Just read: The Black Swan: The impact of the highly improbable

[gomberg]

The Black Swan: The Impact of the Highly Improbable by Nassim Nicholas Taleb.

I couldn't put this one down and am going to read it again. This book was much deeper and more interesting than "Fooled By Randomness" and I'm still digesting the ideas of the book. I may review it in depth later when I have time, but I think everyone should read it.

[SGspecial]

[ QUOTE ]
The Black Swan: The Impact of the Highly Improbable by Nassim Nicholas Taleb.

I couldn't put this one down and am going to read it again. This book was much deeper and more interesting than "Fooled By Randomness" and I'm still digesting the ideas of the book. I may review it in depth later when I have time, but I think everyone should read it.

[/ QUOTE ]

I read the excerpt from the Random House site:

http://www.randomhouse.com/catalog/d...p;view=excerpt

I have a couple of comments just based on this (not including the irony of the name of the publishing house).

1. His assertion that the "inability to predict the improbable" renders the "ability to predict the probable" useless, is itself useless. He uses the 9/11 attacks as an example. He seems to be saying that the only releveant factor in the collapse of the trade towers was that they were attacked in a completely unanticipated way. Of course, he neglects the fact that had the engineering and construction of the towers not focused on the "minutiae" of building skyscrapers, they would have collapsed long ago due to high winds or shifting foundation (or never would have stood in the first place).

2. He uses the Nazi "end run" around the French Maginot line in WWII as another example of a "Black Swan". While I agree the french did a poor job anticipating this tactic, and they did suffer a terrible defeat because of it, there is no way to tell whether or not building the line was the proper strategy since no one knows what would have happened if they DIDN'T. It may have been their best defense, it just didn't work. His argument is purely "results oriented" and as poker players it is our duty to point out that results oriented strategy is hogwash.

Otherwise, it sounds like a very interesting book [img]/images/graemlins/smile.gif[/img]

[gomberg]

SG - good points.

1. I think the assertion is not useless if you can use it to at least know that things outside of your model "can" happen. He mainly is bashing bell curve enthusiasts who misapply it to social science situations and economics.

2. I think he used this as an example because it was outside of the realm of possibilities from their perspective.

But yeah, there are some fascinating chapters in this book. And his style of writing is quite funny and witty. Not to mention a lot of the book makes you think and will change the way you look at the world (at least a little).

[SGspecial]

[ QUOTE ]
SG - good points.

1. I think the assertion is not useless if you can use it to at least know that things outside of your model "can" happen. He mainly is bashing bell curve enthusiasts who misapply it to social science situations and economics.

2. I think he used this as an example because it was outside of the realm of possibilities from their perspective.

But yeah, there are some fascinating chapters in this book. And his style of writing is quite funny and witty. Not to mention a lot of the book makes you think and will change the way you look at the world (at least a little).

[/ QUOTE ]

1. I guess his point is that events that disrupt the day-to-day equilibrium that we create (or anticipate) will have an effect that is tremendously more significant than their relative likelihood. That I'll buy. I'm just not buying the claim that our day-to-day details are irrelevant in the face of those possibilities.

2. I picked up on this example precisely because it's the one historical subject I may have an edge against the author on (having done a theme paper on it in HS [img]/images/graemlins/smile.gif[/img] ). The French had no idea of the scope of the Nazi war machine and I believe in retrospect the Germans could have punched through the Maginot line with very few casualties and beat the French army in a race to Paris in any event. I don't know what the highest levels of French intelligence knew at that time, but I suspect neither does he. That's what makes this a poor example.

The 9/11 example of airline security is the one that seems to fit his thesis the best. Since the airlines already had many security measures in place to prevent terrorism, they could haven taken a few more simple steps to prevent the 9/11 attacks. The didn't take them because they didn't anticipate the possibility of a 9/11 style attack, or felt that it was too unlikely to worry about.

Still it is a fascinating concept. In 2p2 terms, just ask Greg Raymer or Joe Hachem about the impact of a Black Swan.

[JasonP530]

got it today and am super excited to read. will post thoughts when I am done.

[denflo60]

I also just finished the book. If you accept the black swan thesis, you can skip most of the book and just read the last few chapters which read quickly compared to most of the book and you will get the drift of his argument. I find his thoughts intrigueing especially with respect to real world use of the bell curve. I wonder whether his theory has any application to 100 way video poker in the short run. Denny

[LA_Price]

I finished this book the other day and it is one of the best I have read in awhile.

Nassim Taleb is the man and is now on my list of people I most want to play poker against.

I was trained in mathematics and have my degree in actuarial science. This book strikes at the very heart of what statisticians can and can not do.

The most important ideas of this book to me are-

The black swan problem-This is basically the idea that random "outliers" can have a dramatic effect and are impossible to forecast. They do not fall into the normal distribution and thus we often discount them in projections. As Taleb says we never learn that we never learn. Each new unexpected event is always seemed to be "predictable" because of the

Narrative fallacy- which is the idea that we construct our histories backwards. Poker players have sort of known this idea for awhile but usually use the term "results based thinking". News media are the biggest culprits, usually playing the part of the monday morning quarterback, and analyzing only the results of what happened. History could have happened in infinitely many possible ways and yet we only see one of them. He uses the example of what if there was a person who had the idea of putting secure locks and procedures on all the cockpit doors on September 19,2001? 9/11 would not have happened. This person would be a hero because they would have prevented thousands of deaths, but the problem is we would never know it or give him credit.

Ludic Fallacy- This is the idea that we can model the stock market, business projections, etc based on games. The reason this is not possible because there are two different types of randomness which Taleb calls

mediocristan/extremistan- mediocristan is sort of a "predictable randomness". For instance if I roll a fair six sided die the probability of getting any number is one in six. If I roll once I will get a random number, but we know the laws the govern dice. The results after many trials will follow very closely to the statistical predictions.

Extremistan is the type of randomness you encounter in the real world, like the stock market. We do not know the laws that govern the stock market, at least not well enough to predict what it is going to do with any great degree of accuracy. Taleb shows mathematically why it does not adhere to a normal distribution and why many Nobel Prize winning investment methods are idiotic and many expert are nothing but empty suits.

I'm going to read this one again as well but I think it(along with Fooled by Randomness) is one of the most important books I have ever read.

The part of the book I am unsure about is his investment advice where he says to put 85-90% in something like treasury bonds and 10-15% in extremely risky investments. Whether this will truly give you the highest long term expectations seems open to a lot of debate. I know nothing about investing so I'd like to hear some thoughts on this. I think what you should invest in depends a lot on your age, and what level of risk you want in your life. These seem to be the levels that he is comfortable with, but a lot of people would be willing to take a lot more risk. Some people want the largest chance to be in the extremes and are willing to risk going broke and being a nobody if they also have some chance of being rich and famous. The reward is worth it to them, even they say had a 90% chance of being broke but a 10% chance of being rich and famous the magnitude of the reward and living the high life is worth the gamble to some.

This is a very good book overall. I'm reading Overblown by Jim Mueller right now and I find that they are both talking about the same subject. How do we make and evaluate decisions in an unsure world?