What's the Pitchfork distribution? You've probably seen this chart:
[INSERT CHART OF THE PITCHFORK NEW REVIEW HISTOGRAM]
Interesting. Says some things. But I think it's incomplete.
Classic reviews are much higher. Obviously there's selection bias here. But they're like way higher. Like take 1994. There were 8 10s that year! We would never see 8 10s in a new release year.
Pitchfork never updates their reviews (well, they did this one time, but it's rare). But they do create charts. And opinions change a lot from their reviews to their charts. For example, SZA's SOS was their #1 album of 2023 but only got an 8.7. Solange's A Seat at the Table was #1 of 2016 — also an 8.7. So what would the distribution actually look like if they updated their reviews?
And besides, how does an end of year chart position translate to a review value? What does #5 of 2017 mean out of 10? Seems like it's probably higher than 8.8, which is what Melodrama originally got that year. Right?
This is abstractly interesting, but it's also necessary if you're trying to construct a retrospective all-time chart from Pitchfork data. Ranking all the albums they've ever reviewed! Somehow, we have to compact all their information into a number out of 10.
How would you go about doing this?
One way would be to take a decade where they have given retrospective reviews and compare it to their charts that year.
And in fact, they've given lots of retrospective reviews. 588 in the 90s alone. You could make a whole chart out of that! So what do they say?
It turns out, yeah, reviews would be a lot higher.
Look at this table of Pitchfork 10s. And this isn't even done! They haven't reviewed Daydream Nation or Closer — #7 and #12, respectively, on their 1980s list.
Retrospective 10s by year:
| Year | 10s | Year | 10s | Year | 10s | Year | 10s | Year | 10s | Year | 10s | Year | 10s | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1960 | 1 | 1970 | 7 | 1980 | 3 | 1990 | 3 | 2000 | 2 | 2010 | 1 | 2020 | 1 | ||||||
| 1961 | 1 | 1971 | 5 | 1981 | 3 | 1991 | 4 | 2001 | 0 | 2011 | 1 | 2021 | 0 | ||||||
| 1962 | 0 | 1972 | 5 | 1982 | 2 | 1992 | 1 | 2002 | 2 | 2012 | 1 | 2022 | 0 | ||||||
| 1963 | 4 | 1973 | 2 | 1983 | 2 | 1993 | 4 | 2003 | 0 | 2013 | 0 | 2023 | 0 | ||||||
| 1964 | 2 | 1974 | 0 | 1984 | 4 | 1994 | 8 | 2004 | 0 | 2014 | 0 | 2024 | 0 | ||||||
| 1965 | 5 | 1975 | 5 | 1985 | 4 | 1995 | 3 | 2005 | 0 | 2015 | 2 | 2025 | 0 | ||||||
| 1966 | 5 | 1976 | 4 | 1986 | 3 | 1996 | 5 | 2006 | 1 | ||||||||||
| 1967 | 4 | 1977 | 8 | 1987 | 4 | 1997 | 7 | 2007 | 0 | ||||||||||
| 1968 | 7 | 1978 | 3 | 1988 | 5 | 1998 | 5 | 2008 | 1 | ||||||||||
| 1969 | 7 | 1979 | 5 | 1989 | 6 | 1999 | 0 | 2009 | 0 |
Compare that to their contemporary 10s:
Contemporary 10s by year:
| Year | 10s | Year | 10s | Year | 10s | ||
|---|---|---|---|---|---|---|---|
| 1999 | 1 | 2008 | 0 | 2017 | 0 | ||
| 2000 | 1 | 2009 | 1 | 2018 | 1 | ||
| 2001 | 1 | 2010 | 2 | 2019 | 1 | ||
| 2002 | 0 | 2011 | 0 | 2020 | 1 | ||
| 2003 | 0 | 2012 | 0 | 2021 | 0 | ||
| 2004 | 1 | 2013 | 0 | 2022 | 0 | ||
| 2005 | 1 | 2014 | 0 | 2023 | 0 | ||
| 2006 | 0 | 2015 | 1 | 2024 | 0 | ||
| 2007 | 0 | 2016 | 0 | 2025 | 0 |
So how do we create this distribution?
Well, a good place to start, I think, would be the retrospective reviews. For example, if the 90s had an average of 4 10s per year, then we should have at least that many in our true distribution.
So why does this matter? Well, it doesn't really. Music is obviously subjective. But I do think we all have a sort of subconscious rewriting of history. We think Liquid Swords or the Blue Album were always 10s. But we would never give ourselves permission to say 2024 has 6 10s (which it might have, in retrospect!).
The Biggest Climbers
A cool offshoot of this is that we can track the large differences between the original score that Pitchfork gave an album and the revealed score in the final chart position. Here's a table of the biggest climbers:
| Album | Artist | Year | Original Score | Chart Position | Implied Score | Δ |
|---|---|---|---|---|---|---|
| A Seat at the Table | Solange | 2016 | 8.7 | #1 of 2016 | ~9.8 | +1.1 |
| SOS | SZA | 2022 | 8.7 | #1 of 2022 | ~9.8 | +1.1 |
| Melodrama | Lorde | 2017 | 8.8 | #5 of 2017 | ~9.5 | +0.7 |
[TODO: Fill in more examples]
These are albums where, over time, the critical consensus shifted dramatically upward from the initial review. It's a fascinating window into how opinions evolve — and how hesitant reviewers can be to commit to a high score in real-time.