I've posted previously about the movie 'The Illusionist' and the marked differences that buzz volume for that movie has compared to the more common distribution of buzz regarding a movie. The graphs below captures this difference by showing mentions of 'illusionist' for 41 days starting from 1 week prior to release and the same for 'superman returns':
The graph for 'superman returns' shows the classic initial peak on the opening weekend followed by repeated dips and peaks for each week/weekend.
One question we might ask is: does discussion around The Illusionist have some clear difference to that around Superman Returns? To answer this, we must first isolate discussion around the two movies. This requires the ability to resolve any ambiguity for the two search terms. It turns out that 'illusionist' is the more ambiguous of the two terms as it can be used to refer to a character type in a number of role playing games as well as the title of a stage magician or other conjurer (e.g. David Copperfield). Once we can distinguish messages that are about the two movies from these and other sense associated with our search, we can begin to look at the difference in conversation.
By looking for passages of text discussing the movies (via disambiguated terms) that also include terms that indicate (positive) recommendations (via terms such as 'recommend' and 'worth seeing') we can get a picture of the intent within these messages. The results are very interesting. For blogs, approximately 2.1% of mentions for Superman Returns had a positive recommendation associated with it. For the Illusionist, however, this number was 5.9%. This appears to give a clear indication that posting on blogs around these two movies has some clear difference - specifically, the positive recommendation of the film in question.
Another clear signal found in the data was that a measure of sentiment around the two films indicated that The Illusionist is regarded more favourably than Superman Returns.
There are a number of caveats to consider here. For example, the terms used to recognize positive recommendations may well be skewed towards one type of audience (an art house audience) rather than another (the summer block buster audience). However, I believe that directionally the findings are correct:
- The Illusionist began with only moderate expectations (hence the small number of screens it initially opened on)
- It has the traits of a sleeper hit - increasing popularity over time indicated by increases in screens allocated and buzz around the movie.
- Posts in blogs about The Illusionist are more likely to recommend it that do posts around our control movie Superman Returns.
- The movie has been more positively received by those who have seen it (a point that is also reflected, for example, in the reviews found on BoxOfficeMojo).