Which factors forecast most accurately which of the nominees will win Best Picture? Total Revenue (adjusted) Total Budget (adjusted) Running Time Director Experience Source Material Studio Genre Release Date
Summary Statistics Winning Nominees MeanMinimumMaximumMedianStandard Deviation Revenue Budget Running Time Days released before ceremony
MeanMinimumMaximumMedianStandard Deviation Revenue Budget Running Time Days released before ceremony Losing Nominees Summary Statistics
P Values of Numerical Data at 95% Confidence VariableH0H0 H1H1 P Value Adjusted Revenue μ winners -μ losers =0 μ winners- μ losers > Adjusted Budget μ winners- μ losers =0 μ winners – μ losers > Days from Ceremony μ winners - μ losers = Running Time μ winners -μ losers = Hypothesis Testing
The Models VariableCoefficientsP-value Intercept Adjusted Revenue E Adjusted Budget E Days from release to award Running Time Experienced Director Comedy Drama Thriller True Story Original Major Studio I II VariableCoefficientsP-value Intercept Adjusted Revenue E Adjusted Budget E Days from release to award Running Time Experienced Director Comedy Drama Thriller Book Original th Century Fox Paramount Universal United Artists Columbia Warner Bros VariableCoefficientsP-value Intercept Running Time Experienced Director Thriller True Story R Squared F-Testn Model I Model II Model III III
Ŷ= x x x x x 1 = running time x 2 =experienced director (1,0) x 3 =Thriller (1,0) x 4 =True Story (1,0) Our Equation
Overall… Average ŷ Winners.358 Losers.242
Filmŷ Patton*.465 Airport.08 Five Easy Pieces.23 Love Story.23 M*A*S*H
Filmŷ Slumdog Millionaire.335 The Curious Case of Benjamin Button*.54 Frost/Nixon.04 Milk.0629 The Reader
Conclusions The strongest correlation we found in the variables we tested was a positive correlation between running time and Oscar Winnings We showed that the common assertion that films with release dates closer to the awards ceremony are more likely to win is likely a myth Overall, our model can only account for 11% of the variability in Data. You cannot quantitatively measure how “good” a movie is.
Credits Lead Excel Technician Warren Brown-Pounds Assistant Excel TechnicianBryce Gerard Data CollectionGrisel Zuniga Warren Brown- Pounds Bryce Gerard Regression AnalystsWarren Brown-Pounds Bryce Gerard Data Provided By imdb.com boxofficemojo.com PowerPoint DesignBryce GerardBased on an Idea byBryce Gerard Special Thanks to: Rajat Gupta