[Worldwalker]

According to the publishers, everyone who downloads a 3,000-book torrent would have bought every single one of those 3,000 books at full retail price if they hadn't gotten them off the darknet. Because, y'know, people routinely spend $45,000 a year for books -- even people who only make $20,000 a year.

This is the same logic the BSA uses to claim losses due to software piracy -- the idea that a teenage warez d00d would have bought Photoshop at its full $1k+ price to put stupid captions on his Facebook pictures if he couldn't get a cracked copy.

It's not a lost sale if it was never going to be a sale to begin with. That hypothetical teenage warez d00d would have used some online graphics editor (they're getting remarkably good) or even Paint if it came down to it. The idea of the copying of software or content being "theft" comes, as near as I can tell, from the idea that someone is in effect stealing back the money that they should have paid for the stuff. But if they were never going to pay it to begin with, is there anything to be stolen?

A while ago, I looked at the listing for the contents of a small bundle of books. I don't remember the exact numbers -- I think I posted them in another thread a month or two back -- but it came out to a total of about 3% of the books were ones I would actually want to read. And out of those, nearly half were books I already had on dead trees. Let's be generous and call it 2%. The collection in question was focused on one of my favorite genres, and my reading tastes are quite wide-ranging, so it's reasonable to assume that a similar number of books -- about 2% -- out of any given bundle of books on the darknet would appeal to any given reader, and would be books they didn't already own. Less if they're one of those people who never reads a book twice, since about half the books that appealed to me were books I'd already read at some point in my life. So out of 1,000 books in a torrent, Joe Schmoe is going to be interested in 20 ... and probably have read 10 of those already. So instead of $45,000 worth of lost sales, we're looking at $149.50 in lost sales, maximum. The other $43,850.50 does not represent lost sales because they're sales that would never have been made; the files probably didn't even survive Joe's first scan through the newly-unzipped collection.

And the odds are, if Joe could have bought those books hassle-free, DRM-free, at reasonable prices from his favorite ebook store, he wouldn't have taken the time and effort to find them on the darknet. He would have been happy to have a nicely-formatted, illustrated, legitimate book instead of an illicit book of dubious OCR quality. If they'd been on Smashwords for $2.99, he'd probably have bought them, and the author would have gotten more profit than if he'd bought a mass market paperback.

So, in short, there are no legitimate, reliable numbers on lost sales. Anyone who claims to have them is pulling them out of a very dark and smelly place. Any shift in sales figures that might be due to illicit downloads is so entangled in so many other factors that it cannot be distinguished. For instance, I've spent less on books since I got my ebook reader. Probably a lot less. The economy has a lot to do with this. Yet I'm reading a lot more, partly because the aforementioned economic downturn has made alternate forms of entertainment relatively more expensive. In the past year, I've probably read a hundred items (novels or short fiction) from public domain sources and free sources such as the Baen Free Library, and except for one or two MM paperbacks a month, all of my spending for recreational reading has gone to Baen, Smashwords, BVC, directly to authors, occasionally to other publishers and outlets, and so on. The publishers are making less off of me, definitely -- but authors, collectively, are making more. No illicit downloads involved.

Roll the dice. Whatever you come up with will be as reliable as any other numbers you find.