Epagogix

email this story
read comments (16)
post a comment

There will always be outfits offering formulas and stratagems that will cut down on the risk factors in movie-making. A New Yorker article by Malcolm Gladwell is focusing on a a new permutation of this -- a British- based company called Epagogix. If every producer and creative executive in this town were to embrace Epagogix, the effects upon the withered soul of Hollywood would be nothing short of demonic. And yet, to be perfectly honest, if I were running a studio I would probably take a look at it.

The principals are Dick Copaken, Nick Meaney and Sean Verity. They're basically offering a "neural network" program that, in a Hollywood sense, counts cards at the casino. Their site says that Epagogix "works confidentially with senior management of major film studios to assist in identifying and developing scripts, and in transforming scripts with low box-office revenue potential into properties that can be profitably produced and distributed."

"If you were developing a $75-million buddy picture for Bruce Willis and Colin Farrell, Epagogix says, it can tell you, based on past experience, what that script's particular combination of narrative elements can be expected to make at the box office," writes Gladwell. "If the formula says it's a $50-million script, you pull the plug."

The neural network, says Meaney, "can sometimes go on for hours. If you look at the computer, you see lots of flashing numbers in a gigantic grid. It's like 'The Matrix.' There are a lot of computations. The guy is there, the whole time, looking at it. It eventually stops flashing, and it tells us what it thinks the American box-office will be. A number comes out."

"The way the neural network thinks is not that different from the way a Hollywood executive thinks: if you pitch a movie to a studio, the executive uses an ad-hoc algorithm--perfected through years of trial and error--to put a value on all the components in the story. Neural networks, though, can handle problems that have a great many variables, and they never play favorites -- which means (at least in theory) that as long as you can give the neural network the same range of information that a human decision-maker has, it ought to come out ahead."

Innarritu vs. Ariagga<< previous | next >>Innaritu to Pearson

Posted by Jeffrey Wells on October 26, 2006 at 12:27 PM

comment #1

Mark G. says ...

Whoever uses this thing is wrong for the job. You have to trust your gut instinct and looking back is certainly the wrong choice. That's part of the problem of what's wrong with today's movies. Instead of sensing, smelling, anticpitating what moviegoers would like to see two years from now, the dealmakers try to copy what's working today, not realizing that the movie will be outdated 18-24 month after the greenlight when it hits the screen.

Posted by Mark G. at October 26, 2006 1:26 PM

comment #2

Rich S. says ...

I'm more shocked that they haven't been using something like this already. Don't they have the first week's grosses plotted out long before the movie ever hits theaters?

I still think it comes back to the blockbuster mentality. No one knows when a picture is going to be a big hit or a big flop. But smaller pictures should be fairly easy to predict based largely on the past performance of the actors/writers/director, etc. Again, pictures can hit or flop, but those should be aberrations. And that should also be built into the equation.

Posted by Rich S. at October 26, 2006 1:48 PM

comment #3

jeffreywells says ...

Wells to Mark G: Exactly. Epagogix ignores -- has no choice but ignore -- that life is a moving train and nobody knows precisely what's coming 'round the bend,

Posted by jeffreywells at October 26, 2006 2:04 PM

comment #4

giantman says ...

Jeffrey just said what I was thinking, the past is dead and while that might (and that's a big might) be an interesting piece to the puzzle, it isn't the entire picture. I'd prefer to see Hollywood moving forward, this system would guarantee the same-old-same-old.

Posted by giantman at October 26, 2006 2:22 PM

comment #5

Craig Kennedy says ...

This one time I went a whole week without reading an article about movies that quoted William Goldman's "Nobody knows anything" and then leprechauns and unicorns came flying out of my butt.

This is a fascinating article, but the whole Epagogix thing is troubling. If I was a studio executive and my mission was to make movies with as broad of an audience appeal as possible, I'd be sold, but as a viewer I'm more interested in the unique films that fall into the nooks and crannies. I want to be challenged and surprised, not shown what has worked before.

On the other hand, if they could tailer the system to fit my tastes, similar to what Netflix does only more complicated...that would be interesting.

Posted by Craig Kennedy at October 26, 2006 2:43 PM

comment #6

JD says ...

There are literally thousands of people with an acute understanding of movies and yet the movie industry insists on hiring lawyers, MBA grads, and marketing people with little understanding of movies to laborioulsy reach conclusions that your average cinephile could reach with far greater ease, efficiency, and success. They're living under the illusion that movies are a science when all the evidence suggests otherwise.

The movie geeks that read and write Aint it Cool News could have told you a decade ago that directors like Peter Jackson and Sam Raimi were going to make enormously successful blockbusters (not because of past successes, but because of their talent), while the industry was still more confident hiring "proven" directors like Renny Harlin and Joel Schumacher to direct their tentpoles. Inevitably, the industry catches up with the fans, but why play catch-up when you can actually set and anticipate trends? All I'm saying is, there should be some kind of mandatory (and difficult) film comprehension test that every Hollywood executive must pass before taking a position of power. That's the only way Hollywood will ever regain the credibility and success it had back in the days of movie-loving moguls like Louis B. Mayer. Fortunately, their spirit (and sense of taste) seems to live on in indie execs like John Lesher and James Schamus.

Posted by JD at October 26, 2006 2:43 PM

comment #7

christian says ...

how utterly frightening. and predictable.

jd is exactly right. the executives are usually exactly wrong.

but this what happens when you let marketing scum from big colleges control the creative side.

thassright, i called you marketing people scum. and you are as you know it's your job to collectively rid the country and world of originality in favor of branding us like cattle.

marketing people have utter contempt for the consumer. i've worked with them and most are soulless drones with the morals of a GIRLS GONE WILD producer. watch tv ads today and the proof is there. oh, i'm sure some great cookie-cutter cross-pollinated synergistic films will come from this computer program.

Posted by christian at October 26, 2006 3:05 PM

comment #8

christian says ...

and that's why execs love directors from videos and commercials. they know how to be good li'l corporate drones. spare me the one or two exceptions.

Posted by christian at October 26, 2006 3:07 PM

comment #9

Sean says ...

"He didnÃ¢â‚¬â„¢t like Ã¢â‚¬Å“The Bodyguard.Ã¢â‚¬Â Even the title made him wince. He was the sort who liked Ã¢â‚¬Å“Yojimbo,Ã¢â‚¬Â after all."

If I thought he meant that as a joke, it would be pretty funny.

So, basically, these two guys are the screenplay equivalent of eHarmony? The thing the music guy is doing makes some sense, but these guys seem to be suggesting that only they can properly judge the script along the specific lines, and assign the scores. I don't like that.

Posted by Sean at October 26, 2006 6:10 PM

comment #10

Michael says ...

These guys sound pretty smart. But they're not smart enough to get the last scene in The Bodyguard? They pretty explicitly do not end up together.

Posted by Michael at October 27, 2006 6:52 AM

comment #11

Midwest Doug says ...

Neural networks have bit a hot thing with datamining in the past 10 years or so. One of the banes of neural nets is that they can 'overpredict' the data. In other words, they get very very good at predicting the data that's used to build the neural net, but they're poor at extrapolating out to new things (similar to Mark G's original point). From reading the Gladwell article, the guys at Epagogix have been careful about over-predicting their sample and have done well at predicting new things. However, they'll tell you that they cannot predict outliers, whether they be my Big Fat Greek Wedding or the Blair Witch Project or whatever.

The other part about Epagogix that's hard to take is that their calculations are based only on the screenplay. It doesn't care if Michael Mann or Michael Bay or McG or Brett Ratner or your buddy down the street is directing the movie, and it doesn't care if Cary Grant or Cary Elwes is the star. That's completely counter-intuitive, and I can't buy it completely. But it's their 'Moneyball' premise.

Still, I'd like to take the heretical stance that this could be good for movies. What kills studios is the big bomb. Epagogix could prevent the big bomb by making big investment pictures safer plays. Would they be great art, or anything? Perhaps not. But if studios are making money on their big movies, that gives them more leeway to gamble on low-investment pictures that might break out.

The newyorker.com has a video of a Gladwell speech on this topic. It's interesting, because Gladwell is an engaging speaker, and because he portrays Copaken as quite a movie buff. Copaken recognizes that small pictures he likes will not be favored by the neural nets; he's aware of the dilemma that his project is creating.

A studio could certainly use Epagogix and only make safe plays. But even in dollar-dominated Hollywood, I can't believe all studios would follow that path. They are trying to create art, and they employ artists. So yes, there would be noxious, homogenized product, but there always is. And studios would be more comfortable taking risks knowing that part of their revenue stream is certain.

Posted by Midwest Doug at October 27, 2006 7:01 AM

comment #12

Sean says ...

"And studios would be more comfortable taking risks knowing that part of their revenue stream is certain."

I thoroughly disagree. There's one *big* assumption in your post, Doug. It's this:
"But even in dollar-dominated Hollywood, I can't believe all studios would follow that path. They are trying to create art, and they employ artists."

The studios rarely, if at all, care about art, except in terms of how they can market it, and they're getting *further* from the art paradigm. Disney, for instance, is now run by a marketing guy.

What you're saying, essentially, is "a however-many-billion dollar company which is guaranteed almost a billion dollars a year [from having multiple big-hits] is going to risk a hundred million of that in order to maybe make as much as a hundred and fifty (in lieu of, not addition to, that hundred the spent on risky stuff). Stuff their computers tell them won't do well.

It doesn't make sense. The studios have repeatedly shown how uninterested they are in small movies, except that each studio takes one "pet" director and that becomes their smaller-movie guy.

The thing is, the way it's set up now, they're already choosing to throw 150-200 million at terrible ideas like 'Poseidon' rather than funding 3-5 smaller pseudo-indies, even if the smaller movies could each gross back their budget and more easily. They're just not interested in the smaller paydays, but they realize that a few of them can hit, and then they look brilliant, so they fund a few of them. The computer would hurt that.

And, on the other hand, you'd *never* have an interesting $200 million movie again. I know a lot of people here hated 'King Kong' or 'Superman', but they at least represented filmmakers giving in to their own excess (as opposed to Brett Ratner who flat out can't direct doing 'X3'). 'King Kong' was too long, but all the best stuff were the parts that could've been streamlined out. Gone will be the remotely artistic $200 million movie, because they can say "add 10% more love story, a few jokes, and if you could make Jimmy Olsen black, it would add ten million bucks out the door, easy."

The studios primary goal is to make money. If they had a literal formula for how to do that, they won't feel they *need* artists anymore.

I'd love to see that program try to argue that the grosses of 'Pirates of the Carribean' are only the story, setting, characters, etc., and have nothing to do with Johnny Depp.

Posted by Sean at October 27, 2006 7:18 AM

comment #13

Midwest Doug says ...

Sean,

Your points are well-taken. I agree that we would lose 'interesting' big-budget movies. But so few big-budgeters are interesting, I'm not sure it's a great loss.

I share your skepticism that the program doesn't need to account in the actual talent involved. And I share your concern that studios would all start operating around this central area of themese instead of testing the limits.

But presumably, if Epagogix works well, it could have helped Warners avoid a Poseidon-like disaster. Maybe they'd spend that money on another big-budgeter. Or maybe they'd spend it on a few smaller budget movies that have some upside potential.

There's one key point that the studios may or may not get. For Epagogix to continue to work well it needs a lot of 'spread' in the data. If all the new films start coming out the same, the neural net will begin to bomb. It's like a gene pool that suffers from incest -- too much in-breeding and the whole thing falls apart. A neural net always needs fresh, new data points. The studios will need to put out 'mutations' -- movies outside the Epagogix mainstream, or movies where the confidence level about the accuracy of the results is lower. If the studios don't test new ideas, the whole thing will eventually fall apart.

I would never want a world where Epagogix alone is used to greenlight movies. And I have an optimism that however money-hungry the beancounters at the top, we will have art and artists making great films (digital video and distribution help this cause). But frankly, for big-budget movies, it makes sense to try to get more cost certainty. And if Epagogix can help, a studio would be silly not to try it.

Posted by Midwest Doug at October 27, 2006 8:06 AM

comment #14

christian says ...

"But presumably, if Epagogix works well, it could have helped Warners avoid a Poseidon-like disaster."

but one could have done that by hiring execs and development people who have a clue and not a netflix cue of old b.o. hits.

your premise ignores the obvious:

who greenlights a bonehead 100 mill plus remake of an old disaster movie that was a true product of its time? even harry knowles could have predicted a bomb.

Posted by christian at October 27, 2006 1:49 PM

comment #15

kimi98 says ...

Posted by kimi98 at August 31, 2007 5:10 AM

comment #16

cheapetiffany says ...

I am acutely aware that you have not elected me as your President by your ballots, and so I ask you to confirm me as your President with your prayers. And I hope that such prayers will also be the first of many. If you have not chosen me by secret ballot, neither have I gained office by any secret promises. I have not campaigned either for the Presidency or the Vice Presidency. I have not subscribed to any partisan platform. I am indebted to no man, and only to one woman -- my dear wife -- as I begin this very difficult job.cheap tiffany co jewelry

Posted by cheapetiffany at April 25, 2011 1:42 AM