computer brain, would have foreseen the genius of his gamble.

Or maybe not:

The title 11/22/63 has a 14.6% chance of being a bestselling title!

It was only then that I read the disclaimer:

This is not an exact science. Far from it. In fact, Dr. Winkler advises that the Lulu Titlescorer should, in practice, always be combined with use of your own low-tech judgement.

This is because, for all the work that went it, the Lulu Titlescorer is capable of giving high scores to titles that most of us would rate as weird, if not terrible. Meanwhile, of course, it also gives low scores to the titles of novels (e.g. The Da Vinci Code) which, in fact, topped the New York Times bestseller list for long periods.

Thanks a lot, Dr. Winkler. So in other words, I’m on my own with Bad Little Falls. I felt betrayed by Lulu, the faithless vamp.

I wonder if it’s too late to cancel those tickets to Vegas.