Recently, I read this fascinating book by Naseem Nicholas Taleb ( NNT for short).

NNT delves into why predictions & forecasts go terribly awry in particular economic / poicy forecasts and debunks a number of investing myths. Also delves into why risk models are not accurate.

First some background, NNT himself is currently a professor of Uncertainity, in his past life he has been a trader ( a quant - someone who is paid to apply principles of mathematics to trading strategies to eliminate risk).

This is his second book, the first being Fooled by Randomness.

In this book, NNT shows us how Induction is a myth, that it only takes one incidence of a contrary event to disprove a hypothesis. The title derives from the hypothesis prevalent till about 1600s in the West, that all swans are white. The discovery of a black swan in Australia puts paid to this hypothesis. The Black Swan then becomes a metaphor , a symbol of all that is unexpected and likely to have a huge impact. Black Swan events shape our lives. Examples abound, NNT in particular cites the tragic events of Sept 11, 2001.

We as humans convince ourselves that such events were explainable - in
hindsight whereas almost all consequential events in history were
unexpected.

NNT goes on to show us how the Bell Curve / Normal distribution is a fallacy, and how the structured randomness of models / games does not manifest as such in life. ( the Ludic fallacy) - how mistaking the map / model for the real causes us to lose perspective.

He shows us how many events are just without precedent , therefore past as a predictor of future is not very useful. Our philosophy / thinking patterns have many fallacies and NNT uncovers such myths and fallacies.

NNT cautions us not to mistake absence of proof as proof of absence.

World events happen in two realms for him, the Mediocristan where the bell curve rules and Extremistan where the bell curve does not rule, where experts are merely empty suits and stuffed shirts. In Mediocristan, predictions are possible and attaining expert level knowledge is possible, however for better or for worse, our worlds are a mixture of the two.

He extols the virtues of other tools like fractals and scalable distributions that may be used to temper expectations and make black swans grey. He argues for counter factual reasoning in an effort to uncover real reasons, and combat risk.

I really liked his style, the prose is not difficult to follow, it is very engaging and a very thought provoking book. It has altered my way of looking at things, though I still am uncomfortable with its conclusions but that is just what the book intends to do - to break away from your old patterns, to take a fresh look at things.

I wish there was more of here's what you can do type of material in the book but I guess that would take away the fun of thinking for oneself. ( It does have a section 3 devoted to some of this but there is not much direct apply these principles and be happy instructions in there.)

Overall : A must read !

Interested folks can find more info here :

Amazon:http://www.amazon.com/Black-Swan-Impact-Highly-Improbable/dp/1400063515

On wikipedia :

http://en.wikipedia.org/wiki/The_Black_Swan_%28book%29