Here are some of the links I’ve been following over my first cup of coffee:
- CNET reports that Netflix anticipates at least some backlash over their price increases. In response, Blockbuster has stepped up their efforts to market their DVD subscription plans.
- Ryan Lawler traces the anticipated sources of advertising revenue for TV advertising as more and more people turn to connected TV sets and streaming TV.
- The Economist has launched its new documentary film website, providing documentary filmmakers with yet another site for finding an audience for their films.
- The IFC Blog reports on an MIT study that considers biases in the “star systems” that are typically used to evaluate movies on sites such as Netflix, where those rankings inform recommendation algorithms. The author of the study, Devavrat Shah, suggests that external factors, such as the user’s mood, may affect ratings and proposes that more accurate rankings would be produced if films are paired.
- On a related note, USC’s Annenberg Innovation Lab has launched a new tool, the USC Annenberg Film Forecaster, that seeks to quantify anticipation for an upcoming movie. The tool measures both the number of tweets and the sentiment expressed in those postings. One interesting example: they were able to track the fact that Green Lantern was getting quite a bit of buzz but that approximately 40% of it was negative. One of the difficulties they are addressing is how to teach the tool to recognize sarcastic tweets. The tool is also limited by Twitter’s ostensibly younger demographic, suggesting that adult-oriented films may get fewer mentions.
- This is more for fun, but someone has created an amusing infographic showing which Netflix films are popular in different parts of the country.