"Although The Devil's Teeth sold just 36,000 copies in hardcover, according to Nielsen BookScan, which tracks about 70 percent of book sales, it was a New York Times nonfiction hardcover best seller," Times book reporter Motoko Rich wrote last week in a weird article that had book editors around town scratching their heads. "Although"? "Just"? Many Times bestsellers have similar and, sometimes, lower Bookscan numbers. Is it possible that the Times, arbiter of the most influential bestseller list in the country, doesn't really have a set or sensical definition of what constitutes a "best seller"? An article in yesterday's Business section reveals: yes, that is possible. Likely, even!

An editor's note informs us that "the editor of the Sunday Business section is under contract to Random House and did not edit this article," which maybe explains sentences like "Based on those figures and some analysis—about the popularity of the genre, the likely audience, the possible newsworthiness of the topic of the economy—they work up profit and loss projections." It kind of seems like no one edited this puppy.

For starters, what kind of article sets out to explicate a mystery—in this case, the mystery of why some books sell a lot of copies and others sell very few, and how indirectly these sales typically correlate with the size of authors' advances and publishers' predictions—without even attempting to define the term that the article is purportedly about. Is "making a best seller" really so mysterious that no one can even begin to know what a "best seller" is?

It isn't until almost three quarters of the way through the article that we get anything resembling answers. "There are two ways for a book to become a best seller. One is to make it on to a best-seller list by selling many copies in a week. Other books sell steadily over months and years, eventually outselling many official best sellers," reporter Shira Boss informs us. (Say it with us, Shira: "backlist.") So! There are best sellers on "best-seller" lists and then there are 'unofficial' best sellers? How many copies does a book have to sell to be considered in any way "best?"

Maybe the Times is being cagey because they don't want to publicize how ultimately non-authoritative their bestseller list's rankings really are. The methodology behind the list is a "trade secret," based not on plain vanilla sales figures, as this article seems to imply, but on sales figures from "a selected sample of independent and chain bookstores."

The article's ostensible point—that publishers base their acquisitions too much on authors' sales track records and too little on anything resembling reader feedback—is a good one, of course. But how on earth are publishers or readers meant to get a sense of the book sales landscape if everyone is confused about what kind of numbers they're meant to be chasing?

Is poor Curtis Sittenfeld a failure because her second novel sold "only" 42,000 copies? Is Charles Frazier a failure because his follow-up to Cold Mountain, in spite of its 8 million advance, "only" sold 240,000? Or are the real failures the people who write about this stuff and can't seem to wrap their heads around the seemingly illogical but ultimately sort of inevitable way that book publishing actually works—that is to say, differently than every other industry, especially those in which people actually make lots of money?

The Greatest Mystery: Making A Best Seller [NYT]
NYT Bestseller List [Wikipedia]