“House of Cards” and Our Future of Algorithmic Programming

[Jopa]

The bigger, possibly darker omen for the future of TV is found in several articles about why Netflix decided to make its original programming bet on House of Cards, specifically, as opposed to some other series about, say, zombies or teenagers. It bought House of Cards based on what it knows about the viewing habits of its 33 million users—it knew which and how many users watch movies starring Kevin Spacey and the director David Fincher, and, through its tagging and recommendation system, how many sat through other similar political dramas. It has shown different trailers to people depending on their particular viewing habits, too.

 

Jonathan Friedland, the company’s chief communications officer, said, “Because we have a direct relationship with consumers, we know what people like to watch and that helps us understand how big the interest is going to be for a given show. It gave us some confidence that we could find an audience for a show like ‘House of Cards.’ ”

 

 

[rolf]

And that is not algorithmic programming. That is nothing more than a very hard database query. More like collaborative filtering.

Can easily be done in SQL. Get the top 100 most interesting actors by number of viewers. Get the top 100 most interesting directors by number of viewers. Get natural joint. As a result, you'll probably have about 10 results. Check which one is most relevant. Buy it.

Of course, you can make a bit more dependencies than only actors, for example give a counter for every actor and for every user that watches a series give that counter a +1. Then make a list of all series rated by the sum of all counters of actors that play in it. Then, sort by score, and work top down on best series (on actors now only, doesn't matter).

 

Note the difference between hard and easy. Hard means because it takes a long time to calculate. Easy because it is easy to write. Only reason I don't give an example to that is because I hate SQL.

 

 

Let's just say that I'm working on algorithms right now, and stating this is an algorithm is just wrong. Probably written by someone who doesn't know the difference

 

Also, this happens already a lot. Facebook uses this method as giving examples of which persons you could also know. The only difference to this is summarizing this approach over all users.