Tuesday, September 11, 2012

We Are Now One Year Away From Global Riots, Complex Systems Theorists Say


Short description of 2011 study. We may be getting close to the tipping point, although it is difficult to predict effects of complexity with precision.

Mother Board
We Are Now One Year Away From Global Riots, Complex Systems Theorists Say
Brian Merchant
(h/t Naked Capitalism)

5 comments:

Roger Erickson said...

not sure this is a useful application of so-called complexity theory

seems rather superficial to me

If you ask a simplistic question, it hardly matter how complex the analysis is.
Wasn't the famous answer "47" - or something to that effect?

ps: only British complexity theorists would publish something like this;

would any Russian complexity theorists publish something so superficial? French? German?

ps: ps: based on this premise, every time an individual gets hungry, all their cells are expected to riot?

How did humans get this far with so much "complexity" expected?

Roger Erickson said...

Oh no, this is from MIT, not the UK.

Didn't Shewhart say that data is meaningless w/o context? When was the last time you saw Americans rioting over food prices? Swedes?

Their simplistic model is highly context dependent, and the model doesn't take enough variables into account.

Besides, it's still asking a useless question. Ask a better quetsion, say, how to improve the quality of distributed decision-making, and see if the model has anything useful to say that isn't already obvious.

Roger Erickson said...

oops, meant to say that


Their simplistic QUESTION is highly context dependent, and the model doesn't take enough variables into account to say much that's very useful.

Sounds like a solicited paper by some police force, maybe DHS.

Magpie said...

I'd say that their figure 1 (which is reproduced at Motherboard's site), is quite suggestive: it involves low income countries geographically dispersed (one in the Americas, a few in sub-Saharan Africa, many in north Africa and the Middle East, one in south Asia). An explanation to those riot clusters should account for these characteristics and the explanation offered does that.

Of course, it raises several questions:
(1) are the data represented reliable? how is a riot defined? are those all the riots registered?
(2) the occurrence of riots and the price of food may be related, but this doesn't need imply that one causes the other: they could be caused by a third variable. Are we sure a possible third variable also points to incoming riots?

In any case, we'll know soon enough whether these guys are onto something: in about one year's time.

Matt Franko said...

"many in north Africa and the Middle East, "

Imagine what is going to go down over there if the west ever goes off of petroleum..... a magnitude of human carnage the world has never seen....

rsp,