Oscar Prediction: 2017's Best Picture

Introduction

This is a prediction for the winner of the year's Best Picture Oscar using a mathematical formula. Inspired by data-centric entities like the Football Outsiders, BoxOfficeQuant, and the VFX Predictinator, I devised an algorithm to accurately forecast the annual Best Picture Oscar winner.

Prior to the 2011 Academy Awards, I cooked up a recipe using quantifiable data (box office, critical reviews, award nominations & wins) that successfully predicted the previous decade of Best Picture winners (with one exception, Crash in 2005). This formula then accurately projected The King's Speech as the Oscar winner in 2011, and went on to properly foresee 4 winners in 6 years. See the past predictions and nominees from all years pitted against each other.

After incorrect predictions in 2014 (Gravity instead of 12 Years a Slave) and 2016 (The Revenant instead of Spotlight), I tweaked the formula to correctly predict the winners of those years and this 2.0 version of the formula has been retroactively applied to all years. Unfortunatly the 2.0 algorithm failed to predict the 2017 winner Moonlight.

I can no longer pretend to think this prediction formula actually works. I'll be as surprised as you if it comes up accurate this year.

Data

Prediction

Out of 9 nominees, Lady Bird has the best box office indicator, which propels it to 4th place overall. Dunkirk didn't take any of the top individual scores, but an overall strong showing places it in 3rd. The Shape of Water holds the greatest awards indicator, landing it in 2nd place. And Three Billboards has both the top critical success indicator and the top spot overall. If this damn formula works this time, then Three Billboards will be the next Best Picture.

Predicted Winner: Three Billboards Outside Ebbing, Missouri