AUTHOR=Christoforou Christoforos , Papadopoulos Timothy C. , Constantinidou Fofi , Theodorou Maria 

TITLE=Your Brain on the Movies: A Computational Approach for Predicting Box-office Performance from Viewer’s Brain Responses to Movie Trailers

JOURNAL=Frontiers in Neuroinformatics

VOLUME=Volume 11 - 2017

YEAR=2017

URL=https://www.frontiersin.org/journals/neuroinformatics/articles/10.3389/fninf.2017.00072

DOI=10.3389/fninf.2017.00072

ISSN=1662-5196

ABSTRACT=The ability to anticipate the population-wide response of a target audience to a new movie or TV series, prior to its release, is critical to the film industry. Equally important is the ability to understand the underlying factors that drive or characterize viewer’s decision to watch a movie. Traditional approaches (which involve pilot test-screenings, questioners and focus groups) do not perform at par. In this study we develop a novel computational approach for extracting neurophysiological (EEG) and eye-gaze based metrics to predict the population-wide behavior of movie-goers. We further, explore the connection of the derived metrics to the underlying cognitive processes that might drive moviegoers’ decision to watch a movie. Towards that, we record neural activity - through the use of electroencephalography (EEG) -  and eye-gaze activity from a group of naïve individual while watching movie trailers of pre-selected movies for which the population-wide preference is captured by the movie’s market performance (i.e. box-office ticket sales in the US). Our findings show that the neural based metrics, derived using the proposed methodology, carry predictive information about the broader audience decisions to watch a movie, above and beyond traditional methods. In particular, neural metrics are shown to predict up to 72% of the variance of the movies’ performance at movie’s premiere and up to 67% of the variance at subsequent weekends; which corresponds to 23-fold increase in prediction accuracy compared to current state-of-the-art neurophysiological or traditional methods. Moreover, the derived neurophysiological metrics are related to cognitive states of enhanced arousal and focused attention, the encoding of long-term memory, and the synchronization of different components of the brain’s rewards network. Beyond the practical implication in predicting and understanding the behavior of moviegoers, the proposed approach can facilitate the use of video stimuli in neuroscience research; such as the study of individual differences of attention-deficit disorders, and the study of desensitization to media violence.