Thursday, February 11, 2016

Who are the contrarians?

Iyas Daghlas


As much as people would like it to be true, it turns out that Pitchfork isn’t the worst offender when it comes to being a blatant contrarian.


I looked at the top 100 rated albums on Metacritic (by their aggregated, weighted Metascore from Metacritic). These are albums that have been rated very highly by several musical publications; there has been a near consensus amongst the critics that the album is very strong. For many of these records, however, there are some detractors that give significantly lower scores. I then created a list of reviews that gave these critically acclaimed albums a score of 60/100* or less


As it turns out, the most frequent** contrarians are The Guardian (6/48) and Rolling Stone (5/48). Q Magazine had a notable 3/48, and every other publication was either listed once (Pitchfork was listed once for its review of PJ Harvey's Stories From the City, Stories From the Sea) or twice in the database.You’ll find links below to the Guardian and Rolling Stone reviews. If you’re going to pick one to read, pick The Guardian’s Kanye West review.


The Guardian:


The Roots - Undun


Kanye West - My Beautiful Dark Twisted Fantasy


Bring Me the Horizon - That’s the Spirit


My Morning Jacket - Z


Swans - To Be Kind


St. Vincent - St. Vincent


Rolling Stone:


Deafheaven - Sunbather


QOTSA- Songs for the Deaf


Spoon - Kill the Moonlight


Outkast - Speakerboxxx/The Love Below


The Magnetic Fields - 69 Love Songs


Hopefully I can piece together some of the trends in these contrarian reviews and make that into a blog post in the upcoming months.


Methods:
Used a web-scraper to grab all scores from top 100 albums reviewed on Metacritic, then subsetted the dataset to only include reviews scoring 60/100 or lower. Album reviews that were not available online or did not have a score were not included. For the latter, Metacritic’s staff reads the album review and arrives at a consensus about what the score sounds like it should be.

* An assumption of this analysis is that album scores mean similar things for the different musical publications (see my previous post about the distribution of Pitchfork reviews). My intuition tells me this isn't an unreasonable assumption given that the left skewed distribution centered towards the right of the review scale is something you also see in distributions of IMDB's TV shows (see previous post; also see the following link I just found that coincidentally does a very similar analysis to the one I did on Pitchfork's reviews: [1] ).


**Based on this sample of 100 albums. In the future I might look at the top 1000 and see if this trend holds.

No comments:

Post a Comment