New York

SXSW 2006: Day 3, ‘This Is Art, Right?’


The above is a direct quote about the reluctance to understand music via science from one Mike McCready, CEO of Platinum Blue Music Intelligence in Barcelona, Spain. But we come to SXSW to learn, no? So bear with me as I try to explain McCready’s panel yesterday on “Quantum Music Analysis (QMA) of Hit Songs” (and I will necessarily misrepresent some of these heady concepts which I imagine McCready would forgive).

Turns out that the “scientists” at Platinum Blue have discovered that 85 percent of all hit songs conform to certain “hit clusters” which themselves conform to mathematical patterns. They arrived at this conclusion by analyzing every song in the Billboard Top 30 (although I didn’t catch how far back in time the analysis went) with a set of musical criteria (melody, beat, chord progression, etc.). Lyrics are not analyzed for several reasons: computers cannot detect word play and double entendres and like the FBI, they often cannot understand the words. Top 30 songs come from only a few clusters.

So for $5000 per full CD (they’re working on discounts for indies), Platinum Blue will analyze your songs to determine if they conform to any of these hit clusters. If your song sounds like a hit to you (and you should know; according to McCready, human ears have not evolved much since the time of . . . wait for it . . . Beethoven) and you can afford to promote it AND it conforms to a hit cluster, then Platinum Blue maintains that your song has an 80 percent chance of becoming a hit.

Platinum Blue has a non-disclosure agreement with their clients. But since several majors have come out in a recent Harvard business school study, I asked McCready if he could reveal some of the songs that Platinum Blue predicted would be hits. Instead of simply telling us, McCready felt compelled to find something on his computer that would answer my question. In the meantime, he took other questions. But that meantime went on for about fifteen minutes. So I interrupted someone else’s question to get McCready to answer mine. Turns out he forgot and proceeded to futz with his computer and take more questions in another meantime. A few people around me doubted he would ever give us an answer.

Finally, he produced a copy of the Harvard report (with “DO NOT COPY” splayed all over it) which listed about ten songs that Platinum Blue successfully predicted as hits. But I was in the back row and could not read it. So I had to ask McCready to read it out loud. It felt like getting a hit song out of a turnip. But here are some winners (read very quickly, I might add): Monica: “So Gone;” Uncle Kracker: “Drift Away;” Brad Paisley: “I Wish You’d Stay;” and Pazz & Jop 2003 winner Outkast: “Hey Ya!”

OK even with this cursory explanation, I know you’re dying to poke holes in all this. But a thorough poking would take us to the end of the internet. I will admit that his talk of “false positives” and “false negatives” (Platinum Blue rarely get the latter) reminded of the Sex Negative and Positives in Rinse Dream’s creepy, post-apocalyptic porno Café Flesh. But in a kinder cut, I’d say that McCready is no less confused about gross cultural overproduction than any of the scurrying SXSW participants are. And on one level, his presentation was wackier than anything Ariel Pink offered up. More on that later.

Most Popular