brianeyci wrote:Stas Bush wrote:What is the hindrance to seeing the U.S. distribution as a pyramid? Reduction of numbers in the lowest quintille?
It's because Master of Asses is a big fat fucking moron who thinks that models have to be perfect, or is a big fat fucking liar.
Again, the central tenant of your idiocy is that the world's poorest people outnumber all higher levels of income. This is not true.
Check out this graph he says
proves his point repeatedly. Take a look at the x-axis. See anything off with it? Anything wrong? That's right, the x-axis is not to scale. It treats the distance between 100 and 1000 as exactly the same as the distance between 1000 to 10000. More ridiculously, the distance between 10000 and 100000 is the same too.
You are, honestly,
this stupid, aren't you? I mean, how did I make a mistake like missing the fact that a graph's axis was logarithmic? Oh, wait, I INTRODUCED the graph as using a logarithmic scale. Here are some quotes:
I wrote:Here's the chart for the world from 1970-2000. It's actually really cool. You can see that the world basically gets richer and more bell-shaped (keep in mind, the axis is logarithmic), but still skews to the right... funny... I though I said that early on the last page... somewhere... when I was arguing against some dumbass....
I wrote:Look at this graph. This shows the distribution of income across the entire world. I know that I used a big word when I introduced it as having a logarithmic scale, but regardless it shows that you are INARGUABLY wrong. You also could've gotten this from the CIA World Factbook listing of GDP/capita, as I pointed you to earlier. You totally ignored this evidence, as well. I therefore invoke DR 6. Put up or shut up.
Hilariously, Brian STILL doesn't get it. He's honestly this dumb. Moreover, he treats it as being somehow dishonest that I would present a graph that used a logarithmic scale (an imminently reasonable thing to do when working with density functions like this) as a... graph that uses a logarithmic scale! My god! I'm such a liar!
What grade are logarithms introduced in, again? Oh, yeah.
eighth, apparently. What is that? Junior HS? And what was it that he said to me earlier in this very thread?
brianeyci wrote:Could it be that you don't understand Grade 10 math? [...] Wow you're a fucking moron
Your graph is a lot better. You can see the difference between rural China and urban China especially, where the purest kind of capitalism exists.
How is that graph better?
1. It treats each country/region's population as being discrete--that's how it creates those seemingly-large discrepencies (also, it doesn't seem to account for every country). The graph that I originally presented takes into account the distribution WITHIN COUNTRIES as well as the distribution amongst countries--when comparing the income of the entire world's population, there is no question which is the superior graph.
2. It just disguises the actual relationship because the high end has much lower density, and by treating this as a line graph rather than a density curve.
3. It makes some seriously erroneous claims, such as lumping "rural India" in with "Africa," even though (as you can see from my chart) the poorest 2% of Indians still make more income annually than many of the poorest world citizens who live in Africa (or even many citizens of Pakistan or Bangladesh). Thus, it totally disguises the LENGTH of the left-hand tail, which is the whole focus of our analysis.
It's not a bad graph--it would be useful in a pinch (it still shows the tail), but it clearly pales in comparison to the one that I provided, especially for the use you are attempting to put them to.
It's modeled by y = a(1/x). The US portion of the graph is modeled by y = a(1/x) too. The entire world could be modeled by y=a(1/x), but there's mitigating factors such as lack of pure free trade, war, governments and... sensibility. See, what Master of Ossus is too moronic to understand is that in attempting basic modeling, you strip as many extraneous variables as possible, like how when you encounter the heat equation for the first time in physics you ignore the heat that leaks out of the pipe, or the heat that leaks in the opposite direction. That doesn't mean the basic heat equation is invalid any more than it means that y=a(1/x) is invalid in describing that graph.
Do you seriously look at a graph like
this and say, "gee... that looks as if it would be well-modeled with y=a(1/x)? Or do you say, "gee... that approximates a normal distribution?" Seriously. Your model ignores everything to the left of the peak, in which a substantial fraction of the world's population lies. If by stripping away "as many extraneous variables as possible," you mean ignoring everything to the left of the peak, then you're right--I'm leaving that in there. Moreover, what in the hell do factors like "lack of pure free trade, war, governments and... sensibility" have to do with the data that has been presented, which unequivocally is better modeled with a normal curve than with your idiotic hyperbolic function?
Master of Asses seems to think it's devastating to my argument that the shape of the graph isn't a perfect pyramid, when almost the entire graph, the first graph he posted
is a pyramid.
Your graph is even better at illustrating the point.
Brian, there is NO graph here that is more accurately modeled using y=a(1/x) than by a bell curve--all of them have tails to the left, which your model completely ignores. Sorry, that's not a curve of best fit--that's sheer intellectual dishonesty. And that IS devastating to your claim when you're arguing unequivocally that the poorest people are also the most populous. Again, you may be able to DEFINE people below the peak as constituting the poorest group, and thereby ignore this difficulty, but when you're arguing against someone whose hypothesis PREDICTS such a tail, it destroys your argument.