I understand why they use test audiences. And it can have value. But if you hire someone with an off-kilter approach, and pick a random sample of people to review your film, don't be shocked if the test audience doesn't like or understand it.
The problem is that there is only so much information that the studios can plug into their predictive models here, and you can't necessarily get all the right people in to give you the right kind of input in situations like test screenings. And frankly, i'm not sure that there is much of a real science to it from what I've seen. It seems more pseudo-science, where people who speak with authority try to convince execs with money that they know what they're doing. Evilface's example is a good one. It's not a bad idea to get people who wouldn't typically like a film into a test screening. But that should be selective and intentional (e.g., what can we do to minimize the boyfriend's boredom while not pissing off the girlfriend with this romcom?), and you shouldn't have too many of those folks there, because chances are they're never going to want to see a film like that unless they are forced to.
As for Black, if this movie fails spectacularly, so be it. But I'm a bit tired of predictable, generic movies. I'll happily shell out money on something that takes a risk and fails in the geek genre nowadays over one more Star Wars or Avengers wannabe. And if you try to do that with Predator, no one's going to see it anyway.
