The Bigger Picture: Combining Econometrics with Analytics Improve Forecasts of Movie Success
There exists significant hype regarding how much machine learning and incorporating social media data can improve forecast accuracy in commercial applications. To assess if the hype is warranted, we use data from the film industry in simulation experiments that contrast econometric approaches with tools from the predictive analytics literature. Further, we propose new strategies that combine elements from each literature in a bid to capture richer patterns of heterogeneity in the underlying relationship governing revenue. Our results demonstrate the importance of social media data and value from hybrid strategies that combine econometrics and machine learning when conducting forecasts with new big data sources. Specifically, while recursive partitioning strategies greatly outperform dimension reduction strategies and traditional econometric approaches in forecast accuracy, there are further significant gains from using hybrid approaches. Further, Monte Carlo experiments demonstrate that these benefits arise from the significant heterogeneity in how social media measures and other film characteristics influence box office outcomes.
We wish to thank Chris Hansen, seminar participants at the Young Econometricians around Pacific (YEAP) 2017 annual conference, the Canadian Econometrics Study Group (CESG) 2017 annual conference, Carleton University, Chinese Academy of Sciences, Northeastern University, Renmin University, Xiamen University, and Zhejiang University for helpful comments and suggestions. Xie’s research is supported by the Natural Science Foundation of China (71701175), the Chinese Ministry of Education Project of Humanities and Social Sciences (17YJC790174), the Natural Science Foundation of Fujian Province of China (2018J01116), the Fundamental Research Funds for the Central Universities in China (20720171002, 20720171076, and 20720181050), and Educational and Scientific Research Program for Young and Middleaged Instructor of Fujian Province (JAS170018). Lehrer wishes to thank SSHRC for research support. The usual caveat applies. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.