The Wisdom of the Illiterati


	lies, damn lies, and… Our personal memories of pop history are always suspect; unassisted by charts and history books, no one can ever seem to remember exactly when a record came out or how long it was around. This no doubt has something to do with the built in ephemerality of pop music and its insistence on the all-importance of the present. While you’re caught up in the music, or caught up in whatever you’re doing while the music is playing (working, playing, dancing, getting laid, etc.), all other considerations are driven out of your head. For most people, pop history is personal history, and personal history rarely comes with a timeline. So it shouldn’t have surprised me when last week, commenting on Michaelangelo’s latest Project X piece on Idolator, someone complained about the seemingly glacial pace of the top ten this year. He seemed to think that this was either something new, or that “Girlfriend” was a particularly egregious example of how slowly things were moving. It is true that summertime, when the labels take a deep breath and try to relax before the fall onslaught, is a slow season for the pop charts. But not only is this summer no slower than previous years, it’s considerably faster, and “Girlfriend” can hardly be taken, historically (and by historically I mean the last three years) as an emblem of chart stagnation. And though I’m no statistician, I have the numbers to prove it. So far this year, 43 records have made the top ten, 33 of which have finished their chart run (or I assume they have; there’s no reason that “Girlfriend” or even “Before He Cheats” couldn’t come roaring back again). Five of those records spent more than 10 weeks on the chart. 11 more spent between four and nine weeks, and the remainder three or less. The average amount of time any record spent on the chart is around five and a half weeks (I know, there are no halves in the top ten, but those are statistics for you). That means, on average, it takes about six weeks for the charts to turn over completely. That seems like a long time, but it’s nothing compared to previous years. In fact, the average length of time records stay in the top ten has been dropping every year since 2004. In that “Year of the House of Usher”, the average record stayed on the chart for nine and a half weeks, and 26 records, over half of the 55 that made the top ten, were there for over ten weeks, four of those for over 20 (if you combine the numbers for all of Usher’s releases that year, you get a whopping total of 86 weeks). Each year since then, however, the number of records to make the chart has increased, while the average length of time they stay there has dropped. In 2005, 56 records made the top ten, only one more than the year before, but the average stay in the top ten dropped to slightly less than nine weeks. In 2006, 68 records made the chart, with an average stay of slightly less than 8 weeks (the jump in total records may largely be attributable to Billboard including digital sales in their chart for the first time). As I mentioned before, the change in numbers this year is even more dramatic. By the end of June, 39 records had already made the chart—with the volatile fall release schedule yet to come—and the average chart run was only slightly more than 5 weeks. At this pace, we will most likely have 80, or more, top ten records by the end of the year. Even more fascinating than the average, however, is the median length of stay—that is, the length of time the majority of records actually appear in the top ten. In 2004 through 2006, the median is fairly close to the average: nine, eight, and seven, respectively. That means that there were roughly as many records spending over ten weeks on the chart as there were spending three weeks or less, pretty close to a normal distribution. This year, however, the median is three. I’ll repeat that: three. In 2004, over half the records to make the top ten were there for more than ten weeks; this year, the picture is reversed: over half the records to make top ten this year have spent three weeks or less on the chart. That low median is unusual not only when compared to the last few years, but to just about any year. In 1975, for example, 109 different records hit the top ten, none stayed on the charts for more than ten weeks (and only three managed that), and the median stay was almost exactly the same as the average: five. I haven’t checked any other years that far back yet, but will try to do so soon. I have a feeling, though, that even in the busiest years, the average, and the median, never dip much below five weeks. Obviously, much of the change this year has to do with the growth of digital distribution and the change in people's buying habits that come with it. While album sales (both physical and digital) continue to drop, track downloads have increased by nearly 50% over last year. At the same time, the people who keep the charts, i.e. Billboard and Nielson Soundscan, are trying desperately to reformulate their calculations to give a more accurate picture of what’s popular at the moment. Billboard has announced that they’ll reformulate their charts in the next couple of months, increasing the weight of digital purchases in their calculations. At the same time, they’re trying desperately to come up with a way to accurately gauge streaming internet traffic and are even considering monitoring P2P services, a move that could earn them a lot of flack from the major labels, who might see it as encouraging piracy (there are already companies, however, selling P2P info to radio stations). What this means for the music, of course, is pretty much nothing. More records on the charts means more opportunities for new artists, but buying habits change slowly, and changes in cultural taste can take even longer. It will be interesting to see what happens in the coming months, when the majors roll out their big guns: Usher, Mariah Carey, 50 Cent, Kanye West. It’s doubtful that any of them will rule the charts the way they did in the past—and I’ll go out on a limb and predict that both 50 Cent and Usher will essentially tank; that is, big first week album sales, no truly big hits, and a sudden plummet—but even if they do, they’ll be the freaks in the pack, the established artists who still benefit from name recognition and the old, atrophied distribution system. Meanwhile, a couple of dozen others will be thrashing around in the lower reaches of the charts, waiting for a chance to cut the big guys off at the knees. It’s only a matter of time before they do. \| [Archives] Search entries:		top ten 1966 top ten OTHER OCCASIONAL RAMBLINGS Hail Blogoslavia! ~~100~~ 45 or so things about JSM The Saga of the Lone Biscuit OTHER BLOGS Blogroll Me

lies, damn lies, and…

Our personal memories of pop history are always suspect; unassisted by charts and history books, no one can ever seem to remember exactly when a record came out or how long it was around. This no doubt has something to do with the built in ephemerality of pop music and its insistence on the all-importance of the present. While you’re caught up in the music, or caught up in whatever you’re doing while the music is playing (working, playing, dancing, getting laid, etc.), all other considerations are driven out of your head. For most people, pop history is personal history, and personal history rarely comes with a timeline.

So it shouldn’t have surprised me when last week, commenting on Michaelangelo’s latest Project X piece on Idolator, someone complained about the seemingly glacial pace of the top ten this year. He seemed to think that this was either something new, or that “Girlfriend” was a particularly egregious example of how slowly things were moving. It is true that summertime, when the labels take a deep breath and try to relax before the fall onslaught, is a slow season for the pop charts. But not only is this summer no slower than previous years, it’s considerably faster, and “Girlfriend” can hardly be taken, historically (and by historically I mean the last three years) as an emblem of chart stagnation. And though I’m no statistician, I have the numbers to prove it.

So far this year, 43 records have made the top ten, 33 of which have finished their chart run (or I assume they have; there’s no reason that “Girlfriend” or even “Before He Cheats” couldn’t come roaring back again). Five of those records spent more than 10 weeks on the chart. 11 more spent between four and nine weeks, and the remainder three or less. The average amount of time any record spent on the chart is around five and a half weeks (I know, there are no halves in the top ten, but those are statistics for you). That means, on average, it takes about six weeks for the charts to turn over completely.

That seems like a long time, but it’s nothing compared to previous years. In fact, the average length of time records stay in the top ten has been dropping every year since 2004. In that “Year of the House of Usher”, the average record stayed on the chart for nine and a half weeks, and 26 records, over half of the 55 that made the top ten, were there for over ten weeks, four of those for over 20 (if you combine the numbers for all of Usher’s releases that year, you get a whopping total of 86 weeks).

Each year since then, however, the number of records to make the chart has increased, while the average length of time they stay there has dropped. In 2005, 56 records made the top ten, only one more than the year before, but the average stay in the top ten dropped to slightly less than nine weeks. In 2006, 68 records made the chart, with an average stay of slightly less than 8 weeks (the jump in total records may largely be attributable to Billboard including digital sales in their chart for the first time). As I mentioned before, the change in numbers this year is even more dramatic. By the end of June, 39 records had already made the chart—with the volatile fall release schedule yet to come—and the average chart run was only slightly more than 5 weeks. At this pace, we will most likely have 80, or more, top ten records by the end of the year.

Even more fascinating than the average, however, is the median length of stay—that is, the length of time the majority of records actually appear in the top ten. In 2004 through 2006, the median is fairly close to the average: nine, eight, and seven, respectively. That means that there were roughly as many records spending over ten weeks on the chart as there were spending three weeks or less, pretty close to a normal distribution. This year, however, the median is three. I’ll repeat that: three. In 2004, over half the records to make the top ten were there for more than ten weeks; this year, the picture is reversed: over half the records to make top ten this year have spent three weeks or less on the chart.

That low median is unusual not only when compared to the last few years, but to just about any year. In 1975, for example, 109 different records hit the top ten, none stayed on the charts for more than ten weeks (and only three managed that), and the median stay was almost exactly the same as the average: five. I haven’t checked any other years that far back yet, but will try to do so soon. I have a feeling, though, that even in the busiest years, the average, and the median, never dip much below five weeks.

Obviously, much of the change this year has to do with the growth of digital distribution and the change in people's buying habits that come with it. While album sales (both physical and digital) continue to drop, track downloads have increased by nearly 50% over last year. At the same time, the people who keep the charts, i.e. Billboard and Nielson Soundscan, are trying desperately to reformulate their calculations to give a more accurate picture of what’s popular at the moment. Billboard has announced that they’ll reformulate their charts in the next couple of months, increasing the weight of digital purchases in their calculations. At the same time, they’re trying desperately to come up with a way to accurately gauge streaming internet traffic and are even considering monitoring P2P services, a move that could earn them a lot of flack from the major labels, who might see it as encouraging piracy (there are already companies, however, selling P2P info to radio stations).

What this means for the music, of course, is pretty much nothing. More records on the charts means more opportunities for new artists, but buying habits change slowly, and changes in cultural taste can take even longer. It will be interesting to see what happens in the coming months, when the majors roll out their big guns: Usher, Mariah Carey, 50 Cent, Kanye West. It’s doubtful that any of them will rule the charts the way they did in the past—and I’ll go out on a limb and predict that both 50 Cent and Usher will essentially tank; that is, big first week album sales, no truly big hits, and a sudden plummet—but even if they do, they’ll be the freaks in the pack, the established artists who still benefit from name recognition and the old, atrophied distribution system. Meanwhile, a couple of dozen others will be thrashing around in the lower reaches of the charts, waiting for a chance to cut the big guys off at the knees. It’s only a matter of time before they do.