Do albums tend to get worse towards the end?

[sp4cetiger]

Do you find that your favorite tracks on an album tend to be towards the beginning or the end? I know artists have a tendency to start albums with bangers, but what about the rest?

I wanted to try to get some actual numbers for this, so I compiled the ratings for individual tracks on all the albums in the BEA top 500 (with a script, not by hand). Anyway, I got the following results:

Thumbnail. Click to enlarge.

Basically, this just shows how highly people tend to rate tracks as the album goes on. As you can see, people tend to rate the beginning of the album higher than the end -- by about 2 points, on average. This effect isn't just restricted to the first track, tracks continue to get ratings higher than the album mean (horizontal dashed line) for the first third of the album. I thought maybe the last track on the album would get higher ratings than the average, but no. There is a jump in rating towards the end, but it's still below the album mean.

So what does this mean? Do artists tend to throw their best tracks at the beginning of an album or does listener fatigue cause people underrate tracks later in an album? Maybe a little of both?

[besteveruser]

Dunno. Some combination of attention spans just not wanting to listen intently for a full fifty minutes straight and the artist's and label's inclination to give the audience the hits they want first. It's been working pretty well for music for a while

But that chart there looks like a perfect visualization of my feelings about many albums, even a lot of the albums I like. That line graph is the shape of my love for music.

And I gotta say I've always preferred songs to albums. It's hard to find a great album. Finding a great song is easy. Just gotta YouTube or iTunes sample that stuff. Kinda want a site called besteversongs tbh.

[Defago]

What's the variance? Run a significance test to see if it's not just randomness. 500 albums is also not such a large sample compared to the whole database. Do you even statistics.

[benpaco]

Actually, I tend to like albums better towards the end. In fact, if you split albums in my top 10 simply in half, I only like the first halves of Unknown Pleasures and The White Album better than their second halves. Everything else I prefer the second half of. I'd go more into depth but I'm running around a bit, I'll check for my top 100 later
_________________

. . . 2016 . . . 2015 . . .

Things I Make

[sp4cetiger]

[CA Dreamin]

20 standard deviations is no accident, so you've def proven Side As are higher rated than Side Bs. But to answer your question about why, I'd say it is a combination of artists front-loading their albums and listener fatigue, but more so on the artists. Yes, listener fatigue can make one overlook a late-album gem, but fans of BEA's top 500 have listened to every song on them, probably several times (or so we'd like to think). Musicians obviously want people to keep listening to the end, so putting the better, catchier songs earlier makes perfect sense. Later album tracks are slower and tend to be more forgettable. How interested would a listener be if those kinds of songs were in the beginning of an album? This isn't to say those kinds of songs aren't good. 'Porcelain' on Californication, '4th of July' on Superunknown, and many others are terrific late-album tunes. However, besteveruser is right; Great songs are way more common than great albums. Only the best artists can do them. Most artists, however, can't pull it off; and for them the best way to keep people listening is put their best efforts first.

[denmarkman]

Out of boredom, I've gone through my whole top 10 and determined whether I liked the first or second half of each better.

First half:
If You're Feeling Sinister
Swordfishtrombones
Unknown Pleasures
Homogenic

Second half:
Metal Box
Pink Moon
Pinkerton
Perfect from Now On
Enter the Wu-Tang (36 Chambers)

Literally can't even decide:
Surfer Rosa

So based on my top 10 at least I don't see that as much. However, I think that's more because I'm not particularly attracted to albums that are super front-loaded.

I actually have a few thoughts on this, but I have to go pick up my dad and anyway I'm high so I'll give a more in-depth response later.

[Defago]

Yeah I made a mistake on a division. The question now though is of the simultaneity: does being first cause people to like songs more, or do artists put their best songs first? Personally, I think it's both.

[benpaco]

Right so what I'm finding going over my chart is, with a few exceptions, the albums which produced at least 1 big hit have better first sides, whereas the albums which didn't have a pop hit or maybe work only as an album (see Nattvittje) have better second sides. So I did some searching because I got curious.

Of the best selling 9 singles of all time, 1 is a double single and 2 were never released on studio albums, so we'll discount those. But of the 6 remaining best selling singles ever all are on the first half of the album except for Celine Dion's "My Heart Will Go On".

Digging deeper (as that was barely scratching the surface), I decided to add in all singles which have sold at least 10 million copies. I found that, of the 32 singles to sell at least 10 million copies, only "My Heart Will Go On", "I Will Survive", "The Last Farewell", "I'm a Believer" and "Whiter Shade of Pale" appear on the second half of an LP. That's a measly 16%.

Finally I decided to look at more recent examples. I decided to look at the highest selling 10 iTunes songs of 2013, as what could be a more cliched version of the present state of music? Well, the only three songs which are on the second half of their LPs are "When I Was Your Man", "Stay", and "Mirrors". 30% but still low. If you follow this pattern, it seems to continue (I can go only off brief glances and prior knowledge here, I really am prepping for a week trip). My thoughts are one of 3 things:

1 Marketing. If you have a song as the first track on an album, it's the first a person hears. They associate that as being the song of the album then, especially when the best song is placed first and it's blared on the radio nonstop. It's the "best" so it must be the best.

2 Lack of attention span. This is more sales than quality, but sales at least partially dictate what we end up hearing. I have friends who I don't think have ever finished a full album. They'll start one and just get bored and leave around track 4 or 5. Therefore, it's the first half of the album that they remember, as it's the only bit they could remember.

3 Marketing based on lack of attention span. You want your album to sell, right? If "albums are dead", as some of claimed, and people really are abandoning albums before they're done, you want to make sure they hear the best stuff first, the guff can come after. So you order the album in such a way that guff will come later so that the album, or more specifically the singles off the first half of the album, sell well.

Anyone else have thoughts on this?
_________________

. . . 2016 . . . 2015 . . .

Things I Make

[satiemaniac]

benpaco, I really appreciate your legwork and the amount of info you stuffed into that post. I'll respond in full to it after a good night's sleep 'cause you deserve some discussion generation.

Another facet I'm interested in throwing in for discussion is how BEA users (since they are the sample we're looking at) rate "songs." I (and this is the RYM in me talking) rate things as "releases." If an artist has chosen to release a single, I will listen to it as a whole (hip hop and dance singles with multiple remixes/instrumentals/vocal tracks depending on context and genre sometimes notwithstanding but always getting a listen through each individual track, sometimes spaced out so I don't hear two very similar songs and have them bleed together) and rate it as such. Same for an EP, same for an album, same for a compilation. The result, for example, is that THEORETICALLY, I could see Pole's 1 as better or worse than 2 and 3 and as better or worse than the compilation that has all three as one entity. I like the compilation enough as a full experience (not taken in one sitting but as a tryptych of complementary works intentionally released as one retroactively) to have it as my second favorite album of all time, and tend not to be too interested in parsing apart if there is a superior "part" of it. Now, to apply this logic to an album, the "parts" that make it up to me are almost irrelevant. Unless I am listening to a pop album, I enjoy things as unified wholes (even if this tends to be a marketing schtick or plain revisionism in some cases - more on that when I'm articulating using a brain that isn't running on Warheads candies, McDonald's fries, too much soda, and at 4:13am). Thus, rating individual songs requires me to temporarily lift something out of a context, and that seems counter to my way of appreciating music.

So my question is, do you "experience" an album and then go back and hold it under a microscope? Do you do that the first time around? What is your reason for doing this? Is this a rockist (popular music as a whole-ist) approach? Would you pick apart classical works and highlight crescendos (or whatever your favorite part of your favorite symphony or whatever is) as "better" parts? What about jazz improvisations? Are specific solos in improvisations reaching past a certain minute mark something you would rate independently of the flow of rhythm that segues between them? What about folk music (not acoustic popular music, but in the musicological sense of the term)? Non-European classical music like drone-filled Hindustani and Carnatic classical music? What about field recordings (be they sonic landscapes crafted from miscellany of a trip or "day in the life" slabs or ethnographic materials as assembled by Folkways, for example)? Just to belabor this last question, there is a Folkways release that documents the Head Start program for children in Mississippi. Some individual tracks are a bit dull and serve more of a historical than musical utility, but cutting out one part would negate the purpose of the album. Sure, as a listening experience, the few wonderful group performances of major spirituals would be more interesting, but without this detritus (kind of a harsh word and not really the one I'm looking for but eh) and other assorted material, I wouldn't be faced with the historical element that Folkways is getting across. There are other ethnomusicological compilations from them and others that serve as dual history lessons and listening excursions, such as their recordings of American Indian music for pleasure and sacred/ritual musics or the A Casa Edison samba compilation I'm working through currently that spends a disc essentially providing (in my mind as a listener, to a fault, but in my mind as a historically interested and persuaded person as fascinating and very necessary, as well as showing admirable striving for completeness) context for the later recordings that start to evolve more into what we now recognize as samba music.

Sorry for all this food for thought, and I know it's largely just taken an interesting topic and turned it, potentially, into a wide variety of tangents. To answer the OP, for me, if the album is really good and something I consider pleasing, I tend to be enthralled from front to back. As I said before, only pop records ever have filler in my mind. It's why I feel no shame in skipping around The Beatles' early albums as I did as I was listening through them yesterday (or, technically, the other day). But if we're talking rock and pop music specifically, a lot of them have the potential and the real expression of uninspired second sides. I'll leave off on one historical example that's interesting if you don't mind checking out RYM release history. This album by Docteur Nico was originally released with a certain tracklist, and the reissue flipped the A and B side. I would love to see a double blind test (yay, more statistics!) of whether or not respondents (who of course were already at least interested in if not fluent in the idiom of Soukous music) who listen(ed) to one version or the other first preferred it typically to those who listen(ed) to the other first.

Blah, I need to sleep. Sorry for rambling, y'all.