Software predicts which songs will be hits
I have to say two things. One, it would probably work. Two, it shows the main flaw in the music business: Music is not bought directly by listeners, but by promoter/producers. The lawyers in the article. Lawyers bet on past decisions, that’s the basis of common law.
One would hope that the internet would cut them out, but I still see a new layer of “taste makers” that you have to appeal to in order to have anyone find your demo in a sea of wannabes.
[snob]
Most popular music is so formulaic and boring that it could be made by a computer program, so it’s not that surprising that a computer program can analyze and it determine if it hits the lowest common denominator points musically and have mass appeal.
[/snob]
I wonder how different the lowest common denominator points are between the decades; what musical aspects are great predictive factors for hits in, say, the 70’s vs. hits of 2009?
I wonder if you could alter the various weights and algorithms to find the lowest common denominator points for songs that are hits with snobby music critics!
50’s - Sounds like Elvis
60’s - Sounds like Beatles
70’s - Sounds like Bee Gees
90’s - Sounds like Teen Spirit