wisdom of crowds in movies today

It has been widely acknowledged that the “wisdom of crowds” as demonstrated in prediction markets (Surowiecki, 2004, Manski, 2006) is a surprisingly accurate mechanism to predict future trends. Large groups of “ordinary” people are better in predicting trends than a single expert. At the same time, the Web has turned into a major platform for information exchange, thus becoming a mirror of the real world: Millions of volunteers post latest news on Web sites such as Wikipedia, and political blogs such as dailykos and instapundit. In addition people express their opinions in forums and online communities, and tell openly what matters to them. Approaches such as “Netnography” (Kozinets,2002) make use of this fact for marketing research, proposing analysis of statements of “devotees” and “insiders” in online forums and other Web sites. This paper proposes combining these two ideas, interpreting opinionated discussions and the level of “buzz” about the movie business on the Web as some kind of a prediction market. Our approach offers an automated, efficient, and cheaper way to tap people’s opinion than polling people over the phone. Our method calculates levels of “Web Buzz” by mining discussions in movie- related online forums, combining information about the structure of the social network with an analysis of the contents of the discussion. This paper demonstrates our approach by predicting the success of movies based on the communication in the online community IMDb.com. We analyze its communication patterns in regard to metrics like “intensity” and “positivity”. While intensity means the frequency of the subject in discussion, positivity refers to the degree of positive feelings towards a movie expressed by contributors. Thereby we factor in quantitative and qualitative dimensions of discussion allowing us to extract an aggregated community opinion about individual movies. These measurements are the basis of our two hypotheses. First, we assume that the chances of a movie to win an Oscar can be determined by the communication structure of the IMDb community. Second, we speculate that there is a relationship between the communication intensity about a movie and the performance of the movie at the box office.