When it comes to predicting a movie's success or failure, as William Goldman so memorably wrote, nobody knows anything. But that's not for lack of trying. Here at indie ground zero, Matthew Prince and his merry band of techies at Unspam Technologies Inc. are pretty sure they've come up with a formula to reliably predict success in two of the Sundance Film Festival's competition categories: drama and documentary.
Good news for "TV Junkie" and "Flannel Pajamas." Not so good for "Puccini for Beginners" and "small town gay bar."
"Our engineers were thinking that determining whether a movie is good or bad could be similar to determining whether e-mail is spam or not," said Unspam Chief Executive Prince, 31, who loves the festival and uses it as a recruiting tool. "We had the last 10 years of the festival's film guides, which are like inputs, and then a bunch of outputs, like how many people saw a film, did it win anything at Sundance, did it have commercial success. If you could figure out the pattern between the inputs and the outputs, then you could actually predict future winners."
They scoured the Sundance film guides, which include descriptions of each movie signed by one of about five Sundance selection committee members, the number and location of screenings, as well as information available on the Internet Movie Data Base and box office figures. They crunched the data using a "naive Bayesian classifier," a statistical method of figuring out whether something is more like one thing or more like another thing, Prince said. "That's the principle behind the most advanced spam filtering technology today."
And here's what they found: Among the best predictors of failure is the number of producers on a film. "If you only have one or two producers on your film, you're statistically more likely to have a stinker," said Prince, which is the opposite of conventional wisdom on the topic (too many cooks, etc.). Five or more producers, it turns out, is best.
If, during the festival, a film is preceded by a short, it has a higher chance of being a flop (43%) than a hit (35%). If a film has its Park City premiere at the Eccles Theater, one of the largest venues here, it's less likely to go on to success. And though Sundance has pushed digital filmmaking, movies shot on Sony HD only have a 19% chance of being a hit. Prince admits that what constitutes success is "fuzzy," but, he wrote on the project's website, www.deconstructingsundance.com, "[t]esting the system with known data from previous years, we have established an approximately 81% typical accuracy rate on a year-by-year basis."
The most entertaining part of Unspam's analysis is the deconstruction of the film descriptions in the festival's program written by selection committee members. Prince and his crew came up with two lists: words that "make you golden" or are "the kiss of death."
Golden: academic, accomplished, bedroom, complex, dialogue, dream, death, focus, girl, human, high, journey, love, mother, narrative, romance, relationship, superbly, sex, ultimately.
Kiss of death: Africa, America, American, beautiful, black, best, emotional, fascinating, great, inspired, lake, new, riveting, Sundance, sexy, story, subtitles, truth, vision, world.
"A lot of adjectives, like 'riveting' ... that's a bad indication," Prince said. "Whereas words that are tangible tend to be pretty good things."
They also discovered that those blurbs' bylines correlate with how well films do. Festival director Geoffrey Gilmore, festival programming director John Cooper and programmer Shari Frilot tend to write descriptions of movies that go on to success. Programmers Caroline Libresco, Lisa Viola and Trevor Groth tend to, as Prince puts it, "pick stinkers."
Cooper, who was striding into a theater to introduce Lauren Greenfield's eating disorders documentary "Thin" on Saturday, chuckled at the mention of the Unspam project, Deconstructing Sundance. "I've heard about it, but I haven't seen it. More power to them."
This year, predicts Unspam, the most successful dramas will end up being "Come Early Morning," "Flannel Pajamas," "The Hawk Is Dying," "Quinceanera," "Somebodies," "Stay" and "Stephanie Daley."
Least successful will include "Puccini for Beginners" and "Steel City."
In the documentary competition, they are predicting success for "God Grew Tired of Us," "The Grand Truth: After the Killing Ends," "So Much So Fast," "Tis Autumn -- The Search for Jackie Paris," "TV Junkie" and "Wide Awake."
As Unspam worked on this project, Prince also obtained the 2004 audited financial statements of the nonprofit Sundance Institute, the festival's corporate parent, filed with Salt Lake County as part of a grant request. Highlights are posted on the website. Revenue from ticket sales was more than $4 million, total assets jumped 254% from $7.2 million in 2002 to $25.8 million in 2004. (Tickets cost $10, with an additional $1-per-ticket handling fee, which is new.)
Meanwhile, Prince and his colleagues have crafted a disclaimer that appears on their website: "You shouldn't make multimillion dollar decisions with a studio's money based on this analysis. On the other hand, if you buy one of the movies we've predicted to be a winner ... our engineers would be happy to attend your acquisition party to talk about the incontrovertible proof of box office success the math provides!"