The Meaning of 6.8
YACHT's new record, I Believe In You, Your Magic Is Real, was rated a 6.8 on Pitchfork in a review that was as much a reminder about the 2002 release by Max Tundra than the 2007 release by YACHT. Obviously I think the record is better than 6.8, but I'm admittedly biased.
Rating systems are designed to give an overview of the quality of a work. Sometimes it's as simple as a thumb up or down, and sometimes it's a complicated rating of every aspect of the work. Regardless, you rarely get an opportunity to know what the reviewer likes and dislikes (unless you do a lot of homework), which is an important aspect of a recommendation. The specific reviewer is typically a small byline at the bottom bottom of the review. It wasn't Rob Mitchum that gave YACHT a 6.8, it was "Pitchfork".
Music reviews (perhaps more than other artistic works) are subjective. One person's perfect 10 might be a harder-to-please critic's 7.9. I know I can ignore Cousin Bill's rant against Justin Timberlake because I know he only likes bluegrass music, and likewise I can recommended the latest Bobby Birdman to Aunt Sue because I've seen her music collection. What does a 6.8 really mean? I want that same context for Pitchfork, and looking at the history of reviews, I can sort of get it.
Method: Finding the reviews via the Pitchfork site proved tedious, so I used Google and searched for "record review" "rob mitchum" site:pitchforkmedia.com and went through the first 10 pages collecting 93 reviews.
The average score is 6.6. This is slightly higher than the May 2006 6.45 site average and a teeny bit less than the 6.7 site average from October 2003, but that seems to follow the trend of the whole site. Looking at Mitchum's reviews year by year there is a significant drop in scores between 2004 and 2005, leading me to wonder, did something happen to change Rob's perspective on music?
2002: 6.82003: 6.8 (site average 6.7)
2004: 7.0
2005: 6.4
2006: 6.5 (site average 6.45)
2007: 6.4
When I graphed all the scores out, I saw that the lows seem to have stopped about a year ago. May 2005 he gave a 0.4 to Weezer and has rated nothing lower than 4.4 since then (The 4.4 was an A.I.H. remix album). Was there a backlash from Weezer fans that scared him into not ripping too hard on an album? Or perhaps Mitchum just likes most of the music he hears?
Rob Mitchum doesn't seem to review much terrible music. Only six albums were lower than a 4, while 31 albums were between 7 and 7.9. In fact, almost half the albums (48) were between 7 and 8.9. Odd.
0.0 to 0.9: 11.0 to 1.9: 1
2: 1
3: 3
4: 9
5: 14
6: 14
7: 31
8: 17
9: 2
Perfect 10: 0
All of this makes me very interested in Rob Mitchum. Googling him is difficult since he has written so much for Pitchfork, but perhaps he is the same Rob Mitchum who attended the University of Michigan and is (or was) a big fan of Phish in 1997? If so then it makes sense that he would be the former rhythm guitarist and vocalist for a college jam band called Hypersonic in 1997. There is another article from 1998, so maybe he graduated that year? Reviewers should be public figures. The more I know about Mitchum the more I would understand the ratings.
I looked at his ratings by month, to see if maybe sunny summer weather would boost his appreciation of music. It doesn't appear to be seasonal, but there is an increase in scores for the second half of the year.
Jan: 6.9Feb: 6.0
Mar: 6.7
Apr: 6.4
May: 6.0
Jun: 7.6
Jul: 6.1
Aug: 7.4
Sep: 7.1
Oct: 6.8
Nov: 7.2
Dec: 7.4*
*Only one album from December, Sondre Lerche. Faces Down.
Conclusions
1. I'm certainly not the first person to analyze Pitchfork reviews, and the results seem to be surprisingly the same (Mid-level average scores, very few hated albums). If nothing else, Rob Mitchum is as consistent as Pitchfork itself.
2. Prediction: I'd guess Rob Mitchum's next review is going to be a 6.3.
3. Though it's not the most fair way to use the numbers, it's interesting to me how the numbers allow comparison between artists. For example, YACHT's latest is better than albums by Weezer, Moby, The Postal Service, The Who, Nine Inch Nails, The Killers, Of Montreal, Mogwai, and Beck.
4. The 0.0 to 10.0 rating system is dumb. UrbanHonking believes in the simple beauty of the Netflix 5 point system (Hate, Didn't Like, Like, Really Like, Love) and I find the idea that an album is .3 points better than another very silly. It implies a technical and scientific application of what is only ever a subjective system.
5. I recommend buying YACHT's album. (Marriage Recs Shop)
<< | Posted by kmikeym at 5:32 PM | >>
Here is what I've heard about Pitchfork. Admittedly, I am not sure if this is actually true or not, but this is what I've heard: A reviewer writes the review. The rating is then given by Ryan Schreiber, who is the cheif editor of Pitchfork. And, according to my source, he pretty much only gets a boner over the Wrens.
This explains why a band can get a really great review with a poor rating and vice versa.
excellent analysis!!!!!!!!
This may be simply more subjective analysis, but I have sometimes wondered about the seasons when albums are released/reviewed. It seems to me like late summer/early fall is a good time, as people are ready for something new, fresh, etc. for their back-to-school.
Just a thought.
What is the point of a 0.0 to 10.0 scale anyway? If you need the extra digits, why not just use 0-100?
This is an awesome entry.
I second Carlos and Lucie's enthusiasm for your research!
Just a few hours ago I was thinking there should be a database of critics. We need a brief biography, how long they've been wherever they are, where to find what they've published and whether they've ever done anything interesting. I like the idea of somehow charting all their responses to things.
That led to thinking about a kind of cultural survey experiment:
One could build a database of all the things that had ever been critiqued (perhaps limited to specific set of critics: let’s say just reviews in the WWeek, Mercury, Seattle Weekly, The Stranger, since 1995 for example). It would be interesting to map the positive/negative/descriptive responses across the population of critics to see if there were specific camps or 'clumps' of anomalous aesthetic response. Then to identify the specific works that served as the best statistical indicators for determining structural correlations between the content and the various “species” of critics represented by each clump. These indicator works might be our “classics”. As we started to list and analyze the thematic and qualitative content of these indicators, I wonder if it would be possible to build strong theories to predict which specific characteristics determined how the various species of critics responded to work.
I was also musing about a 'Critique the critique' journal.
And then I found your remarkable analysis. Uncanny.
Ritchey might write her dissertation on 20th century music criticism.
Metacritic is a pretty good "db for critics" or at least a collection of them. Erik turned me on to the idea awhile back of using it to find the critics that seem to have your vibe, and then paying attention to those critics to try to find new stuff that you would like.