THE FOURTH QUADRANT: A MAP OF THE LIMITS OF STATISTICS - Nassim Nicholas Taleb

THE FOURTH QUADRANT: A MAP OF THE LIMITS OF STATISTICS [9.15.08]
By Nassim Nicholas Taleb

http://www.edge.org/3rd_culture/taleb08/taleb08_index.html

A great article on how the use of conventional statistical tools have made us very very susceptible to the impact of rare events.

The article starts with reference to the ongoing Global Financial Crisis and talks about in general people quantifying risks in a financial system almost dumbed it down to – “It has not happened in the last 30 years and hence its probably not likely to happen anytime soon” while almost the reverse was true.

The gist of the article though is how the current state of Financial Statistics dumbs down the improbable but highly dangerous event while modelling and deciding the future mainly because the probability based on past data fools them. A classic case is the Turkey Story.

“A Turkey is fed for a 1000 days—every day confirms to its statistical department that the human race cares about its welfare "with increased statistical significance". On the 1001^st day, the turkey has a surprise.- ITS THANKSGIVING”

He goes on to create a Mapping this based on two distinct types of decisions - true- false binary type and yes-no but how much - and two distinct types of randomness – mediocristan (the measurable randomness) & the extremistan (the event that hits you out of the blue).

He goes on to explain how because of our fragile theories, a game breaking event can throw everything up in the air. And that’s what seems to have happened in general with the current Financial Crisis. Wise men bet so heavily on the a set of probabilities that were self reinforcing that at the first instance of an anomaly, the whole system collapsed like a house of cards caught in a tornado.

The last section deals with what to do with such extremistan items.