IT HAS HAPPENED: Update 2014 Thread

[Luke JR68]

If AM would show what albums have best stood the test of time, then what about albums that were used to be seen as they stood the test of the time well, but nowadays less so? Should such albums still be higher than albums that are highly acclaimed, but have only been judged for a couple of years?

I guess what I wanted to say was that which works that will "stand the test of the time" can never be written in stone either.

So if I'm understanding you correctly Henrik, the general opinion of a work changes as time passes, and it is best to use a variety of sources (EOY Lists, Decade Lists, All-Time Lists, etc) to not only show the current opinion, but show how that opinion has evolved over time?

[Henrik]

Wow thanks for all the hard work you do Henrik! This update is fantastic. It might take some getting use to seeing so many ranked songs. Quick question about that though. Since it has been expanded to a top 6,000 songs, it appears more songs may have been added to the bubbling under as well, is this correct?

There are over 9,000 songs on the site and I hope to expand the list to 10,000 songs in 2015 or 2016.

[BleuPanda]

The issue of putting a 5 year limit is it limits diversity, which I don't think a recommendation guide should do. As has been said before, nothing will ever be set in stone. And really, I don't think drastic changes in opinion happen that often, at least in the negative direction. Critics have enough taste to tell when a classic album comes out. The real change over time is when a sleeper hit finally gets the notice it deserves (which has thankfully happened with In the Aeroplane Over the Sea). We're still seeing albums from the 2000s steadily climbing upward, so I think Henrik has found a proper way to rank new stuff without overranking them. We might get a handful of heavy falls now and then, but I think it's worth it for the overall consistency displayed.

[Henry]

Henrik - I very much appreciate the effort and the result.

Thanks for being so patient with all the nit-picking.

I am looking forward to listening to the added songs that I am not familiar with and re-acquainting myself with songs that I barely remember because I do not have them in my i-Tunes library.

[irreduciblekoan]

Ever the diplomat, I find myself in the middle of all the arguments.

I think EOY lists are given a little too much weight, but I still think they should be included. Jimmy, you said that decade lists are good because they take 10 years into account. Someone can likewise say that EOY lists are good (better than reviews or monthly lists) because they take 11-12 months into account. They all have their place in the formula, I just think the formula could be improved even further.

I think songs are given a little too much weight, but I still think they should be a large part of an artist's rankings, and I don't think albums are the be-all-end-all of an artist's value. After all, there were many great musicians before the era of albums and they should be justly recognized. Still, I consider albums the culmination of an artist's talent, like films. So they should be given a bit more weight than songs. Especially if you consider that a great album means having 9+ great songs, whereas a great song is just that, one great song.

Ratings from Metacritic and Allmusic showed that 2012 were on a similar level with other recent years, so I think it makes sense that the 2012 list had a similar distribution to 2011, 2010,... On top of that, "Channel Orange" was unusually superior in the eoy lists for 2012, and therefore it makes sense to me that it topped the 2010s list and made the top 150 of all time in the previous update.

I guess me and Henrik simply have different opinions about what "top 150" truly means. For me, top 150 is canon, it's a huge accomplishment (or should be), a very, very high ranking, and to be in the top 150 would mean you've reached the level of all-time classics of the last 70 years. To you, Henrik, "top 150" isn't such a huge thing, and therefore you don't mind when an album from the previous year, which wasn't on any all-time lists, entered that high. I guess the real question we must answer is, what do the rankings actually mean? What does it "mean" to be the 130th most acclaimed album of all time, or the 400th? Should 130 be held as a lofty ranking for all-time classics, or should we just focus more on yearly rankings and the relative positions between albums within the same era (60s, 80s, 00s, etc)? I have always found that AM is more logical when I just look at yearly and decade rankings, rather than try to make sense about why Channel Orange, from 2012, is already "more acclaimed" than Bob Dylan's trailblazing, revolutionary Freewheelin'.

[Mattceinicram]

The issue of putting a 5 year limit is it limits diversity, which I don't think a recommendation guide should do. As has been said before, nothing will ever be set in stone. And really, I don't think drastic changes in opinion happen that often, at least in the negative direction. Critics have enough taste to tell when a classic album comes out. The real change over time is when a sleeper hit finally gets the notice it deserves (which has thankfully happened with In the Aeroplane Over the Sea). We're still seeing albums from the 2000s steadily climbing upward, so I think Henrik has found a proper way to rank new stuff without overranking them. We might get a handful of heavy falls now and then, but I think it's worth it for the overall consistency displayed.

I'm so glad to see ITAOTS going up in ranking. I always thought it was way too low on this site.

[Bruce]

The issue of putting a 5 year limit is it limits diversity,

When they list the all time leaders in batting average in baseball there is a minimum number of at bats for a player to qualify. Same thing for field goal percentage in basketball and quarterback rating in football.

People could still look at each year's list to get recommendations on more current items, but perhaps the all time list should only contain items that have been around for a certain amount of time.

[Harold]

The way I see it is that either Kanye or his PR team have successfully managed to trick nearly every single music critic at every single major publication in the western world into thinking that he music is good, or those critics are legitimate fans of his work.

And if media attention was such a big factor in his acclaim, why aren't we seeing the same sort of acclaim for Lady Gaga, Katy Perry, Rihanna, Justin Bieber, etc.?

I had to shake my head a little at Jimmy's remark. Let's not forget that West's first two albums (The College Dropout and Late Registration) both topped the Pazz & Jop poll, in consecutive years (by the way, the only other artist to ever accomplish that was The Clash), well before he became the "social media" superstar he is now. (By the end of 2005, he'd made his first real headlines outside of music for his incendiary comments about President Bush, but I doubt that had any impact on LR's acclaim.)

[irreduciblekoan]

I just want to say though that I don't want Henrik to feel like he's being attacked, or that we don't appreciate everything he has done. The last two pages of this thread has become a sort of debate and storm of complaints, but if I'm joining in, it's only because it's something to talk about, and I'm agreeing or disagreeing with certain viewpoints here. But the important thing is, I've been a fan of AM for several years now and it has long been an invaluable source for me. So again, thank you Henrik, and don't take my posts the wrong way. They are simply meant to be constructive, but it isn't to fix what's broken, but to improve what's already great. If you don't take any of my suggestions, I will still be a very happy visitor to this site for years to come.

[BleuPanda]

Well, there's a recognizable zeitgeist in Channel Orange that I think is reflected in its ranking here. Sure, we might not yet have the proof in lists that critics consider it among the all-time greats, but there's enough talk surrounding it that it feels like a huge deal. I don't think it's impossible to estimate an album's all time rank based on extrapolation from EoY lists. And part of the deal of this is, albums like Channel Orange are even bigger to people in high school and college, the ones keeping up with more current music, which includes those who will grow up and become the next batch of music critics. If Channel Orange made this big of an impact with current critics, just wait for those who had "Thinkin Bout You" play at their Senior Prom to start making lists. It's happening with stuff from the 90s and 2000s now, and will happen with the albums we're discussing now in another decade or so. Music has the tendency to get more acclaim from those who grew up with it. The use of EoY lists has been an effective way of predicting the ultimate impact.

This message took way too long to type up...Either way, the main point is I think Henrik is doing a fantastic job with this site, and as much as we want to debate things like EoY lists, the data comparing how the list has evolved over the years shows a generally strong consistency.

[Henrik]

If AM would show what albums have best stood the test of time, then what about albums that were used to be seen as they stood the test of the time well, but nowadays less so? Should such albums still be higher than albums that are highly acclaimed, but have only been judged for a couple of years?

I guess what I wanted to say was that which works that will "stand the test of the time" can never be written in stone either.

So if I'm understanding you correctly Henrik, the general opinion of a work changes as time passes, and it is best to use a variety of sources (EOY Lists, Decade Lists, All-Time Lists, etc) to not only show the current opinion, but show how that opinion has evolved over time?

Well, currently it is only possible to see how the opinion about albums and songs has evolved over time in the spreadsheets. And even those are not truly showing the opinion changes over time, as lists are often included a long time after they were made. In fact, this update included a list from 1956, which is the oldest list in my database!

The weight of a list increases a lot during the first years after an album was released and then continues to increase but not as much (a logarithm function). For an album from 1990, lists from 1996 have higher weights than lists from 1991, and lists from 2001 have somewhat higher weights than lists from 1996. For an album from 1970, lists from 1991, 1996 and 2001 have almost the same weights.

This way, an album can debut very high in the AM all-time list, but if it is not included in lists from the following years, which will have higher weights, the album will quickly drop.

[Henrik]

I also want to add that as long as there are no personal attacks, I enjoy these kind of conversations a lot. This happens to be an area where I have a strong opinion, so I am not likely to change anything. But I have made many changes through the years based on other forum discussions.

[Listyguy]

The issue of putting a 5 year limit is it limits diversity,

When they list the all time leaders in batting average in baseball there is a minimum number of at bats for a player to qualify. Same thing for field goal percentage in basketball and quarterback rating in football.

People could still look at each year's list to get recommendations on more current items, but perhaps the all time list should only contain items that have been around for a certain amount of time.

Albums and songs are more like seasons in a player's career than the careers of players though. Obviously no one would argue that Albert Pujols has had a better career than Mike Trout to date, but Trout is clearly having a better season that Pujols.

[Luke JR68]

Well, currently it is only possible to see how the opinion about albums and songs has evolved over time in the spreadsheets. And even those are not truly showing the opinion changes over time, as lists are often included a long time after they were made. In fact, this update included a list from 1956, which is the oldest list in my database!

The weight of a list increases a lot during the first years after an album was released and then continues to increase but not as much (a logarithm function). For an album from 1990, lists from 1996 have higher weights than lists from 1991, and lists from 2001 have somewhat higher weights than lists from 1996. For an album from 1970, lists from 1991, 1996 and 2001 have almost the same weights.

This way, an album can debut very high in the AM all-time list, but if it is not included in lists from the following years, which will have higher weights, the album will quickly drop.

Oh... For whatever reason it didn't strike me that the weights for lists increased the more recent they are relative to the work itself, thank you for clearing that up Henrik

[Henrik]

And really, I don't think drastic changes in opinion happen that often, at least in the negative direction.

I think this is very true.

The current all-time top 100 albums by decade:

1950s - 1
1960s - 25
1970s - 35
1980s - 18
1990s - 13
2000s - 6
2010s - 2

I will be surprised if this distribution will not even out in the future. I think most of the albums of the 2000s and 2010s in the top 100 will stay there, and they will be joined by other 2000s and 2010s albums. Mostly at the expense of albums from the 1960s and 1970s.

[Bruce]

The issue of putting a 5 year limit is it limits diversity,

When they list the all time leaders in batting average in baseball there is a minimum number of at bats for a player to qualify. Same thing for field goal percentage in basketball and quarterback rating in football.

People could still look at each year's list to get recommendations on more current items, but perhaps the all time list should only contain items that have been around for a certain amount of time.

Albums and songs are more like seasons in a player's career than the careers of players though. Obviously no one would argue that Albert Pujols has had a better career than Mike Trout to date, but Trout is clearly having a better season that Pujols.

Yes, but listing Radiohead among the 5 greatest acts of all time would be like listing Trout among the 10 greatest players of all time already. Radiohead has only made 8 studio albums in 20 years but are ranked ahead of Elvis Presley.

[Listyguy]

When they list the all time leaders in batting average in baseball there is a minimum number of at bats for a player to qualify. Same thing for field goal percentage in basketball and quarterback rating in football.

People could still look at each year's list to get recommendations on more current items, but perhaps the all time list should only contain items that have been around for a certain amount of time.

Albums and songs are more like seasons in a player's career than the careers of players though. Obviously no one would argue that Albert Pujols has had a better career than Mike Trout to date, but Trout is clearly having a better season that Pujols.

Yes, but listing Radiohead among the 5 greatest acts of all time would be like listing Trout among the 10 greatest players of all time already. Radiohead has only made 8 studio albums in 20 years but are ranked ahead of Elvis Presley.

Why is 20 years not enough time? Presley's career lasted only 24 (and he did very little in those final four to improve his musical rank).

[Henrik]

I think Usain Bolt is the greatest athlete of all time. The time that has passed since his world records and the length of his career are irrelevant.

[BleuPanda]

You're too busy looking at numbers and missing the quality. Sure, they've 'only' made 8 albums, but three are in the all time top 100 with another right outside it, with another 2 in the top 500. Meanwhile Presley doesn't have a single album in the top 100 and only 2 in the top 500.

[Henrik]

You're too busy looking at numbers and missing the quality. Sure, they've 'only' made 8 albums, but three are in the all time top 100 with another right outside it, with another 2 in the top 500. Meanwhile Presley doesn't have a single album in the top 100 and only 2 in the top 500.

That's a lot of numbers though.

[BleuPanda]

You're too busy looking at numbers and missing the quality. Sure, they've 'only' made 8 albums, but three are in the all time top 100 with another right outside it, with another 2 in the top 500. Meanwhile Presley doesn't have a single album in the top 100 and only 2 in the top 500.

That's a lot of numbers though.

I should have probably worded that better. How about "looking at quantity over quality"?

[Listyguy]

You're too busy looking at numbers and missing the quality. Sure, they've 'only' made 8 albums, but three are in the all time top 100 with another right outside it, with another 2 in the top 500. Meanwhile Presley doesn't have a single album in the top 100 and only 2 in the top 500.

Presley was never really an albums guy though. Albums didn't become the popular form until after his peak. He was the master of singles (3 in the top 100, 4 more in the top 200).

[Bruce]

Why is 20 years not enough time? Presley's career lasted only 24 (and he did very little in those final four to improve his musical rank).

It's not a question of how much time, it's a question of how much quality material did the act release.

An act that has only released less than 100 studio songs should not be ranked among the 5 greatest acts ever. They are just not prolific enough.

[Bruce]

I think Usain Bolt is the greatest athlete of all time. The time that has passed since his world records and the length of his career are irrelevant.

Then why do you see the time that has passed in evaluating records as relevant?

[Bruce]

I think Usain Bolt is the greatest athlete of all time. The time that has passed since his world records and the length of his career are irrelevant.

Bolt is only great at one thing, sprinting. Why would you see him as a better athlete than guys like Jim Thorpe and Jesse Owens, and Carl Lewis, who were great sprinters but were also great at other things, like the long jump and strength events?

[Rob]

How do you guys go about exploring the new updates?

I always go through them from bottom to top (it's more exiting than the other way around), song by song and album by album. Slowly. This usually takes a few days, but this time I had the afternoon off and spend till now to finish the whole thing in one sitting. Somewhat addictive, these updates. Luckily we only get them once a year. Many thanks for the hard work, Henrik!

Interesting discussions here, some reflecting some observations I had, because it was hardly missable that especially in the albums department (and the artist ranking, but I take these with a grain of salt) the update favored new works heavily, although that might have something to do with the fact that this year just had many lists with more recent work (the fingerprints of NME can't be missed). Especially in the top 1000 almost everything that went up was from at least the nineties and mostly from the 21st century, while albums from previous years were largely the ones taking the hit. It's an interesting shift and I'm on two minds of it. Better let it sink in first. Only the slaughter of jazz is a bit of a downer, but it's not really anybody's fault that lists of last year overlooked them.

Mostly I'm happy with it though. Glad to see the expanded list of songs that makes room for left field entries (Man in the Long Black Coat or Isis by Bob Dylan for example) and it's always nice to see favorites gaining more ground. My number 1 song, The Boxer by Simon and Garfunkel finally made the top 500. On to the top 100. Other jumps that come to mind that I loved where the great leap of Baba O' Riley, Video Games and U2's One into the top 100. And Tom Wait's Tom Traubert'S Blues is also a very welcome new addition. In the album department I was glad to see my favorite artist of the last ten years, Laura Marling, gaining not one, not two, but three albums. Even if they're not particularly high (and I guess, never will be), it's great to have them. And Morricone's holy score for Once Upon a Time in the West too.

Lots to explore. Again, thanks!

[Bruce]

I think Usain Bolt is the greatest athlete of all time. The time that has passed since his world records and the length of his career are irrelevant.

Let's see if he can do it again like Carl Lewis did over several olympics.

World records get broken because conditions and equiptment and diets and training methods continue to make it easier over the years. If you go strictly on times then Jesse Owens would not even make his college team nowadays. Nobody believes that.

Besides, if you want to go with the best current athlete as the greatest of all time you'd be more accurate if you chose the best decathlon athlete rather than the best sprinter.

[BleuPanda]

Why are we discussing sports players in a music thread? We get it, Bruce. You like consistency in acts over extended periods of time. Other people who have experienced the impact of a single outstanding work feel otherwise. This is beating a dead horse.

[Bruce]

Other people who have experienced the impact of a single outstanding work feel otherwise.

Are you saying that I have never experienced the impact of a single outstanding work?

[Listyguy]

the update favored new works heavily, although that might have something to do with the fact that this year just had many lists with more recent work (the fingerprints of NME can't be missed).

I was noticing this as well. Almost every song that wasn't well ranked in 2009 (top 1000-ish, I'm going through them for bracketology) is a 2010-2013 song or was on the NME list.

[Kingoftonga]

How do you guys go about exploring the new updates?

I always go through them from bottom to top (it's more exiting than the other way around), song by song and album by album.

I admire your restraint. I usually start with #1 and work my way down. I generally explore the albums first, and I find the Top 500 particularly interesting (though, admittedly, its a pretty arbitrary criterion). I also check in on personal favorites when I remember (Manuel Gottsching's E2-E4 is back in the Top 3000! At #2853, but at least he's back in!). It's also a good time to make note of albums I've been meaning to listen to; ones with large jumps mean I'll probably put them higher on my "To Listen" list.

Then I check out the songs, and I have to update my iTunes playlists to reflect the update. (I keep a running playlist of every AM song I have in the proper order). This might take longer than usual this time around, but I'm not complaining.

[Bruce]

By the way, occasionally the site turns me on to a good record I did not know before. This just happened with "Easy Snappin'" by Theo Beckford from 1959. It clearly took its beat from "No More Doggin'" by Rosco Gordon.

Speaking of records that had a huge influence on reggae, I'm shocked that "Be My Guest" by Fats Domino is not listed on the site. It's generally considered the most important record ever in the development of the reggae beat.

[Listyguy]

How is "St. Elmo's Fire" not ranked, or even a bubbler? That song is great, easily the best of Eno's career.

[BleuPanda]

How is "St. Elmo's Fire" not ranked, or even a bubbler? That song is great, easily the best of Eno's career.

Yeah, with such an expansive list, it's really interesting to find the big songs still missing. As I said in the other topic, I'm surprised "Oh Comely" by Neutral Milk Hotel hasn't popped up yet.

[Listyguy]

How is "St. Elmo's Fire" not ranked, or even a bubbler? That song is great, easily the best of Eno's career.

Yeah, with such an expansive list, it's really interesting to find the big songs still missing. As I said in the other topic, I'm surprised "Oh Comely" by Neutral Milk Hotel hasn't popped up yet.

Agreed. It's definitely my favorite song from "In the Aeroplane Over the Sea".

[JR]

I know the artist lists aren't to be "taken too seriously," but...

I am surprised to see Britney Spears ranked higher than Janet Jackson (even though just six positions separate them). Spears has two tracks ranked much higher than Jackson's highest-ranking one, but the latter has four albums in the top 3,000, while Spears has none (and Jackson has more tracks in the top 6,000 than Spears).

[GucciLittlePiggy]

I still firmly feel he is simply a big act of the moment who will drop once he produces more lackluster work, and the media's love affair with him comes to an end, as I don't care what anyone says, that is a clearly major reason he receives such acclaim currently

Am I misunderstanding something? Are you saying that publications are rating his albums more highly simply because he's a big figure in the media? I mean, MBDTF and Yeezus are my 2nd and 5th favorite albums of all time, respectively, and it's not because the media told me they were great.

[Jirin]

I think critics rate albums higher when they are more saturated in the media, and are more comfortable placing albums higher when they perceive them as being more recognized, yes. Can't speak for your own personal motives. All the critics who rated Kanye high probably genuinely like him, they're not lying about their tastes or are just fully manipulated by the hype. But there are other artists they enjoy just as much but rate lower because they perceive them as being greater risks, or feel less confident in their own feelings because they don't see other people talking about them. That's just diffusion of responsibility to the group, and it's the marketing and hype that created the inflated perception of group opinion.

Let's take a look at X's album Pazz N Jop ratings versus AM all time ratings.

Los Angeles: EOY #26, currently #429
Wild Gift: EOY #2, currently #643
Under The Black Sun: EOY #10, currently #2307
More Fun In The New World: EOY #4, currently bubbling under

Los Angeles was the lowest ranked of X's first four albums on Pazz n Jop but now it's by far their most praised album. It's not because they enjoyed Los Angeles any less in 1980 than they do today, but it's because they felt less comfortable rating it high in 1980 because nobody was talking about X yet.

As for including EOY lists, I'm comfortable including them just because I know the more time passes, the less they count. If I understand Henrik's formula correctly, how long after the album was released the list came out is taken into account in weighting the lists.

[DocBrown]

Much interesting discussion, with much predictable bluster from the usual suspects. I will add my voice (it's the bullfrog in the corner) to the chorus of thanks, Henrik, with the added proviso that I think the results are perfect!

Let me defend that indefensible position.

Whether Steely Dan drops (boo!) or Nirvana rises (yay!) has nothing to do with the inherent value of the music or your personal taste. The music is just as great today as it was on Saturday, and deserves the same spot in your personal pantheon. Henrik has simply added more data to the dataset (and perhaps tweaked the algorithm in mysterious ways) resulting in changes, mostly incremental but occasionally massive (boo, Jazz). It's nothing personal, and doesn't affect the war in Gaza one bit.

And personally I believe that some music being made today is every bit as good as music made 20, 50 or 100 years ago. The genius of AM is that Henrik compares critics voices to place that very recent music, as best as can be statistically determined, in its correct relative position to the music of other eras. If you believe that all music of the past 35 years is unlistenable, you will never be happy with that approach, regardless of what percentage of the music in the list is "new" or "old". But new and old are moving targets, and as has been pointed out before, changing tastes will often relegate former supergroups to the dustbins of musical history. The example of Chicago was a great one... beloved in 1974, (when I saw them live and owned most of their discography,) ridiculed by the mid-80's and largely ignored today. Changing tastes account for much of that, but I think it might also be that their music went down a path that went nowhere and they have not "stood the test of time", despite still being active after a 45 year career. Has any group of the past twenty years cited Chicago as a primary influence?

On the other hand bands like Nirvana, or Jimi Hendrix, who produced only a few albums each, continue to be massively influential twenty and forty years later. If they continue to influence musicians, and critics, today, then they should continue to climb the charts.

Debating whether Janet Jackson should be rated ahead of Britney Spears (obviously, duh) or which ITAOTS song is the greatest is irrelevant. This statistical model produces these results; as Henrik explains in the preface the lists are not definitive and never can be. By adding a massive amount of new data, Henrik has plugged some holes, which for fans of the still-missing songs, artists and albums will make their exclusion all the more glaring. That gives you something to anticipate for the next go round.

[Jonathon]

How is "St. Elmo's Fire" not ranked, or even a bubbler? That song is great, easily the best of Eno's career.

I thought the same thing when I looked at the list.

As for the "Don't take the artists list seriously" argument, isn't that ultimately the point of the site? To tabulate the greatest artists of all time, based on the strength of their albums and songs?

It's really hard to say either way how it should work. If you weigh albums too heavily, Elvis Presley, Chuck Berry, Little Richard, Robert Johnson, Hank Williams Jr, and countless other A-list legends suffer. If you weigh the songs too heavily, Tom Waits, Joni Mitchell, Wilco, Captain Beefheart, Frank Zappa, and Steely Dan suffer. In a perfect world, there would be a formula that adheres to both strengths equally, without penalizing for lacking in the other. I just don't think such a formula exists.

[audioclectic]

This statistical model produces these results; as Henrik explains in the preface the lists are not definitive and never can be.

Some time ago, Bruce objected to the inherent legitimacy of music industry critics (or criticism / critique) and suggested that the opinions of studied individuals might be as relevant. That notion - at the time, part of a contentious dialogue (surprise) - was quickly slapped down by Henrik.

I've read many critical posts in many industry websites since finding AM in 2010. I agree with Bruce. So what?

This statistical model, as well as it can be designed, is only a reflection of an aggregate bias. As time goes on, bias changes. Those are facts, and the facts neither validate nor invalidate Henrik's algorithm. At the same time, the factual bias and Henrik's algorithm do not invalidate the considered opinions of those who take exception.

We all have different aesthetic values leading to judgements. Henrik knows this. I have never seen him suggest otherwise. If any of us are really aggrieved we can start our own website, or we can go away, or suffer through our objections. And then again, we can focus on the positives, and those are not in short supply.

[Bruce]

This statistical model produces these results; as Henrik explains in the preface the lists are not definitive and never can be.

Some time ago, Bruce objected to the inherent legitimacy of music industry critics (or criticism / critique) and suggested that the opinions of studied individuals might be as relevant. That notion - at the time, part of a contentious dialogue (surprise) - was quickly slapped down by Henrik.

I've read many critical posts in many industry websites since finding AM in 2010. I agree with Bruce. So what?

This statistical model, as well as it can be designed, is only a reflection of an aggregate bias. As time goes on, bias changes. Those are facts, and the facts neither validate nor invalidate Henrik's algorithm. At the same time, the factual bias and Henrik's algorithm do not invalidate the considered opinions of those who take exception.

We all have different aesthetic values leading to judgements. Henrik knows this. I have never seen him suggest otherwise. If any of us are really aggrieved we can start our own website, or we can go away, or suffer through our objections. And then again, we can focus on the positives, and those are not in short supply.

Great post.

Rather than starting my own website I have published my feelings on the rankings of songs and artists at Digital Dream Door where I edit many lists.

[Henry]

This statistical model produces these results; as Henrik explains in the preface the lists are not definitive and never can be.

Some time ago, Bruce objected to the inherent legitimacy of music industry critics (or criticism / critique) and suggested that the opinions of studied individuals might be as relevant. That notion - at the time, part of a contentious dialogue (surprise) - was quickly slapped down by Henrik.

I've read many critical posts in many industry websites since finding AM in 2010. I agree with Bruce. So what?

This statistical model, as well as it can be designed, is only a reflection of an aggregate bias. As time goes on, bias changes. Those are facts, and the facts neither validate nor invalidate Henrik's algorithm. At the same time, the factual bias and Henrik's algorithm do not invalidate the considered opinions of those who take exception.

We all have different aesthetic values leading to judgements. Henrik knows this. I have never seen him suggest otherwise. If any of us are really aggrieved we can start our own website, or we can go away, or suffer through our objections. And then again, we can focus on the positives, and those are not in short supply.

Does anyone believe that bias can be removed from musical rankings? Do those who have studied music academically or otherwise claim to have minimized their biases? Is there a model that has been subject to peer review that demonstrated how to minimize bias in this endeavor? If not, I do not yet understand why opinions by studied individuals would be particularly helpful. Even if objective measures could be established they would likely be for limited parameters. Perhaps I am misunderstanding the value of opinions of studied individuals as suggested by Bruce and supported by audioclectic. I have no objection to including them in the algorithm, but do not think that their absence is costing us much.

[GucciLittlePiggy]

I think critics rate albums higher when they are more saturated in the media, and are more comfortable placing albums higher when they perceive them as being more recognized, yes. Can't speak for your own personal motives. All the critics who rated Kanye high probably genuinely like him, they're not lying about their tastes or are just fully manipulated by the hype. But there are other artists they enjoy just as much but rate lower because they perceive them as being greater risks, or feel less confident in their own feelings because they don't see other people talking about them. That's just diffusion of responsibility to the group, and it's the marketing and hype that created the inflated perception of group opinion.

I guess I understand what you're saying. I haven't really analyzed trends in acclaim over time as you did with X, but I see what you're getting at. I've probably seen these trends myself, but I never really noticed or put much thought into it. I think most publications, though, have the integrity to give the most honest rankings for their lists. Larger market publications are probably the most likely to succumb to the supposed "hype" surrounding an artist. I can't tell you how much it annoys me when Rolling Stone will fail to rank an album by an up-and-coming artist on their end of year list, but then will give high praise to the same artist for their next album even if it's inferior work. Even still, I can't imagine Kanye dropping too far in critical acclaim several years down the road, but maybe I'm just a bit biased.

[BleuPanda]

I forgot to lament the fact that "Gangnam Style" is now ranked on acclaimed music.

[Nassim]

I think critics rate albums higher when they are more saturated in the media, and are more comfortable placing albums higher when they perceive them as being more recognized, yes. Can't speak for your own personal motives. All the critics who rated Kanye high probably genuinely like him, they're not lying about their tastes or are just fully manipulated by the hype. But there are other artists they enjoy just as much but rate lower because they perceive them as being greater risks, or feel less confident in their own feelings because they don't see other people talking about them. That's just diffusion of responsibility to the group, and it's the marketing and hype that created the inflated perception of group opinion.

I think you have a point but that it's incomplete.
First I don't see why that would apply only to recent albums, a magazine is more likely to put Pet Sounds or Revolver in top of a 60s list than Odessey and Oracle or Os Mutantes. They will most likely put Blonde on Blonde way ahead of John Wesley Harding.
I also think it goes further than "being comfortable because someone else validated it", you have a lot of other factors than peer approval :
First, visibility. For what I know wasn't the issue for Los Angeles just the fact that less people had heard it ? Odessey and Oracle, Spiderland or In The Aeroplane Over the Sea took time to get acclaim because they were not well known (whether it is for disbanding just after, no promotional support, bad timings or any reason). For instance, the French magazine "Les Inrocks" put Funeral on their EOY list two years in a row : in 2004 it was barely released, not many critics had heard it, it landed around #30 I think, in 2005 it was album of the year : partly because now all the critics knew it, and partly because Les Inrocks like to vote for the most "trendy" album of the year. Much like your EOY or decade polls, lots of magazine and websites lists are made by polling through their team. The most an album is known, the most it is likely to appear in lots of lists, especially in the bottom of lists, this actually really shows in OUR polls too !

Second : selective listening : you don't know any album of the 2000s, you stumble upon AM, which album are you the most likely to listen first, Funeral or Bows + Arrows ? You do know a bit your 60s album and dislike both The Velvet Underground & Nico and Cosmic Tones for Mental Therapy, which one are you most likely to give a 2nd chance to ? It's the same for critics : if all your colleagues are raving about an album, it's more likely you give it more spins than something you never hear being talked about, so sure if you are from Rolling Stone it's more likely you gave a few chances to Yeezus than to a Shabazz Palace album.
This strangely seems to be less obvious than I think, there was a lengthy article somewhere (popmatters ?) trying to analyze the results of the Pitchfork Reader Polls (for albums between 96 and 2011) and never did the writer go for the obvious one : those are Pitchfork Readers, the albums they have listened to are most likely to be informed by the website, so it's even more likely that the albums a lot of them have given time to are some that Pitchfork like ; the same goes for critics, it's more likely that the albums that the most critics have listened to are either those from household names or those with rave reviews. That's a bigger problem in EOY when some critics might not have heard enough albums to have a full list of definite favorites and will fill the bottom with whatever acclaimed stuff they heard, but it also impacts all-time lists where some albums are more likely to have been heard by everyone than others... I'm sure if all the critics in the world had heard Perfect From Now On, it would be top 500 of all time

Last things is, unfortunately, editorial line. LesInrocks put on the cover of their EOY special magazine the band/artist who ended first, so it has to be something that appeals to their readers : cue to MGMT toping the list or The Black Keys #1 with an album not even released yet. Pitchfork needs something critically revered but still not too straight forward, cue Low topping the 70s. Fact needs to fill its lists with stuff you don't know, NME must put something trendy and cool etc... I don't think the lists are heavily influenced by editorial decisions, and maybe all those examples are just the real results of the poll they did internally (you pick your writers based on taste similarity), but I would be surprised if there never had been a few switches to make the results more satisfying to a list's core target.

[Nassim]

Why is 20 years not enough time? Presley's career lasted only 24 (and he did very little in those final four to improve his musical rank).

It's not a question of how much time, it's a question of how much quality material did the act release.

An act that has only released less than 100 studio songs should not be ranked among the 5 greatest acts ever. They are just not prolific enough.

So why should we in your opinion usually only care about the singles, the songs that people know and paid for, because if it wasn't successful it's not relevant, and now suddenly we should care for over 100 songs by artists, including B-sides and deep cuts ?
How many songs did Buddy Holly record by the way ?
I'm still baffled by how often you can say the exact opposite of something you previously stated just because it validates the opinion you're currently defending... have you thought about politics as a career ?

[Rob]

How is "St. Elmo's Fire" not ranked, or even a bubbler? That song is great, easily the best of Eno's career.

Yeah, with such an expansive list, it's really interesting to find the big songs still missing. As I said in the other topic, I'm surprised "Oh Comely" by Neutral Milk Hotel hasn't popped up yet.

The song who's absence keeps baffling me the most Child in Time by Deep Purple. Maybe it's only huge in The Netherlands (though the seventies poll on AM suggests otherwise), but I actually grew up believing that was one of the indisputable classics, on par with things like Stairway to Heaven, Hotel California or Imagine. Yet critics seem to completely ignore it, even though it's by a famous band and from a famous album. When I saw it was going to be included this update I thought it would certainly be placed in at least the top 6000 songs, but no: it's still bubbling under. What gives?

It's actually funny to look at the bubblers. When the list of inclusions was revealed I remarked that most songs I thought were missing would finally make the list. Now it seems really a lot of them are just bubblers. Oh well, at least they are visible on the site now.

On another note, I want to comment a little on the discussion about whether or not albums should be given more weight in the artist ranking or not. One argument in defense of higher album weighing I have not come across here, but that I always thought was the reason of giving them more recognition is simply that an album is just a bigger work. An album is a collection of multiple songs, usually round 8 to 12 or something. A song is just, well, one song. So to make a great album you need to make many great songs and as a extra challenge these songs have to cohere in one way or another. So if songs and albums would be rated the same, you're basically saying that making one great song is as big an achievement as making about ten great songs. That's why I think it doesn't hurt to be a little more biased in the rankings towards albums, even though I admit it isn't fair for acts from before the album era. Nothing personal against singles artists too, because I admire them. I also think Henrik went a long way to balancing things more this time.

One thing that bothers me about the artist ranking though is what I hereby dub the Pharrell syndrome. This means that apparently an artist can rise high coasting on the works of others. I've noticed this before, but it got worse this time around, with Pharrell Williams suddenly coming out of nothing and taking the 365th place in the artist rank! This is based almost entirely on the guest work he did on songs by other artists. He has two songs of his own included, but they are placed at #3933 and #3998, so they probably wouldn't have influenced his position to much. I'm not saying it is completely unfair that he got points for Get Lucky or Blurred Lines, but I wonder is reduced points for featuring is in order. Admittedly, that might not be fair either, because Jay-Z for example is credited as "featuring" on Crazy in Love and Umbrella, but was also a mayor creative force on both songs. On the other hand, John Lennon was a mayor force in the creation of Fame by Bowie and also contributed vocals on it, but wasn't credited so the song doesn't appear on his page. Same goes for Sting on Dire Straits' Money for Nothing and no doubt others. The point is that how guest work is credited can significantly change an artists position in the artist list and I'm not sure if that is fair. However, I don't really have a solution for it and perhaps it's best to take that into account when saying you should take the artist ranking with a grain of salt.

[Nassim]

The song who's absence keeps baffling me the most Child in Time by Deep Purple. Maybe it's only huge in The Netherlands (though the seventies poll on AM suggests otherwise), but I actually grew up believing that was one of the indisputable classics, on par with things like Stairway to Heaven, Hotel California or Imagine. Yet critics seem to completely ignore it, even though it's by a famous band and from a famous album. When I saw it was going to be included this update I thought it would certainly be placed in at least the top 6000 songs, but no: it's still bubbling under. What gives?

Could be a bit of national biased, I always thought "The Partisan" was the most well known and acclaimed Leonard Cohen song before arriving here, be it for its relative popularity in France or the number of covers it has.

One thing that bothers me about the artist ranking though is what I hereby dub the Pharrell syndrome. This means that apparently an artist can rise high coasting on the works of others. I've noticed this before, but it got worse this time around, with Pharrell Williams suddenly coming out of nothing and taking the 365th place in the artist rank! This is based almost entirely on the guest work he did on songs by other artists. He has two songs of his own included, but they are placed at #3933 and #3998, so they probably wouldn't have influenced his position to much. I'm not saying it is completely unfair that he got points for Get Lucky or Blurred Lines, but I wonder is reduced points for featuring is in order. Admittedly, that might not be fair either, because Jay-Z for example is credited as "featuring" on Crazy in Love and Umbrella, but was also a mayor creative force on both songs. On the other hand, John Lennon was a mayor force in the creation of Fame by Bowie and also contributed vocals on it, but wasn't credited so the song doesn't appear on his page. Same goes for Sting on Dire Straits' Money for Nothing and no doubt others. The point is that how guest work is credited can significantly change an artists position in the artist list and I'm not sure if that is fair. However, I don't really have a solution for it and perhaps it's best to take that into account when saying you should take the artist ranking with a grain of salt.

To be fair Pharrell doesn't get any points for the N*E*R*D* ranked songs and albums. Also he actually wrote Drop It Like It's Hot, he is credited as a featuring but he is the main creator of the song. And he doesn't get credited for other ranked songs he wrote or produced like Grindin', Hollaback Girl, Hot in Herre, Milkshake, Rock Your Body...
And while he might not have done much for Get Lucky (he is still credited as a writer), I would be really surprised if he wasn't the main creative force for Blurred Lines.

But anyway, it would be interesting to see who gets the biggest bumps if all their projects were put together (after the Fab Four obviously). Would Clapton be able to top Damon Albarn ? (Hopefully not) How high would Eno jump ?
And it would also be interesting to see who is the most acclaimed producer, non performing songwriter, studio musician... I think someone did something a bit like that a long time ago in the forum...

[audioclectic]

Does anyone believe that bias can be removed from musical rankings? Do those who have studied music academically or otherwise claim to have minimized their biases? Is there a model that has been subject to peer review that demonstrated how to minimize bias in this endeavor? If not, I do not yet understand why opinions by studied individuals would be particularly helpful. Even if objective measures could be established they would likely be for limited parameters. Perhaps I am misunderstanding the value of opinions of studied individuals as suggested by Bruce and supported by audioclectic. I have no objection to including them in the algorithm, but do not think that their absence is costing us much.

Hi Henry. I can't speak for Bruce, but I'm not advocating the incluson of individual opinions in the data set. I'm just pointing out that it's futile to be frustrated over the bias. At the same time, I would prefer to stop short of venerating the critics. Opinions are opinions. We agree or disagree. No harm, no foul. If we can't get beyond it, find another amusement.