Big Data Predicts Oscar Winners

Last month, the winners of the 85^th Oscar Academy Awards revealed. Ann Lee, one of the best Taiwanese directors, won the best director. I am proud of him, but let go back to the topic. Many people like to do Oscar prediction according to many factors, and this is why big data is so fascinating. Because we are not in the movie industry, we do not know what the factors are influencing the results of voting to the winners. Here, I list some interesting findings from the statistical information.

1.    Of the past 84 Best Picture winners, only 21 movies have had one-word names.

2.    The Best Picture award is worth anywhere from $20 million to $50 million in additional sales to the winner

3.    According to several reports, the multiple award ceremonies from the Golden Globes to the Screen Actors Guild (SAG) Awards are the real predictors of Academy Award winners.

4.    Twitter was also a good predictor of the winners.

But, Oscar winners are not only based on the information I list. Columbus, Ohio-based data sciences company Farsite thinks they got the model. They successfully predicted 5 of 6 big winners including best picture, best actor, best actress, best supporting actor, and best supporting actress. They only missed best director that they voted to Steven Spieberg.  What is their algorithm? 

They put sources that they think useful into the model. It will be an interesting research if we can know the weights from different sources year by year. Then maybe we can examine what the main factors are to let those voters make the decisions.

Add caption

References:

1. http://www.farsiteforecast.com/vanity-fair-polls-and-parties/

2. http://www.farsiteforecast.com/and-the-oscar-goes-to/