Seth Godin has got the right idea:
... the rocket scientists are busy promising Hollywood that they can run the numbers on a script and figure out how to change it to make it more likely to sell. Add a sidekick to that superhero, perhaps, or have that demon be summoned instead of whatever it is that unsummoned demons do...
This rearview window analysis is anathema to the creative breakthrough that we call art. No amount of digital focus group research could figure out that we wanted Memento or The Matrix or Amour. Worse, it's based on the flawed assumption that the past is like the future, that correlation and causation are related. By that analysis, every Supreme Court chief justice, US president and New York City police chief is going to be a man. Forever more.
Emph. mine. The whole problem with making — as opposed to just marketing — art of any kind on a commercial scale is that it forces you to be a groupthinker, not a maverick. You're stuck referencing only what exists, instead of dreaming of what could be, because otherwise The Numbers don't add up.
Now, people — especially people with tons of money at risk — have good reason to be skeptical of the idea that only a lone weirdo can save us. Sometimes those lone-weirdo moves turn out to be just that: not only unmarketable, but unwatchable and uninteresting by any measure, no matter how we run The Numbers. But without the possibility of acting as if the future doesn't have to just consist of yet more manqués of present genres, then there really is no future.
Don't take any of this as a sign that I'm all for tossing the Nate Silvers of the world out on their spreadsheets. Number-crunching deserves to be put to good use — better use, certainly, than keeping us landlocked in Sequelistan.