Algorithmic Entertainment
Algorithms might soon be in the director’s chair.
Video streaming sites such as Netflix are propelled by their use of data. With many millions of people watching, the company knows what genres, actors and type of content are popular, and importantly, they know who they’re popular with.
The theoretical advantage of this is Netflix can commission original content with a high degree of confidence that it will be successful, provided the content is pushed towards the right people. In their book Streaming, Sharing, Stealing, Michael Smith and Rahul Telang explore this idea, recounting the tale of Netflix and the TV show House of Cards.
The show’s protagonist/antagonist Frank Underwood was deemed by other networks to be too evil for fans to like him, and thus like the show. But Netflix thought the opposite; their data suggested people liked shows with anti-heroes, political intrigue, and mafioso-esque characters, and believed viewers would like Frank Underwood precisely because he was unlikeable.
They commissioned the show, and House of Cards was a massive success.
This success, however, is perhaps an exception that disguises an interesting tension between commissioning executives and showrunners. Let’s return to House of Cards.
