Chaos theory explodes Hollywood hype

You can imagine the conversations in Hollywood right now. "My idea is The English Patient meets Star Wars... I'm thinking Demi Moore. She's married to an alien, who crashes his spaceship..." Or, "Independence Day meets Trainspotting... they're all high on drugs when the invasion happens, but the aliens can't understand their accents... it's a sure hit if we can get Tom Cruise."

Doomed to failure, of course. Or are they? The trouble, as any film executive who's honest will admit, is that no one can tell. As Oscar-winning screenwriter William Goldman famously said, nobody knows anything.

A big-name star is no guarantee (think of Last Action Hero with Arnold Schwarzenegger) and even a fistful of Oscars doesn't necessarily mean money in the bank. Trying to forecast hits is like trying to predict the weather.

And that, says one scientist, is not just idle imagery but absolutely true, because both are ruled by the chaos theory. Art de Vany, an economist at the University of California in Irvine, says that filmgoers are like particles in gas. You can't predict what they're going to do specifically. Science fiction? Natural disasters? "Genre," Dr de Vany says, "has no predictive value."

With David Walls, also an economist, Dr de Vany has made an econometric study of the film industry, bringing new thinking to a market that for decades has been happy to operate on "feel". (Consequently it's missed some beauties: 20th Century Fox refused to finance The English Patient two years ago, and would have featured Demi Moore instead of Kristin Scott Thomas. Hit or flop?)

The results make unsettling reading for highly paid studio executives.By looking at the revenues and lifespans of 300 films released over an 18- month period, the economists showed that the only guaranteed indicator of how well a film will do in one week is how well it did the week before.

The statistics are stark. On average in the US, less than one in six films runs for longer than two weeks. Only one in 20 survives longer than 15 weeks. Four films out of the 300 pictures studied earned 80 per cent of the total revenue. That leaves just 20 per cent of the total for 296 films - which, in turn, means big losses.

To analyse precisely what was happening, Art de Vany and David Walls had to turn to a complex piece of physics devised earlier this century by Albert Einstein and Satyendra Bose. The equations are used to model the behaviour of large groups obeying simple rules independently - the essence of chaos. It can be gas particles in random motion, ping-pong balls being showered onto buckets or (it turns out) people deciding what to watch on Friday night.

"Audiences tend to behave over the course of the film's run like the particles falling into urns in a statistical physics model," says Dr de Vany. "In the Bose-Einstein process it is equally likely that the particles (audience) will fall into a few urns (movies) as it is for them to be distributed in any other way. Ultimately, the probability that a new viewer will go to a particular film depends positively on the number who have already seen it."

Everything depends on the first week. "Stars and big budgets guarantee only that a film's run will begin on many screens. From there on a film's run is like a parachute jump - if the film doesn't 'open', it's dead."

One element is crucial for success. The opening of a film creates an "information cascade" (what you or I would call word of mouth). That word of mouth can pack a place out - or leave it empty. It is a process which begins with film critics, but spreads as soon as the film opens.

Ask yourself: have you met anyone with a good word to say about Sly Stallone's latest, Daylight? Have you met anyone with a bad word to say about The English Patient? And, given the choice, which would you go to see?

The problem for the studio executives is that it is beyond their control. The bigger the hype, the bigger the potential success - or flop. Think of Independence Day or Mission: Impossible. Then, think of Waterworld.

So it doesn't matter how much the studios spend on a film (such as Titanic, thought to be heading for a $200m box-office bloodbath later this year) or on its marketing. We may just be Bose-Einstein particles - but we're opinionated particles.

An interview with Art de Vany appears in the current issue of 'The New Yorker'