Encyclopaedia Metallum: The Metal Archives

Message board

* FAQ    * Search   * Register   * Login 



Reply to topic  
Author Message Previous topic | Next topic
adroship
Mallcore Kid

Joined: Sun Jan 26, 2014 5:46 pm
Posts: 10
Location: United States
PostPosted: Wed Aug 20, 2014 2:17 am 
 

So, I started working on a side project, and I thought if there was anywhere to share the results, it was here. I did a quick comparison of how many bands were in the archives for each country (per billion people in that country) and that country's nominal GDP/Capita. The results were a little predictable and hardly rigorous Image

but still a little interesting, with some surprises (what's up with Bolivia?). Here is the same data in a graph: Image

Basically, props to Finland.

The red dashed line is the mean correlation. I chose the scale of bands/billion people simply because that worked best for the coloring function on the map. All info for population and GDP came from Wolfram Mathematica, except for Curacao which came from Wikipedia. White on the map means I couldn't find it. I did not label every point for obvious reasons, but I'm happy to give info for any country or point on the graph. The UK is basically right on top of the US.

There is a lovely image for metal bands per capita floating around, but as far as I can find this is the first graph that shows GDP as well. If people ask, I may make something similar for the states (Update: Included several posts down and it's much more interesting!). That being said, this was made in two nights and I'm sure someone with graphic design experience could make something more impressive. Enjoy!


Last edited by adroship on Tue Aug 26, 2014 12:04 am, edited 1 time in total.
Top
 Profile  
LegendMaker
Metalhead

Joined: Mon Jan 18, 2010 11:24 am
Posts: 1603
Location: France
PostPosted: Wed Aug 20, 2014 2:54 am 
 

This could be interesting trivia to look up. Unfortunately, the images you posted are tiny, tiny thumbnails that can't be read; could you link to bigger versions?
_________________
Opus wrote:
If you're going to sell out, make sure you'll get paid at least.

Top
 Profile  
MutantClannfear
Metalhead

Joined: Thu May 27, 2010 12:12 am
Posts: 2237
Location: United States
PostPosted: Wed Aug 20, 2014 3:12 am 
 

The data points in the second graph could stand to be less confusing as well. I'd suggest removing any points for countries which you don't specifically have labelled; it's really hard to tell which bullet corresponds to which country, the farther left and downwards you go.

That being said, it's a pretty interesting concept to look at. I'd be interested in seeing a similar one for the United States in particular.

Top
 Profile  
adroship
Mallcore Kid

Joined: Sun Jan 26, 2014 5:46 pm
Posts: 10
Location: United States
PostPosted: Wed Aug 20, 2014 3:21 am 
 

LegendMaker wrote:
This could be interesting trivia to look up. Unfortunately, the images you posted are tiny, tiny thumbnails that can't be read; could you link to bigger versions?


Huh. No idea why that is happening, they are reasonably sized when I look at them on computers other than my own. Either way, here are links to pretty large versions:

Map: http://postimg.org/image/vzdy8cttp/
Graph: http://postimg.org/image/bhmpo0971/

MutantClannfear wrote:
The data points in the second graph could stand to be less confusing as well. I'd suggest removing any points for countries which you don't specifically have labelled; it's really hard to tell which bullet corresponds to which country, the farther left and downwards you go.


Hmm... I mostly labeled to give examples of dense areas or highlight especially weird behavior. I'll mess around and see if I can find something more aesthetically pleasing without sacrificing data points.

Top
 Profile  
teh_Foxx0rz
Metal newbie

Joined: Tue May 20, 2014 9:38 am
Posts: 293
Location: United Kingdom
PostPosted: Wed Aug 20, 2014 6:52 am 
 

adroship wrote:
MutantClannfear wrote:
The data points in the second graph could stand to be less confusing as well. I'd suggest removing any points for countries which you don't specifically have labelled; it's really hard to tell which bullet corresponds to which country, the farther left and downwards you go.

Hmm... I mostly labeled to give examples of dense areas or highlight especially weird behavior. I'll mess around and see if I can find something more aesthetically pleasing without sacrificing data points.

Put the points you've labelled in different colours.

Top
 Profile  
joppek
Metalhead

Joined: Sun Jan 09, 2011 7:36 am
Posts: 873
Location: Suomi Finland Perkele
PostPosted: Wed Aug 20, 2014 7:35 am 
 

and maybe make another graph where you discard edge cases and only include data points between the origin and denmark (so it's zoomed in there)
_________________
All the best bands are affiliated with Satan. -Bart Simpson

Top
 Profile  
Rompestromper
Metal newbie

Joined: Mon Jun 23, 2014 2:37 pm
Posts: 99
Location: Netherlands
PostPosted: Wed Aug 20, 2014 10:07 am 
 

Nice statistical work man, sounds like something I would do for fun as well!
One question however, did you use all bands randomly or did you only use the active bands because that might give a somewhat different result I think.

Top
 Profile  
narsilianshard
Metalhead

Joined: Fri Aug 14, 2009 12:22 pm
Posts: 621
Location: Seattle
PostPosted: Wed Aug 20, 2014 12:02 pm 
 

Rompe has a good point. Do the years you used for GDP match the "activity" of the of bands? Currently active bands vs 2013 GDP would really be the only correct way to compare.
_________________
Haven't written shit about metal recently, so here's some stupid gifs I made years ago.

Lastfm

Top
 Profile  
adroship
Mallcore Kid

Joined: Sun Jan 26, 2014 5:46 pm
Posts: 10
Location: United States
PostPosted: Wed Aug 20, 2014 2:23 pm 
 

I appreciate the advice everyone.
teh_Foxx0rz wrote:
Put the points you've labelled in different colours.

Done
joppek wrote:
and maybe make another graph where you discard edge cases and only include data points between the origin and denmark (so it's zoomed in there)

and done.

I made a zoomed in version that is hopefully not too cluttered Image
http://postimg.org/image/rksnndoh1/

And then I read Rompe and narsilian's comments. It's a good point. Everything I have put up so far has been the total number of bands, whether active or not. So I went back through the archives, and found the number of active bands for each country. Sometimes it had no effect, sometimes it had a pretty dramatic effect. Might be worth looking into a different project to see how the spread of metal bands changes over time. Anyways, I have updated the globe now so that it only shows active bands:
Image
http://postimg.org/image/5lapw617j/

If you compare that with the original map, it seems like there is slightly less focus in the Scandinavian countries (Finland still dominates) and a bigger focus on South America. The large scale graph didn't change significantly, but here is the zoomed in version only looking at active bands:
Image
http://postimg.org/image/sgne806sd/

I am surprised Japan is so low in both versions. Anyways, I made a list of the top ten countries that deviate from the average in terms of active bands. All of them are above the line. In order from largest deviance to least: Bolivia, Macedonia, Chile, Finland, Paraguay, Maldives, Malta, Costa Rica, Bosnia and Herzegovina, Serbia.

The ten that deviated the least, from least deviance to most: UK, Mongolia, Andorra, France, US, Nepal, Venezuela, Lebanon, Cuba, Armenia.

I'm trying to present this without too much analysis (I have zero experience in anthropology, sociology, geography, etc.) but I wonder if this is an accurate way of demonstrating how much a country "cares" about metal. I wonder if it says something that Macedonia, a country with significantly smaller GDP, puts out a comparable number of metal bands per person to the US, UK, and France. Or how about the fact that Australia has significantly more bands per person despite a similar GDP to those three?

Top
 Profile  
Anal Enforcer
Mallcore Kid

Joined: Tue Aug 12, 2014 3:04 pm
Posts: 19
Location: United States
PostPosted: Wed Aug 20, 2014 3:00 pm 
 

I love this kind of stuff! Nice job adro.

Top
 Profile  
ObservationSlave
Metal newbie

Joined: Fri Nov 05, 2010 6:27 pm
Posts: 388
Location: United States
PostPosted: Wed Aug 20, 2014 5:55 pm 
 

Japan doesn't really surprise me because it is the tenth most populous country and I don't commonly stumble upon Japanese bands. I am aware of quite a few, but not nearly as much as the US, Canada, Australia, and most of Europe and South America.

I have always been interested in what MA uses to define countries. For instance, places like Puerto Rico, Gibraltar and Greenland are generally not considered countries, while Kosovo is, but bands from Kosovo are listed under Serbia.

Top
 Profile  
Rompestromper
Metal newbie

Joined: Mon Jun 23, 2014 2:37 pm
Posts: 99
Location: Netherlands
PostPosted: Thu Aug 21, 2014 4:01 am 
 

Hey man that is awesome, cool that there are some countries that have some different contributions in the last years compared to before!

Top
 Profile  
joppek
Metalhead

Joined: Sun Jan 09, 2011 7:36 am
Posts: 873
Location: Suomi Finland Perkele
PostPosted: Thu Aug 21, 2014 4:29 am 
 

do you have an un-zoomed version of the plot graph with only active bands counting?
_________________
All the best bands are affiliated with Satan. -Bart Simpson

Top
 Profile  
adroship
Mallcore Kid

Joined: Sun Jan 26, 2014 5:46 pm
Posts: 10
Location: United States
PostPosted: Thu Aug 21, 2014 4:30 pm 
 

joppek wrote:
do you have an un-zoomed version of the plot graph with only active bands counting?


Yep. Image

http://postimg.org/image/4riv117sj/

Like I said, the changes are not that dramatic from the original. Some of the Scandinavian countries move around a fair amount.

Top
 Profile  
Scorntyrant
Metalhead

Joined: Mon Nov 15, 2004 5:55 am
Posts: 1081
PostPosted: Thu Aug 21, 2014 8:53 pm 
 

It would be interesting to run he same analysis based on population percentage who speak English and compare the two.
_________________
J_Ason wrote:

Were it up to me, I'd pick apart your nervous system atom by atom and throw each individual particle into a separate black hole so that you may never think stupid shit like that ever again.

Top
 Profile  
rafa_hell
Metal newbie

Joined: Mon Jul 06, 2009 12:55 pm
Posts: 35
Location: Brazil
PostPosted: Thu Aug 21, 2014 10:24 pm 
 

Really interesting. it has potential. I could help you with this one if I wasn't too busy, keep on it.

Top
 Profile  
joppek
Metalhead

Joined: Sun Jan 09, 2011 7:36 am
Posts: 873
Location: Suomi Finland Perkele
PostPosted: Fri Aug 22, 2014 3:30 am 
 

that pretty much straight up column of iceland, sweden, faroe islands and finland is interesting... what is it about that point in gdp that makes a spike

btw, does the plot include countries with zero metal bands?
_________________
All the best bands are affiliated with Satan. -Bart Simpson

Top
 Profile  
yentass
Metalhead

Joined: Sun Nov 23, 2003 9:28 am
Posts: 903
Location: Israel
PostPosted: Fri Aug 22, 2014 1:41 pm 
 

adroship wrote:
I'm trying to present this without too much analysis (I have zero experience in anthropology, sociology, geography, etc.) but I wonder if this is an accurate way of demonstrating how much a country "cares" about metal. I wonder if it says something that Macedonia, a country with significantly smaller GDP, puts out a comparable number of metal bands per person to the US, UK, and France. Or how about the fact that Australia has significantly more bands per person despite a similar GDP to those three?

Well, as the adage says, correlation doesn't imply causation, so probably tying the data to how much a country cares about metal might be futile. Looking on the data on a case by case basis could be quite fascinating though (for example, whether the dominance of the Scandinavian counties is influence by the fact that metal is more mainstream there or not, or, say, whether the influx of Macedonian bands compared to their GDP has something to do with the tumultuous history of the region).
_________________
Voidal, Doom/Death Metulz.
kingnuuuur wrote:
DoomMetalAlchemist wrote:
I know nothing of hair care, so bare with me.

Metal dudes, assemble in the shower!

Top
 Profile  
AYearInExile
Mallcore Kid

Joined: Sun Jun 23, 2013 5:47 pm
Posts: 21
Location: Estonia
PostPosted: Sat Aug 23, 2014 10:38 am 
 

Props to the author of this thread. You had a very interesting idea and actually did something about it.

I can offer one interpretation (based on R. Inglehart World Values Survey). The two factors that you looked at are in causal relation. Economic welfare causes more metal music and proliferation of every other kind of cultural form because people in rich countries generally have a more creative outlook on life. Prosperity gives people time, makes people more expressive and independent and that leads some of them to metal. Poverty causes concerns over survival and affection to religious thought, both of which make metal a less viable option.

I'm aware that it doesn't explain every country on the graph but it explains the mean correlation marked with the red and green lines.

Top
 Profile  
FantomLord17
Metal newbie

Joined: Sun Jan 09, 2011 3:41 pm
Posts: 90
Location: Chile
PostPosted: Sat Aug 23, 2014 2:56 pm 
 

yentass wrote:
adroship wrote:
I'm trying to present this without too much analysis (I have zero experience in anthropology, sociology, geography, etc.) but I wonder if this is an accurate way of demonstrating how much a country "cares" about metal. I wonder if it says something that Macedonia, a country with significantly smaller GDP, puts out a comparable number of metal bands per person to the US, UK, and France. Or how about the fact that Australia has significantly more bands per person despite a similar GDP to those three?

Well, as the adage says, correlation doesn't imply causation, so probably tying the data to how much a country cares about metal might be futile. Looking on the data on a case by case basis could be quite fascinating though (for example, whether the dominance of the Scandinavian counties is influence by the fact that metal is more mainstream there or not, or, say, whether the influx of Macedonian bands compared to their GDP has something to do with the tumultuous history of the region).


It is also important how would you define "caring for metal". As I understand, in fields such as sociology the way you define the construct you are researching and the indicators you use are decisive in the quality of an investigation. I'd say that "caring" for metal would require more indicators such as attendance to metal shows (which would prove difficult if you consider more "underground" shows) and record sales. And of course, it is evident from looking at the graphic that GDP doesn't explain all the variance in number of metal bands, and so other variables would be necessary, as yentass said.

It would be interesting to hear more hypothesis for the most curious cases, maybe a more cultural/political/psychological one to complement the one mentioned by AYearInExile. I'll give it a shot: Chile has its own tumultuous history with Pinochet's dictatorship, and our culture is very open towards accepting products from first world countries in order to live in the illusion of Chile being a first world country. As such, the metal genre has gained quite a wide following for its ambivalent position in culture: aggressive (and as such: cathartic) but not as politically involved as punk -which makes it perfect for the "apolitical" youth-; rebellious and dirty but still technical and musical enough to attract more introverted folks; fun when coupled with alcohol, offering an alternative for people who dislike our more "traditional" party music (cumbia), and so on and so on.

This is a very cool thing you've done adroship. Cheers :metal:
_________________
Last.fm profile
My Best of 2013 list.

Top
 Profile  
Yayattasa
Metalhead

Joined: Wed Dec 12, 2012 7:49 am
Posts: 552
Location: Brazil
PostPosted: Sat Aug 23, 2014 3:22 pm 
 

I think it would be better if instead of bands you counted only musicians (but I think this information is not readily available in the archives).

Also, try log scales in the chart, in either or both axis.
_________________
inhumanist wrote:
Arkhane wrote:
Damn, I thought this thread was headed for closure. Good save, whoever saved it but I'm too lazy to scroll up right now.

oh my god people disagreed on something for several pages stop the presses

Top
 Profile  
vulcan plutarchy
Metal newbie

Joined: Thu Oct 25, 2012 10:36 am
Posts: 108
Location: United States
PostPosted: Sat Aug 23, 2014 4:51 pm 
 

It is unclear what the colors on the map indicate.
_________________
pghedujobs - teaching jobs pittsburgh

Top
 Profile  
kalervon
Metal newbie

Joined: Sun Sep 23, 2012 10:43 pm
Posts: 283
Location: Canada
PostPosted: Sun Aug 24, 2014 9:27 pm 
 

It would be tricly to count "bands X years active" or select only bands that are currently active. Some bands were formed in the 90s then disbanded and now play a live gig every five years and their status is "active". I think the current metrics are as good as one can make them. Kudos !

Top
 Profile  
elf48687789
Metalhead

Joined: Mon Feb 02, 2009 2:03 pm
Posts: 1626
PostPosted: Mon Aug 25, 2014 5:00 am 
 

So basically, what you're saying is that the Scandinavian countries have an enormous amount of bands, irrespective of income, while Eastern Europe and Portugal and Latin America have a lot of bands even if the income is lower than other places?

Top
 Profile  
kalervon
Metal newbie

Joined: Sun Sep 23, 2012 10:43 pm
Posts: 283
Location: Canada
PostPosted: Mon Aug 25, 2014 9:04 pm 
 

I think it shows that there is no clear relationship between # of bands in a country (or per capita) and GDP. Basically, metal does not necessary emerge in countries that are more wealthy or less wealthy.

The working hypothesis could have been: metal is born out of misery; countries where living is harsh spawn angry people who need to let their anger out through various ways, including forming a metal band

Or: metal can only prosper in countries that are well off and where kids have the luxury to afford buying instruments and don't need to otherwise struggle for existence

None hypotheses are supported by the data. It's a case by case basis probably.. something more complex

Though I'm tempted to say there is a correlation between latitude north and the density of metal bands (per capita) :)

Top
 Profile  
adroship
Mallcore Kid

Joined: Sun Jan 26, 2014 5:46 pm
Posts: 10
Location: United States
PostPosted: Tue Aug 26, 2014 12:02 am 
 

joppek wrote:
btw, does the plot include countries with zero metal bands?

Yes.
Scorntyrant wrote:
It would be interesting to run he same analysis based on population percentage who speak English and compare the two.

I like the idea of that, I may tackle it soon. It seems to be a little tricky to find accurate reports of the percentage of English speakers in each country, but I'll look around a little more.

Inspired by the kind words from everyone, I decided to tackle the US states as well. First, thanks to this thread on stackoverflow: http://stackoverflow.com/questions/8957 ... merica-map

from which I got the shapes of the states. Secondly, my source for GSP (I've seen it as GDP for states some places, I think it is worth differentiating Gross State Product) was this site: http://www.usgovernmentrevenue.com/comp ... e_2013mZ0a

I used the info from 2013 even though the active band numbers are from the present. I can't imagine this will skew the data one way or another unless some state had a major surge of metal bands in the past year. Anyways, on to the exciting stuff.

Geographically, I don't see anything too alarming, other than the fact that Oregon has a remarkably high bands to GSP ratio while Maine has a terrible one. Image
http://postimg.org/image/dkdvrdpr7/

Btw, that does not map DC, but I did include it in my analysis and it surprisingly had a low ratio of bands to GSP. Update: I made a stupid. The corrected version of these graphs are a few posts below. I am going to keep up my embarrassing excitement as a personal reminder of my temporary idiocy, and to make it clear how much other people have helped me. Basically, ignore everything else in this post. But, the actual plot is in my opinion much more interesting:
Image
http://postimg.org/image/j44v8j673/

Look at how strongly the two correlate with each other! The dashed green line is actually a linear fit (-22.6398 + 0.707463 x). It is just as convincing zoomed in:
Image
http://postimg.org/image/c6lroo7ff/

I completely agree with Kalveron when he says:
kalervon wrote:
I think it shows that there is no clear relationship between # of bands in a country (or per capita) and GDP. Basically, metal does not necessary emerge in countries that are more wealthy or less wealthy.

In fact, the r squared value for a linear fit of the world map was 27.2%, which is terrible. But for the US, the r squared value for the fit I show is 92.5% which is amazing! Admittedly this is a small sample size of only 51, but I am shocked to get results as good as these. By the way, I am going to call reverse causation and suggest that a large metal scene leads to greater economic prosperity ;)

The ten states that deviated from the model the most were, in order,

Texas (less), California (more), Oregon (more), New York (less), Georgia (less), Washington (more), Florida (less), Illinois (more), Wisconsin (more), Louisiana (less).

The ten states that deviated from the model the least were, in order,

Kansas (less), Minnesota (less), Kentucky (less), Alaska (less), Tennessee (less), South Dakota (more), North Dakota (more), Michigan (less), Wyoming (more), Utah (more).

Hope you all enjoyed this one too. I am thinking of working on the English speaking population map now, or, what I really want to try, a map of the most popular (defined by most number of bands in the archive) genre in each country. This gets tricky because I need to decide what counts as separate genres (sludge/doom, thrash/speed?), and I know whatever I choose will make someone furious. I'd love to hear if you have suggestions.


Last edited by adroship on Tue Aug 26, 2014 1:55 pm, edited 1 time in total.
Top
 Profile  
joppek
Metalhead

Joined: Sun Jan 09, 2011 7:36 am
Posts: 873
Location: Suomi Finland Perkele
PostPosted: Tue Aug 26, 2014 3:54 am 
 

you didn't take population in to account with the states?
_________________
All the best bands are affiliated with Satan. -Bart Simpson

Top
 Profile  
kalervon
Metal newbie

Joined: Sun Sep 23, 2012 10:43 pm
Posts: 283
Location: Canada
PostPosted: Tue Aug 26, 2014 6:20 am 
 

Same question as joppek..

Basically, what we may be looking at on the almost linear plot could be the simple result of a high {population,GSP} correlation

Top
 Profile  
Rompestromper
Metal newbie

Joined: Mon Jun 23, 2014 2:37 pm
Posts: 99
Location: Netherlands
PostPosted: Tue Aug 26, 2014 8:59 am 
 

Well in this case the population is the number of metal bands compared to the GDP

And even though there might be some trends we all known that a GDP is not a function of only one factor but it is still nice to plot it in this way, especially if someone is telling you that metal heads are poor scum

Top
 Profile  
adroship
Mallcore Kid

Joined: Sun Jan 26, 2014 5:46 pm
Posts: 10
Location: United States
PostPosted: Tue Aug 26, 2014 1:52 pm 
 

kalervon wrote:
Same question as joppek..
Basically, what we may be looking at on the almost linear plot could be the simple result of a high {population,GSP} correlation


I have no excuse for thinking last night that dividing by populations wouldn't affect the correlation. I feel rather silly, so I am just going to post the corrected versions and quietly excuse myself. Populations were taken from wikipedia. Since the map of the US shows the ratio of the two variables, it is unaffected. Here is the full version

Image
http://postimg.org/image/p7hinr2nj/

And a zoomed in version
Image
http://postimg.org/image/3zd0xt2y7/

Obviously, no significant correlation. In fact, the r squared value (8%) is not only horrific but worse than in the case for the world. Sorry for the unnecessary excitement everyone.

The ten states that deviated from the model the most were, in order,

Oregon (more), Rhode Island (more), Maine (less), Hawaii (less), Massachusetts (more), Washington (more), Mississippi (less), California (more), Iowa (less), Alabama (less).

The ten states that deviated from the model the least were, in order,

Maryland (more), Utah (more), West Virginia (less), Michigan (less), Wyoming (less), Idaho (less), Vermont (more), Tennessee (less), Nevada (more), Delaware (more).

Top
 Profile  
Display posts from previous:  Sort by  
Reply to topic


Who is online

Users browsing this forum: Dudemanguy, Intraum, SkullFracturingNightmare, Von Jugel and 19 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum

Search for:
Jump to:  

Back to the Encyclopaedia Metallum


Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group