How Pandora Works

Pandora has, at its core - a concept that a musical song comes from a set of genomes. Like our genome, it is built from a simple set of basic elements. In our genome, we're like. A to T and C to G or something like that. But in a musical genome, the song is built from the base characteristics - normally things working together like background vocals, and a lead vocal - or certain types of musical influences that have already been catalogued. It's really just rhythm, tone. A few simple things. Then you combine those into sort of groups -for example, synthesis, electronic backbeat, etc. - and you get an expressed trait. This group is the way that the song gets catalogued and indexed.

Here's the patent for the method. It's kind of deep, so hold your breath and get focus.

A method of determining at least one match item corresponding to a source item. A database of multiple items such as songs is created. Each song is also represented by an n-dimensional database vector in which each element corresponding to one of n musical characteristics of the song. An n-dimensional source song vector that corresponds to the musical characteristics of a source song is determined. A Distance between the source song vector and each of database song vector is calculated, each distance being a function of the differences between the n musical characteristics of the source song vector and one of source database song vector. The calculation of the distances may include the application of a weighted factor to the musical characteristics of resulting vector. A match song is selected based on the magnitude of the distance between the source song and each database songs after applying any weighted factors

In plain english, the song is first sorted into how many trait groups it has (calling it "n-dimensional") and then each element is ranked. Once you have that song ranked, they watch to see what you skip over. If you skip a certain song - they vary one element of the vector to see what you didn't like - but they also go back to something that they think you have already liked. The easiest way to mess up the algorithm is to hate everything. Downrank it all and they won't be able to sort based on where they think you're going. But the music catalogue for your 'channel' will still come out ok, because its really based on grouping like songs by their elements and traits. In other words, there's no way to destroy a channel. You can only lose your identity if you choose to play around with it.

This is why Pandora will inevitably choose embarassing songs when your friends come over. Pandora is really a mathematical system that ignores your friends - it deals with how some unpopular musicians are unfairly overlooked — or how some popular ones are unfairly maligned - and it tries to make musical comparison to things you like, based on the idea that what you like can be taken apart and put back together again. In many cases this is true. But in one case it is not.

In "Super Sad True Love Story" by Gary Shteyngart, we find a world set in the near future - in which almost everyone has been exposed to pornography. A world where personal revelation is more important than personal growth - a world in which books are not read but scanned for text, and people - when they meet - spend more time talking about themselves than actually, to each other (because everything is being used to determine their public rank online - they tend to be constantly promoting themselves).

It is not unlike, perhaps the world described in "TiMER" .. where people might trust finding 'the one' to a computer program. It's just that unlike 'TiMER' , or 'Super Sad' - Pandora doesn't really care what others think - there is only the music, and its attempt to categorize it and sort it so that you can explore the depth and dimension of its sound. Pandora is way to escape the world in which everything is decided for you. As long as you limit your channels, use them as a starting point - and continue to search for, and explore music - on your own.