File talk:World GDP Capita 1-2003 A.D.png

Flawed world regions

The author of this horrible graph provides a source for the information regarding the global economy from 1AD to the 19th century, which is this book here, but it still merits close critical analysis. There's already a lengthy discussion below questioning the validity of the numbers in this graph, so I'm just going to challenge the flawed regional groupings that the author breaks the world down to.

These 7 geographic regions (Africa, Asia, Former USSR, Eastern Europe, Western Europe, Latin America, and Western Ofshots [sic]) are horribly flawed in that they force a modern-day categorization of world regions to different eras when such regional groupings would not have made any sense. For example, the notions of a "Western Europe" and "Eastern Europe" didn't exist in 1AD, when you would have actually had a Southern Greco-Roman Europe (and its sphere of influence) versus a northern/central/eastern "barbarian" Europe.

Additionally, just as these regional groupings are horribly modern-centric, they're also horribly anglocentric. "Western Offshoots" can be understood to refer to any country or region whose culture is derived either entirely, or to a large extent, from Europe. Brazil, Uruguay, and Argentina would fit this description, but because there's already a "Latin America", we're left to assume that "Western Offshoots" refers only to the Anglo societies outside Europe: the USA, Canada, Australia, and New Zealand. This is a flawed sociocultural paradigm that mindlessly repeats a common mistake frequently made in American public discourse: using the term "Western" to refer exclusively to the English-speaking world and -variably- parts of Europe deemed to be similar.

Aside from the fact that Argentina today is ethnically and culturally closer to Canada (let alone Spain and Italy) than to Guatemala, the geo-cultural regions of "Latin America" and "Western Offshoots" wouldn't have even made any sense before the 16th century.

"Asia" is another horribly concocted region on this graph, made up of vastly different civilizational regions from Mesopotamia and Persia to China and Japan. While Europe and its post-1500 sphere of influence are divided into 5 regions (Eastern Europe, Western Europe, Former USSR, Latin America, Western Offshoots), civilizations as vastly different as the Persians and Japanese are forced together into "Asia", rather than grouping them into more sensible categories, such as Middle East, South Asia, Southeast Asia, East Asia, Central Asia. Hell, China deserves to be its own socio-regional category, or at least group it with its sphere of influence (Vietnam, Korea, Japan).

This graph is just horribly "amateurish" (for lack of a better term), and shouldn't be used in any Wikipedia entry

Skyduster (talk) 22:59, 20 December 2012 (UTC)[reply]

Real Dollars

I feel that real dollars (adjusted for inflation) would be more appropriate. If this plot is not adjusted for inflation, then even if real GDP remained constant, the plot would be exponential. Therefore, unless this chart is adjusted for inflation, we cannot tell if there is anything interesting going on besides inflation. --209.131.62.113 (talk) 23:48, 4 February 2009 (UTC)[reply]

Western Offshoot

This Australian says: YOU'RE a western offshoot. Now, how do you like being called names? —Preceding unsigned comment added by 124.168.78.192 (talk) 11:24, 9 May 2008 (UTC)[reply]

More anti-American drivel

How much of the "Western Offshots" is the United States? The maker couldn't bring himself to put in United States? Also, Offshoots is misspelled Offshots. The image will be removed from the Industrial Revolution article. Chiss Boy 17:43, 24 April 2007 (UTC)[reply]

I don't think the author was being "anti-American". We don't know his/her intentions, but I understood "Western Offshoots" to include, the entire Anglosphere (USA, Canada, Australia, NZ). Of course, as I noted above, these regional groupings are flawed. But by assuming that "Western Offshoots" is a reference solely to the United States, you're being just as ethnocentric as you accuse the author of being. Skyduster (talk) 22:59, 20 December 2012 (UTC)[reply]

Image removal

I'll remove this image from most of the articles using it for now. It's not made clear how the data is compiled, and the source given is unreachable.

My main reason for that is that it appears dubious, at best, to me that there should be any reliable data underlying this. There are quite a few other reasons against having this picture, including:

y scale not explained
title is "World GDP", but individual (and quite arbitrary) regions are considered
"Western offshot" appears weird terminology at best. I'm not even sure what's meant here.
bar graphs simply are not appropriate for displaying multiple curves like this.
give thought to a logarithmic scale.

RandomP 11:46, 29 October 2006 (UTC)[reply]

(concerning unreachable) Guess the link to the source data should now be (http://www.ggdc.net/maddison/Historical_Statistics/horizontal-file_10-2006.xls). --Van helsing 12:46, 29 October 2006 (UTC)[reply]

Cool. Okay, so that actually explains quite a bit (which I'm just going to reproduce here so the questions and the answers are in the same place):

the y scale is in "1990 International Geary-Khamis dollars".
"Western offshoots" is defined as (I believe) the current areas of Australia, New Zealand, Canada, and the United States, even before those countries were Westernized
the data is fairly gimmicky until 1820. Sorry, but giving those numbers the major part of the graph area is not appropriate.

I'm also not at all sure what the current best practice is legally for using this document. I don't see a license, so I'm assuming anything more than a very few extracted numbers are a problem. I'd suggest contacting Angus Maddison and asking about using the data.

Please discuss things before reincluding, though!

(As an aside, while the table data might be questionable, the contents of it are extremely interesting, and collecting them must have been a huge effort. This is certainly worth using in other contexts).

RandomP 13:12, 29 October 2006 (UTC)[reply]

Regarding legallity, a data set itself may be copyrighted (even so, a small portion may be cited as fair use, otherwise writting academic papers would be impossible), but not graphs illustraing them and from which the data itself cannot determined exactly (if there is not an already existing artistically very similar graph). Wikipedia has numerous derived graphs and maps based on data sets which itself is copyrighted (like from UN organizations). Regarding validity, Maddison's data has bee3n used in hundreds or thousands of peer-reviewed studies, so it is a valid view. If critical, cite published studies and give an opposing view.Ultramarine 17:01, 29 October 2006 (UTC)[reply]

It's more complicated than that: you can't just take copyrighted data, put it into a graph and claim that that's no longer covered by copyright (for example, there might be an easy way back to the original data). I'm not saying this is a copyright violation, but it's not clearly not one.

I'm still upholding strong objections to this: the graph is mostly based on that bit of the data that doesn't have anything like the purported accuracy, it has several typos and crippling omissions, the copyright status is not at all clear; maybe most importantly, I'm not sure what this is supposed to demonstrate, other than "GDP grew from 1900 to 2003, and was very low before 1820".

At the very least, a new image will have to be generated. Until that happens, please do not use this graph in articles.

Note that unless consensus is an image should be used, well, it shouldn't. Images are less flexible than text, and it is not the removing editor's responsibility to jump through whatever hoops are necessary to regenerate the image without typos, with enough information to be meaningful, etc.

Do not readd the image until those concerns have been addressed.

RandomP 18:08, 29 October 2006 (UTC)[reply]

Again, this data has been cited in numerous peer-reviewed studies. Cite critical ones if disagree. Regarding copyright, see above. Note that your objection would mean that numerous graphs and maps would have to be deleted in Wikipedia. Like probably everything using data from the UN and other international organizations.Ultramarine 18:20, 29 October 2006 (UTC)[reply]

See this: [1] Maddison's papers has been cited by numerous other scholarly papers. If critical, cite published studies, not original research.Ultramarine 18:24, 29 October 2006 (UTC)[reply]

The caption has been clarified. If more clarification is needed, you can easily add it.Ultramarine 18:32, 29 October 2006 (UTC)[reply]

The data is not the only thing being criticised (though the early data that the graph materially relies on is obviously guesswork). The other criticisms are quite sufficient to remove the image for now, sorry. I'd be slightly less inclined to do so if the creator of the image had remembered to include instructions for regenerating it after making corrections, but what we have here is a pile of pixels that is essentially uneditable by anyone but the person who made it. If it doesn't work (and it doesn't), out it goes.

I'm not talking about other images. I'm talking about this specific image, which falls short of basic standards, is not easily fixable, and makes a statement whose dubious notability will have to be discussed if and when a version of it appears that is not immediately removable on other grounds.

Please remove the image instead of engaging in a pointless revert war, unless you actually do believe there are no significant problems with this image that would not warrant edits to a paragraph of text containing the same information. If you do believe there are no such problems, I'd suggest mediation.

RandomP 18:33, 29 October 2006 (UTC)[reply]

I have added your points to the caption. Again, if your are crtical, cite published studies, not you own original resarch.Ultramarine 18:36, 29 October 2006 (UTC)[reply]

I'm not sure which "caption" you're talking about, but it does not matter. The image, not the text accompanying it, or the data it is based on, or the way it is used, is unacceptable, unfixable without significant effort, and should not be used. That's the issue. If you think the information presented in the image is needed, and you cannot generate an acceptable replacement image, write a corresponding paragraph of text and include that in any relevant articles.

RandomP 18:42, 29 October 2006 (UTC)[reply]

You have not explained why, except stating your own orignal research. I have added clarification to the text caption/text summary, as you yourself easily can do. Please state what is wrong with the image or data, citing published studies.Ultramarine 18:45, 29 October 2006 (UTC)[reply]

Yes, I have, but here we go again:

typo in the image (not fixable)
no label on y scale, which is essential for understanding what's going on here
simplistic title
bad representation
nonsensical data (as western offshoots and latin america did not exist prior to 1500, it is quite unclear what's meant here)
years not clearly marked on x scale (they're between tics)
no source to fix all that.

Think of this as the equivalent of a paragraph of text. It has a typo, statements which are inaccurate unless qualified, and it's badly written. We can't fix it. So we'll have to take it out, and replace it by something we can fix: either text, or another image.

RandomP 18:55, 29 October 2006 (UTC)[reply]

Again, it is the image that is at issue here, not any text data whatsoever. The caption was an issue earlier, when the source it gave had become unreachable, but that was just one in a long list of reasons not to use the image. You've fixed one of them. All the others remain.

As for what I could or could not have easily done, it is the including editor's responsibility to make edits that, at least, can be built upon to make Wikipedia better. If that isn't the case, and I feel it isn't here, it gets undone. Original research has absolutely nothing to do with this.

RandomP 18:59, 29 October 2006 (UTC)[reply]

Those reading the text summary will not have any problems. Clarification have been made, like what Maddison means with "Western offshots". A source is given, now working again. Your criticims of the data is original research and uninteresting.Ultramarine 18:57, 29 October 2006 (UTC)[reply]

The text summary cannot make an unacceptable image acceptable. I say the image is unacceptable, and as it is unfixable without significant effort, it should not be used. It's that simple.

Maddison does not use the term "Western offshots".

RandomP 19:03, 29 October 2006 (UTC)[reply]

Yes, he does, that is why I used it. Not also that anyone can make an alternative graph with the free OpenOffice, the data, and a little training. No, there is no rule that anyone should be able to use OpenOffice at once, just like there is no rule that all drawings in Wikipedia made by professional artists should be removed since not everyone can draw well.Ultramarine 19:04, 29 October 2006 (UTC)[reply]

Besides the point. (Fixing should be easy. That's what a wiki is all about. "Just learn to use the application du jour to recreate the image, which will then result in a revert war about which version of the image to use" is not acceptable) In this case, the easiest fix is to remove. Are you actually arguing this image is good enough to be included in an encyclopedia?! RandomP 19:09, 29 October 2006 (UTC)[reply]

No valid reason has been given, original reserach is not interesting. Your points are covered now in the text summary.Ultramarine 19:11, 29 October 2006 (UTC)[reply]

For the last time, I am talking about the image, not the text summary or any other text (that's stored as text in WP's database).

My points are all still valid. The graph doesn't label either axis correctly, fails to accurately say what's even being displayed, and is simply bad independently of those issues.

That readers might be able to detect what the graph should look like by clicking on the image is besides the point.

RandomP 19:17, 29 October 2006 (UTC)[reply]

Anyone that interested would click on the image. There is no requirement that the complete source and explanation should be in the image and not in the text summary.Ultramarine 19:24, 29 October 2006 (UTC)[reply]

Maddison does not use the term. Please check your sources. RandomP 19:10, 29 October 2006 (UTC)[reply]

I have. Line 29, "Total Western Offshots".Ultramarine 19:14, 29 October 2006 (UTC)[reply]

That's the typo I've been talking about! Go read it again. RandomP 19:17, 29 October 2006 (UTC)[reply]

I have. Line 29, "Total Western Offshoots".Ultramarine 19:14, 29 October 2006 (UTC)[reply]

That's the typo I've been talking about! Go read it again. RandomP 19:17, 29 October 2006 (UTC)[reply]

Oh, maybe you are objecting to the typo? Ultramarine 19:15, 29 October 2006 (UTC)[reply]

Precisely. A typo that just happens to make the graph incomprehensible. As I said all along. RandomP 19:17, 29 October 2006 (UTC)[reply]

Eh, I can easily fix that. Anything else?Ultramarine 19:19, 29 October 2006 (UTC)[reply]

Yes. All the other issues. For a start, make sure the caption specifies that 1990 international dollar prices were used, that the year numbers align with tics on the x axis, and that you give basic instructions for recreating the image. Then we have a valid, if still confusing (how is it the "Former USSR" in 1500? What happens to the (very large) intersection of "Former USSR" and "Eastern Europe"?) graph, that we can then hopefully proceed to remove simply because it is bad.

RandomP 19:29, 29 October 2006 (UTC)[reply]

I will not include everthing currently in the text summary in the image. There is no such requirement and those really interested will consult the text summary or even the orginal source, like you did. All information needed for recreation or verification is already given.Ultramarine 19:40, 29 October 2006 (UTC)[reply]

And no one suggested you "include everything currently in the text summary". There are plenty of ways for you (who wants the image in) to go (for example, removing the y scale and saying "1990 prices" would work just as well as keeping it and saying "prices in 1990 international dollars"). However, the current state is clearly unacceptable, and while I'm assuming good faith, if my issues aren't addressed in a "fixed" version I think it would improve Wikipedia to take out the image again. And policy is clear that's what I should do.

(Speaking hypothetically, of course.) RandomP 19:49, 29 October 2006 (UTC)[reply]

What policy are you referring to? So would you find a version acceptable which fixes the typo and states the y scale clearly? Ultramarine 20:01, 29 October 2006 (UTC)[reply]

First of all, sorry, that might have been a bit harsh. There are various WP policies amounting to "if in doubt, improve", the most amusing of which might be WP:IAR.

Fixing the typo, the y scale, and the x scale (align years to tics) are necessary conditions for an acceptable image. While I think they aren't sufficient, they certainly would improve the chances of finding consensus in favour of using the image in any given article. RandomP 21:08, 29 October 2006 (UTC)[reply]

I see no point in changing this if you still would object. What more would you like to change (and not original reserach)?Ultramarine 21:32, 29 October 2006 (UTC)[reply]

Log scale would be better

Hi. A logarithm scale along the vertical axis would be more useful since so much of the graph is so near the 0 horizonal mark. Growth graphs are usually best shown in logarithm scale anyways, because growth is usually percentage of existing growth - expectations and growth potential rise based on prior growth. --70.48.71.143 22:50, 8 January 2007 (UTC)[reply]

Depends on your purpose. A logarithimc graph may be better if comparing year-to-year change, but not if comparing starting and ending points. Also, many people do not understand logarithmic scales.Ultramarine 23:04, 24 April 2007 (UTC)[reply]

I agree with user 70.48.71.143: A log scale would be much better in representing the data that is there. Now whether or not the data is accurate is not my argument here. Whoever decided to go back and do all the research and compile the new data, a log scale of GDP per capita or whatever data ends up on the Y-axis would be more fitting. As a graph that makes use of a log plot, it should also be noted in the updated caption, as well as a brief description of why the log scale is used with maybe a link to the mathematical term "logarithmic" or something similar. ~AK (talk) 06:31, 24 November 2007 (UTC)[reply]

Are you kidding?

Most ambiguous, most flawed graph I have ever seen in my life.

Why? --Vince | Talk 05:54, 19 October 2007 (UTC)[reply]

There is just to much speculation and retrospectivisim in this graph... What was the former USSR doing around time of the Romans? Hah? —Preceding unsigned comment added by 81.183.113.169 (talk) 13:39, 5 April 2008 (UTC)[reply]

terrible image needs deletion

no y axis definition, what are 'western offshots'?, and why is there no data for most of the graph? 'nuff said. —Preceding unsigned comment added by Aformalevent (talk • contribs) 03:25, 3 October 2007 (UTC)[reply]

The graph is commonly used in economic growth theory to illustrate the Malthusian period. Look at the discussion above, I think much of this was discussed. Brusegadi 04:31, 3 October 2007 (UTC)[reply]

Line graph

Shouldn't this be a line graph? It's data over time, right? 66.31.174.231 (talk) 16:34, 13 January 2008 (UTC)[reply]

I don't think a line graph is appropriate because there are so few data points. Connecting data points 1000 years apart with a line adds no additional meaning, and may even be misleading. I agree that the bar graph isn't ideal either. I also think this graph should be on a log scale so we can see most of the data. Rm999 (talk) 02:25, 1 June 2009 (UTC)[reply]

?

This is one of the worst graphs I have ever seen! So amount of work could kix this graph. It's horrible! Delete it! It's beyond bad! 58.178.30.85 (talk) 03:44, 8 June 2009 (UTC)[reply]

Alternative version

I've created an alternative graph from the same data source showing the world average only from 1500-2003 at File:World GDP per capita 1500 to 2003.png. Doesn't show the different geographic areas, but the present uses of this image don't need that. Qwfp (talk) 16:16, 25 June 2009 (UTC)[reply]