News for hackers. Stuff that matters.
Re:Two thousand years in one chart by chime on Saturday September 10, 2011 @11:52PM (#2983155)
The scale is % of economic output and years lived. Assuming the output is inflation adjusted, why would you prefer the scale to be logarithmic?
Re:Two thousand years in one chart by joshuamerrill on Sunday September 11, 2011 @12:32AM (#2983220)
Agreed, since economic output is measured in 1990 Dollars, there's no need for a logarithmic scale here.
Re:Two thousand years in one chart by ars on Sunday September 11, 2011 @01:08AM (#2983265)
Because economic output always builds on previous output. That is the definition of an exponential function, and to reverse those effects you need to take the log.

It's easy to see this - just look at the early years, and notice the curve - it's an almost exact exponential curve. Graphing an exponential curve teaches you very little about the underlying data.

Re:Two thousand years in one chart by GavinB on Sunday September 11, 2011 @01:27AM (#2983288)
A logarithmic view would be great for showing when the population and output stalled or increased faster, but the purpose of this graph is to show that the fact that the increase is dramatically exponential, and at what point in the curve we are right now.

Ideally both would be shown--but just showing the log graph would be misleading in this context.

Re:Two thousand years in one chart by doyoulikeworms on Sunday September 11, 2011 @03:10AM (#2983394)
I agree. I think that the point of this graph is to convey exponential growth, and simply that. It's not meant to be used "seriously," which'd make a logarithmic graph more suitable.
Re:Two thousand years in one chart by hammock on Monday September 12, 2011 @12:31AM (#2985835)
Ars is correct on this one (and also resdirector, the topvoted comment right now)- the issue is not money inflation, it's the simple fact that compounding growth will be exponential, always. So if your point is to show that growth rates increased in recent centuries, then you need to show growth rates, or show this but using a log scale.