A formula typically only works as long as it doesn't feel too much like a formula. Apparently this summer's movies feel formulaic:

The summer movies of 2003 have thus far made roughly 2 percent less at the box office than last summer's fare, and actual attendance has fallen an even more striking 6 percent - a trend that did not spare "Charlie's Angels: Full Throttle," a classic example of the summer movie sequel that opened last weekend. Aimed at young audiences supposedly more interested in sizzle and flash than in such old-fashioned notions as plot and character, the film made only $37.6 million in its first three days.

That may sound like a lot of money, but it's less than the original made in its debut weekend. So in a business where summer blockbusters can open north of $60 million and still be considered disappointments, this was something of a disaster.

"The `Charlie's Angels' case is a fascinating one, because it had all the earmarks of being a phenomenal success," said David Davis, an entertainment analyst for Houlihan, Lokey, Howard and Zukin, an investment banking firm. "A very expensive marketing spend, all of the stars doing publicity - it had everything going for it. I don't know, maybe after so many of these kind of movies so many weekends in a row, it was just one weekend too many."

All of this comes as a bit of a shock to Hollywood executives who, relying on rising ticket prices and a steady diet of sure-thing franchises, had become addicted to breaking box-office records without breaking a sweat. But after a decade of summer movie seasons featuring a steadily increasing number of sequels, prequels, remakes, special-effects extravaganzas and other manifestations of lowest-common-denominator filmmaking, some signs are appearing that the long tyranny of the summer franchise flick may at last be waning. [NY Times]

This follows in any art or entertainment medium: the new Harry Potter is breaking records not only because of pent up demand, but word of mouth and reviews have been very positive, almost universally calling it better than the last one, which did have the slight whiff of formula per general consensus. When the formular doesn't work, then everyone scrambles to come up with a new formulat that doesn't feel too much like a formula and the process begins again.