April 10th, 2010 
David Hill 
- ISBN13: 9780471381976
- Condition: NEW
- Notes: Brand New from Publisher. No Remainder Mark.
Product Description
A must-read for anyone who makes business decisions that have a major financial impact. As the recent collapse on Wall Street shows, we are often ill-equipped to deal with uncertainty and risk. Yet every day we base our personal and business plans on uncertainties, whether they be next month’s sales, next year’s costs, or tomorrow’s stock price. In The Flaw of Averages, Sam Savageknown for his creative exposition of difficult subjects describ… More >>
The Flaw of Averages: Why We Underestimate Risk in the Face of Uncertainty
Posted in 
Tags: 
The author appears to have little knowledge of statistics.
1. There is a whole part of statistics called Non-Parametric statistics which overcome the short-coming of average, and variance. So his statements that statisticians are not aware of some of the short-coming of average, and have no tools to deal with it is wrong.
1. He define a Random Variable (RV) as an uncertain Variable. Well, RV is mapping of representation of events and outcomes to numbers. For instance, a toss of a coin can have two outcome Head or Tail, so mathematically one can represent Head as , and tail by numbers x.
2. He also miss-informs the readers about the concept of dependency with his example of the two ladders that are connected. If ladder 1 has a probability of falling equal to 0.1, and ladder 2 has a probability of falling equal to 0.1, does it mean that the probability of them falling when connected is 0.01, of course not. By adding a connector we have changed the essence of the problem. This can be compared with throwing two dices. While the probability of having a 3 on each dice is 1/6, the probability that the sum (which is the result of throwing both dices at the same time) will be equal to 6 is not 1/36 but it is equal to the sum of the probabilities of throwing a 1 and 5, 2 and 4, 3 and 3, 5 and 1, 4 and 2 or 5/36 (not 1/36 as the author may want us to believe).
He changes the essence of the problem but mislead the readers that he is still using the same initial assumptions.
3. Lastly his description of the Black Swan is wrong. If a wheel has numbers between 0 and 1, and one is turning the wheel in hope that it will stop at a number larger than 0.2 that the wheel goes flying and breaks has no effect on the probability that it will stop at a certain number. The two events are independent, and we have a new problem that is created.
Now if the wheel has one number 0.000001 in it wheel, one may assume that the wheel will stop at this number is extremely slim, and will not even bother taking it into consideration, and it will be discredited. Hence, this number when it occurs it will be the same as the black swan.
4. He states that quant could not predict the outcome of the market meltdown, and that there exists no tool to predict such events. Well, there exists a distribution called (GEV) that does only that, while it is not publicized it does a good job.
Rating: 1 / 5
I had great hopes for this book, because the basic premise is correct, that many people put too much weight upon averages, not enough upon variability, and when averages of combined or mixed events are taken, wrong conclusions are inferred. However, it presents the matter with demagoguery, Mr Savage championing himself as a Luddite, riding on the roadtails of post-economic-crash anti-intellectualism. The book is a sales platform, for Mr Savage’s consultancy, software, and for his role as the medium through which Truth be had, stripped of complexity.
While Dr Savage’s illustrations are often fine and well thought through, both on his Web site and in the book, he is often insulting and name-calling, and oversimplifies. He also breeds distrust in professional advice, and characterizes mathematicians and statisticians as if they were as useless as philosophers or priests.
In the end, by tossing aside some of the greatest thinking and results peoplehood has conceived, he is hobbling his audience, making them incapable of learning more EXCEPT THROUGH HIM, by convincing them these other sources and means are unnecessarily unreachable, and not worth knowing anyway.
He reminds me of a businessman’s version of Peter the Reader, the Nitrian monk, who, with mob of fellow acolytes, brutally slew Hypatia of Alexandria, in league with Christianity’s Saint Cyril. Except Dr Savage worships at the altar of extreme practicality and common sense.
Rating: 1 / 5
If you’re clueless about statistics then this book is for you but if you have superficial knowledge or more about statistics then this is a waste of time to read. The author promotes his website way too much and talks about so many irrelevant subjects. ie writes a paragraph about his father working with Milton Freidman. I don’t see how that is any way relevant to me understanding more about the flaw of averages. Again which I mentioned earlier, this occurs a lot throughout the book. The book could have been condensed into about 50 pages but if that was the case obviously the price of the book would drop dramatically. Ultimately, this is a book you should get from the library and use your savings to buy another worthwhile book. This is one of the worst purchases I have made.
Rating: 1 / 5
Bottom-line: this book introduces a lot of great concepts to those in the population (most) that are clueless when it comes to probability and statistics. I was pleased with the book’s contents until the author couldn’t hold back his liberal anger any longer and had to use the statistical tools that he introduced to push his social agenda. For example, he argues how missile defense is a folly, Reagan got too much credit, the war on terror is ill-founded, global warming is real, Bush-bash, etc.
I lost all respect for Mr. Savage once he wandered off the intellecutal road into the weeds of his liberal nirvana. I guess I should not have been surprised though because one look at his picture on the inside flap of the book gave a shining clue that this man was going to be far to the left of the political distribution.
Rating: 3 / 5
I’ve seen reviews of this book starting with “if you have no clue about statistics…” Unfortunately, that’s a lot of people, even worse, a lot of people who have no clue about statistics think otherwise. If you fall into this category, this book is a good read, which will expose for you a lot of interesting things about what you can and what you cannot do when your data are influenced by a chance. Even if you do, but you have let your knowledge get rusty, this book would be a good reminder.
Also, the actual material of this almost 500 pages book can be delivered in about 20-30 pages, which is a very good ratio for the most US published books.
However, I still did not like this book, specifically, for two reasons.
First of all, the author (according to his own testament) is not a professional mathematician, but software developer. That’s ok by itself, but as a result, while he has a good idea what he is talking about, he occasionally make mistakes or uses completely wrong approaches. For example, in one of chapters he introduces a device to get a random number between 0 and 11 (kind of roulette) and then instead of giving the distribution this device provides, he expects the reader to assume that (which is one of huge flaws when working with data samples) and even worse, to “prove” it without actual data — a certain “F” in any “Statistics 101″ class, and for a good reason.
Second, his style is not that conductive to learning, even if you can recognize when he misses the target. For example, to explain “black swans”, he gives an example with the same device, when “the arrow does not stop but breaks off and hit you in the eye”. The “eye” part is certainly over the board and an amateur educator may think it helps to remember, while it actually hurts digesting the material.
Rating: 2 / 5