Remember that old joke about the new-to-the-industry suit who asks his publishing colleagues why they don’t publish only bestsellers? Well, New Yorker business columnist James Surowiecki suggests – a little optimistically, perhaps – that we may yet reach the days when publishers can accurately determine in advance which books will sell.

That’s because we’re goint to see increasingly sophisticated platforms that will survey consumers on their tastes in a widespread way, producing more reliable results than, say, focus groups. (Which, as Surowiecki notes, “tend to be dominated by the loudest and most opinionated people, to be driven toward consensus decision, and to discourage disagreement, making them of limited usefulness.”)

The days of product-tested books still seem to be a long way off, though – Surowiecki offers only a couple nascent examples of “prediction market” projects and says, “It’ll take a while to work out the kinks.” And Quillblog’s admittedly knee-jerk reaction is still one of skepticism. We already know, after all, that the public likes suspense novels about 2,000-year-old Catholic conspiracies and fantasy series about young wizards, but that isn’t really making publishers’ jobs any easier, much as some of them clearly wish otherwise.