Fooled by Randomness, Nassim Taleb
Taleb's first book and his first identification that there are a lot of things horribly wrong with the understanding of probability generally, mainly because as human beings we are simply not built to calculate it right.
His later books are less apologetic, more eclectic, and to my mind better reads, but this one is the most theoretical of the books I've read thus far. The probability concepts at least made me feel like I was learning something new, but this if you will is the genesis of his argument, which he goes on to develop in later books. This is the second-last book of the
Incerto that I've read and all that I have left now is
The Black Swan, which I'll get to a little later. Some of the material covered is familiar - Kahneman's
Thinking Fast and Slow gets a bit of airplay, but the rest of it is still original and fresh and interesting. He has a certain admiration for George Soros' conjectured lack of humanity as it applies to investments, but sums the man up in one of the best metaphors I have yet heard: 'He uses up people like napkins.'
There are, though, three main things that the book puts across which I didn't get in emphasis from any other place:
(1)
Saying someone may have succeeded solely by luck is not merely sour grapes. It's mathematically supported. Skill has a big role to play, but it is entirely possible if not probable for someone to succeed solely on luck if the conditions are right - and those conditions are nowhere near as remote as you might first think. Simplest illustration can be reasoned out by mathematics:
Take 10,000 stock traders with the same skill, or no skill. Allow that in a given year, half of them will make money and get to stay on at their firm, half of them will lose money and be fired. This is a true toss of the coin, 50/50. How many people are still at the firm after 5 years? It gets down to about 313 or so: 5,000 the first year, then 2,500, 1,250, 625, 313.
That is, about 3% of the original group of 10,000. Assuming you eliminate the skewing of "skill", pure probability says that about 3 in every 100 of a group of 10,000 people under these conditions would succeed solely because of the march of probability, as the group gets smaller.
Would you hire a trader from that group who could point straight at his record saying he'd made money for 5 years in a row? Say, one who proudly proclaimed the deathmatch conditions of his office, right out of Glengarry Glen Ross? The probability illustrates, very clearly, that some impossible-to-compute fraction of his peers, if not him, got by
on nothing but probability, no sour grapes, no jealousy needed, it is a matter of mathematics.
Maybe you would. It depends on the starting group. But that can be impossible to figure out as well. Taleb says he'd be more impressed if the starting group was 10 and the guy was still there after 5 years under the same conditions (this, which Taleb doesn't mention but Kahneman does, is also potentially a problem, is partially because of the Law of Large Numbers: a tiny sample group means you are
more likely to get wild, inconsistent variations from the norm than with large. Also known as why you never trust a social science survey with anything less than four digits as its test group, because most social scientists and most economists don't understand probability and statistics.)
In short: probability is counterintuitive and people should understand it a lot more than they do. I am seriously thinking about looking up an introductory course on probability and statistics because of some of the implications this raises.
(2) The world is much more random than we think, and we have lost the secret of dealing with it. The ancients understood the fickle nature of life on the planet. Karl Popper was a godsend to humanity because his hard limits on science - falsifiability - are going to be the anchor that eventually consigns most social science to the dustbin. I was also surprised to learn that, prior to Bacon, science
was conducting itself in a solely
deductive manner. It was only after Bacon - and philosophers like Hume raged against this approach - that science began to wander down the path of
inductive reasoning, that because A was observed in conjunction with B that A could be said to cause B.
And the ancients' most potent secular way of combating this randomness was stoicism. Taleb spends a lot of time on the philosopher Solon, and in particular a conversation he had with the legendarily-rich kind Croesus. Solon came to Croesus of Syria's court, and appeared utterly unimpressed and unmoved by the great king's magnificent hall and riches. When asked why, Solon simply noted that there was no point being awed by the king's wealth, as all of this could be taken away tomorrow with all the speed with which it did not first come. The legend goes that Croesus did indeed suffer a fairly dramatic turn in his fortunes, defeated in battle, and cried out to the sky on the day of his execution that Solon was right. (Cyrus the Great, the man who had defeated him, is said to have stayed the executioner's hand due to this puzzling outcry, and on hearing Croesus's story, spared Croesus's life and allowed him to live out his days as an advisor instead.)
(3) Absent someone with extremely high qualifications, reviews invariably say more about the reviewer than they do about the work under review. I think I'd already figured this out intuitively from exposure to a lot of the toxic waste of journalism, but this appears to be supported more widely. Taleb starts his crusade against journalism generally right here, and it comes down to this: journalism was never about information. It was always about entertainment.
(He also specified a really good way of cognitively/biologically distancing yourself from the daily shitstream of journalism: you can have the TV turned on and watch the tickertapes or the headlines ... but just have the sound turned off. This is liberating, it exposes the people on the screens as clownish idiots. Your brain is refused the evolutionary trigger of taking things you hear seriously. In general, Taleb's approach is to suggest by the end of the book that people think with their emotions (cf. Gerry Spence and half the books on the Scott Adams persuasion list), and that
this is inevitable. The only way around it is to do the equivalent of a trick from the Orson Scott Card book
Ender's Game or more classically, Odysseus at the isle of Scylla: fill your ears with wax and have yourself tied to the mast while your boat sails the fatal passage.
That, and set yourself up with options. The guy can just about convince you to go into the area off the strength of this book.
In short, avoid bullshit, avoid noise, and avoid journalism -- and I have repeated myself twice there.
I also realise, given these thoughts, that me crapping long reviews out of books is actually a bit pointless. The original way this thread started was actually a good heuristic: when you finish a book, post it up here. Don't include a review. Because I am not an expert in many of these areas, so why should I judge the book to be good or bad? The real impact is
to name the book and get other people interested in reading it. Taleb holds that word of mouth is the only acceptable form of advertising out there. I benefited most from just getting the titles of books that others had read here, I shall henceforth also give that benefit to others. I have used these reviews as a way of trying to write out the kernels of what I got from these books, but this is an abuse of your time, gentlemen, so in future I plan on just on observing a twisted maxim of Wittgenstein:
Whereof one cannot speak, one must remain silent.