Last year, a "motley group of techie folks" came together and used Bayesian analysis (don't worry, I'm not smart enough to figure it out either) to predict which films at Sundance would win the four major awards. Part of their process includes gathering 11 years worth of festival guides, information off IMDb and an assortment of film reviews, throwing it all in some sort of virtual whiz-kid blender and, apparently, out spits which films are most likely to finish on top. It sounds pretty nutty, but in 2006 they accurately predicted all four winners. Yes, I'm serious.
Since the big Sundance awards are being handed out tomorrow -- we're talking the Dramatic Competition Audience Award, the Dramatic Competition Grand Jury Prize, the Documentary Competition Audience Award and the Documentary Competition Grand Jury Prize -- we decided to check back in with Deconstructing Sundance to see what their picks for 2007 are. Basically, they break it down into three hilariously-named categories: Films under the Timberlake Britney category are the best, Federline Britney lists those in the middle and Pantiless Britney, well, you get the drift. Beside each film is a percentage level which is supposed to represent how confident they are in their decision to place a film in a certain category.
So ... you probably want to know which films they picked, huh? Well, according to their 2007 predictions list (and hopefully I'm reading it correctly), The Good Life (98.76%) and Joshua (99.27%) are the two favorites in dramatic competition, followed somewhat closely behind by Grace is Gone (89.13%), Four Sheets to the Wind (86.69%) and Hounddog (66.81%). And the worst film in dramatic competition? Adrift in Manhattan wound up in the Pantiless Britney category with 83.76%.