Will hit science work?

Happy new year!

My friend Rossella in Rome got an email from a company that runs computer analisys on sound files and assesses its potential to be a hit. They have a database of Billboard chart hits and track about 30 different variables on frequencies, melodic patterns, chord sequence etc.

Hi Ms FALCONE, I’m Roger, from Polyphonic HMI, company dedicated to the creation and development of new technology within the musical industry. Now we have designed and developed a software that can predict if a song is going to be a hit! We are sure that this tool will be of enormous help to the record labels and producers

80% RELIABILITY GUARANTEED!!!

If you think you can be interested in our Hit Song Science “technology”, we would be willing and happy to arrange a meeting at Midem. I attach some detailed information about us and I also invite you to visit our web page www.polyphonichmi.com

I am very skeptical, obviously. There are a lot of problems involved. Even the database is skewed towards American taste and the past (also, prior to the intorduction of Soundscan I think the charts were very unreliable). But I am also curious. So I will do two things: I will ask Peter Walsh, our producer, what he thinks about it. And I will ask my friend Raffaele, who is a talented statistician teaching at Padova University, how he reckons the number crunching works.

If you are curious, too:

www.hitsongscience.com

But let it be clear that I am NOT endorsing this. I don't know enough about it.