Black swan theory

[Whomp]

His point is most people would say that the second statement made more sense since it included something. He would say there are more potential outcomes by saying he killed his wife. I will finish the book since I gave it to my gal but she's going to give me hers to see if she should read it. So far I'm not convinced.

 

[C~G]

His point is most people would say that the second statement made more sense since it included something. He would say there are more potential outcomes by saying he killed his wife.

Well, that is like quite simplistic way of looking it but he's partly right.

Then again the second one is the one that offers more stimulus towards inventing new possibilities for the story while the first one lacks basically everything.

A is B. A did C to D.
A is B. A did C to D for E.

Basically we might consider first one being quite insignificant information as it doesn't actually give us any new possibilities to move forward in our thinking.

So it doesn't necessarily mean that we feel that second one makes more sense just because it includes something but because it gives us possiblity to continue. In other words the reason isn't because it's more descriptive but because actually allows us create more connections towards other things while basically first statement relies into entirely pure random factors.

It would be nice to see whether all people see this in the same way or is it more like 50-50.

 

[Whomp]

I think wiki does a good job explaining some of his thoughts. The part regarding the man killing his wife would fall under narrative fallacy. I have to admit I see an extraordinary amount of this go on around the world.

Taleb believes that most people ignore "black swans" because we are more comfortable seeing the world as something structured, ordinary, and comprehensible. Taleb calls this blindness the Platonic fallacy, and argues that it leads to three distortions:

Narrative fallacy: creating a story post-hoc so that an event will seem to have a cause.
Ludic fallacy: believing that the structured randomness found in games resembles the unstructured randomness found in life. Taleb faults random walk models and other inspirations of modern probability theory for this inadequacy.
Statistical regress fallacy: believing that the probability of future events is predictable by examining occurrences of past events.

 

[C~G]

I think wiki does a good job explaining some of his thoughts. The part regarding the man killing his wife would fall under narrative fallacy. I have to admit I see an extraordinary amount of this go on around the world.

Yeah, that explains parts of it.

There could be other explanations as well.

Now, of course I have to get this book...

 

[.Shane.]

Here's a bit of a side-related thing/site I found (via Boing Boing). A website that claims to be interested in the unexpected properties.

The link is to some man-made "sand" that doesn't get wet....

 

[Fifty]

Statistical regress fallacy: believing that the probability of future events is predictable by examining occurrences of past events.

It is utterly insane to call that a fallacy. Almost everything we do is based on that principle.

 

[Julian Delphiki]

Finished the book, and earlier opinion still holds - it was ok read but i'm not too enlightened really. Summary of the book as i understood it: he bashed economists and said that people don't prepare for big unexpected happenings.

Societe Generale fraud seems like a black swan to me, "was not supposed to be possible". While losses told are €4.7 billion, he was playing around with €50 billion which is more than Finnish national budget for 2007 (€40 billion).

 

[Verbose]

It is utterly insane to call that a fallacy. Almost everything we do is based on that principle.

Since we like to be comfortable while guessing...

 

[Fifty]

Since we like to be comfortable while guessing...

huh? 10chars

 

[Whomp]

I think Verbose is saying what the quant guys from Renaissance Capital said in August when their computer statistical models didn't work due to a number of days in a row where the markets had 25 standard deviation event days in a row.
The firm lost a ton of dough and it's the same reason why Long Term Capital Management was destroyed. Black swan event that's an extreme outlier that should never happen.

 

[C~G]

Fifty, it's about the simple fact that even certain thing is extremely probable based into earlier data doesn't mean the result is inevitable.

Also even all possibilities are "mapped" by earlier data there's always the possiblity of something radical. Even though something looks like it is "bound" to happen doesn't mean it will. There's also strong tendency of humans to look for favourable outcomes when there's none available.

Expect the unpextected.

 

[Fifty]

Nobody has told me yet why inductive reasoning ought to be called a fallacy. Nor has anybody explained to me how they manage to live their life without committing this "fallacy" constantly.

 

[Mise]

Just because the sky is blue today doesn't mean it's going to be blue tomorrow.

 

[Whomp]

OK Fitty--I think Taleb being a student of Popper is trying to challenge our biases and prejudices when making decisions based on what we perceive to be perfectly logical when in reality they aren't.

I'll relate to something I think I know about but Taleb would say I've just been lucky predicting which are markets...

He makes the point that if you exclude the 10 largest daily moves in the S&P 500 over the past 50 years, the index would be more than 1000 points higher.
That means an index that's currently at 1,330.61. How could I be so wrong not to see those 10 days? As Don Rumsfeld would say they're "known unknowns"? Take this Wednesday...do you think people have assigned a reason why the Dow Jones swung 600 points in a space of hours? I think many have given their narrative as Societe Generale was responsible or the fed cutting rates or hedge funds covering shorts. I'd say it was one of those black swans that just happen.

To put it another way would a turkey be confident after a year and a half of life that it will continue to be fed to make it happy and content and then one Wednesday in November something happens out of the ordinary...

 

[Verbose]

Nobody has told me yet why inductive reasoning ought to be called a fallacy. Nor has anybody explained to me how they manage to live their life without committing this "fallacy" constantly.

A lot of the inductive reasoning going on tends to be abductive reasoning, which we later verify inductively,when possible.

 

[gallego]

It is utterly insane to call that a fallacy. Almost everything we do is based on that principle.

I don't think the point is to reject statistical inductive reasoning outright. I think he's just saying that it is often overvalued and that we actually miss a lot of good stuff if we shun the statistically improbable.

 

[CCA]

This book from the blurb seems like congratulatory self-wanking.

I particularly dislike the use of strawmen here:

"Yet the person making these statements will likely be addicted to his iPod, wearing a T-shirt and blue jeans, and using Microsoft Word to jot down his "cultural" statements on his Intel-based PC, with some Google searches on the Internet here and there interrupting his composition. If old enough, he might also be using Viagra."

Meh, I'll borrow it from the library, but this book is deinitely not going to be one of my shelf books.

 

[Julian Delphiki]

This book from the blurb seems like congratulatory self-wanking.

I particularly dislike the use of strawmen here:

"Yet the person making these statements will likely be addicted to his iPod, wearing a T-shirt and blue jeans, and using Microsoft Word to jot down his "cultural" statements on his Intel-based PC, with some Google searches on the Internet here and there interrupting his composition. If old enough, he might also be using Viagra."

Meh, I'll borrow it from the library, but this book is deinitely not going to be one of my shelf books.

That chapter was just as bad in the book as it sounds here. In another chapter he claims that no author has sold more than 200 million books, which i think is false.

 

[Hygro]

So I finished reading it almost a week ago. It's a very good book. Quite funny at times, as the author does not divorce his personality from his prose.

I don't have too much to say, except that I'm smarter for having read it. And it's under 300 pages.


Anyway, his "euro-bashing" is mostly in jest and aimed at the French (a language in which he is clearly fluent). He makes a very good point for why human intuition was good for predicting things pre-civilization, but is obsolete in the post-industrial era. I highly recommend reading it.

Thanks, Whomp!

 

[Bast]

Sounds very interesting.