Pazz & Jop Stats: On Kanye West's Dominance, And What The Poll Would've Looked Like Without Him
"If there was any way you could confirm Kanye's inevitable victory early, that'd be a big help art-department-wise." So began Pazz & Jop editor Rob Harvilla's first note to me after voting closed in late December. Inevitability is not a statistical quality, but the magnitude of My Beautiful Dark Twisted Fantasy's win was indeed striking: It got nearly twice as many points as LCD Soundsystem's This Is Happening at #2 (giving it the biggest winning margin in the 38-year history of the poll), with an enthusiasm rating (average points per voter) higher than anything else with even 10 percent as many votes. But a look down the rest of the top 10 reveals more than just Kanye's dominance. Critics seem to be in agreement on lots of things this year.
To put this year's results in perspective, check out this chart of the point totals for the top 20s in 2010 (blue), 2009 (green), and 2008 (red), from #1 on the left end to #20 at the right end, and keep in mind that there were only 16 more voters in 2010 than 2009.
The precipitous leap at the left end of the blue line is Kanye; the winning margins the previous two years were nowhere near as large. But the gap between the blue and green lines all the way across is maybe even more interesting: All the winners had higher totals this year, from approximately the same number of voters. "Consensus" is probably the wrong word for this -- the negative-minded will note that more than 60 percent of voters didn't give Kanye a single point, and this year's #20 appeared on less than 7 percent of the ballots -- but there's definitely something happening.
This breakdown of top-10 voter-overlap gives us another way to look at the data. In previous years there have been obvious holes in the grid, places where the voting factions distinctly do not align. This year, not so much. Arcade Fire supporters didn't tend to vote for Big Boi, but otherwise, the overlaps were more or less proportional. So has the hive mind taken over? Is there an ever-clearer collective P&J bias each year? Or was it just more obvious to everyone in 2010 what the best albums were?
One of the other things we can do in Needle, the database software we used to data-correct and tabulate Pazz & Jop this year and last, is play with variations of our calculations. For example, my own next question, after seeing the scale of Kanye West's victory, was to wonder what the results would look like without him. Not just everybody else moving up a place, I mean, but retabulating the whole poll without the voters who voted for Kanye.
The results themselves don't change a whole lot in this scenario. Big Boi drops from #6 to #10, as almost two-thirds of his voters also voted for Kanye. Joanna Newsom and Grinderman both move up, as more than half of each of their voters didn't. But it's actually even more revealing to re-sort the table not by non-Kanye votes, but by precisely these percentages. So this table ranks albums by the percentage of their points that came from non-Kanye voters. Of the 70 albums with 10+ non-Kanye votes, Big Boi is fourth-to-last on this rescoring, with only 38 percent non-Kanye votes; Sufjan Stevens and the Roots were even lower, and Titus Andronicus is in last place by a surprising margin, with only 27 percent non-Kanye votes.
If we also factor out voters who voted for #2 LCD Soundsystem, on the theory that frontman James Murphy and Kanye combine to embody a sort of P&J party line, and again rank the remaining albums by non-Kanye/LCD voter percentage, we get this list, which is starting to be some kind of measure of relative cultness: stuff you might like if you don't like the stuff that the most voters like. Indeed, #1 on this measure is Oregon dark-atmospheric metal band Agalloch, who are arguably about as far from Kanye and LCD as you can reasonably get while still qualifying as either Pazz or Jop.
Alternately, we could factor Kanye in instead of out. This list retabulates the poll using only the 266 people who did vote for Kanye, figuring that anybody who didn't vote for such a clear winner maybe just isn't to be trusted. Sorting again by percentages, we find of course that Kanye is in first, since by definition 100 percent of his votes came from people who voted for him. But reading down the list is pretty intriguing: Curren$y, Drake, Kid Cudi, Rick Ross, Titus Andronicus, Eminem, Das Racist, M.I.A. For all the apparent breadth of Kanye's support, his supporters in aggregate have identifiably genre tastes.
To see this truth in other forms, click on any chart entry, like Arcade Fire or Grinderman, and look at the Empathies list. These lists show album-similarity rankings based on popularity-weighted voter overlap. People who liked that also liked this, in other words. It's just math -- It doesn't know anything about the bands or their styles -- and yet it picks out stylistically related records with fairly uncanny regularity. All you have to do is follow the data and listen.