Attention All Datamining / Music Geeks

While trying to get the correct Rakim quote for the next post I stumbled upon this: Albums by Year -- Rate Your Music. It looks like it should provide an absolutely stunning database for anyone interested in long term data visualization. How do these lists, which are just surprising enough to feel like a very valid dataset, compare to Billboard's top albums? What about to the Village Voice's annual Pazz & Jop critics poll? And perhaps most interestingly how do they change over time?

For instance 1991, a year I listened to an awful lot of music, is fascinating. At what point did the Main Source record, which almost no one listened to in 91 become higher rated than Nirvana? When did Quest hit #1 and My Bloody Valentine, who were relatively obscure back then hit #2? What about the Swans at 13? They were not just obscure and not just the underground, they were the obscure side of the underground.

It's interesting just how clearly musical peaks shine through, pick a given year and you'll probably get the sense that a certain genre is over represented, death metal in 91 for instance. But if you look over span of years it becomes clear that it is not true, the over represenations are probably accurate representations of a genre's prime eras. For instance hip hops two golden ages, in the late 80s and mid 90s shine super bright on the top of the charts. Yet by 1997 it barely shows up, the classics replaced by only churn.

It should be fascinating to watch this stuff change as bands and retros go in and out of style, classics get forgotten or dug up, tastes go sour then shine bright again and the music continues to play.