This book shows how perceptive people have separated relevant data from irrelevant data.
The author is fond of baseball and baseball statistics. Not being fond of baseball, I had to remind myself from time to time that the book is about data and analyzing data and not about baseball. Sometimes the book moves a little slow, but it is full of good information.
This is not a mathematical book about statistics so much as a book about judgment calls and how to evaluate relevance.
The author's view on Bayes's theorem is particularly interesting. I hope he will elaborate on Bayes's theorem with examples in detail in a future book.
Nate Silver isn't an expert on statistics. As he explains himself in the book, he's an enthusiast who happened to strike gold when he applied statistics to fields where the competition was very weak. Silver's lack of expertise is very much in evidence throughout the book, in the form of two flaws: shallowness and inaccuracy.
Let's start with the shallowness. Most of the book is taken up by descriptions of various fields where statistical prediction has been applied with differing degrees of success: earthquakes, weather, politics, sports, and so on. The main point of these sections seems to be that pure statistics isn't enough -- you need specific knowledge of the problem in order to make predictions. That's a good point, though fairly obvious; to illustrate it, much time is spent simply describing these various fields. We get a description of how contracts work in baseball, a taxonomy of types of poker player, some information on the planning stages of the 9/11 attacks, and more.
Silver doesn't know very much about any of these subjects, so the result is a shallow and unfocused collection of trivia. Worse still, Silver's knowledge of statistics -- the subject of the book -- isn't very good either, so that topic gets a similarly vague treatment.
Worse still is the inaccuracy that plagues the book. I can't speak for the sections on baseball or terrorism, but I noticed many glaring flaws in the explanations of statistics. One important mistake is the treatment of the concept of bias. Bias is an important technical term in statistics, but Silver talks about it as though he were using the colloquial usage, in which bias is always a bad thing to be eliminated. In fact, bias is often useful and important in solving practical problems in statistics.
Another, particularly annoying mistake was the description of David Hume's ideas about induction. Silver insultingly claims that Hume's idea was that if a claim is not known with 100% certainty, it is a mistake to give it anything other than a 50% chance. This is obviously nonsense and unrelated to Hume's actual thoughts, which should have been given a much more thorough treatment if they were to be mentioned at all.
Despite all this, it's a reasonably entertaining book, and the narrator does an excellent job. But I wouldn't recommend it if your goal is to finish the book knowing more than you did when you picked it up.
people interested to hear about past events but not necessarily in a fun or entertaining manner
No, but it did turn me off from this writer
Very long details about historical events and how people failed in predicting them, which is quite obvious, since if those events were predicted they wouldn't have happened. It's good Audible has 2X speed so I was able to reduce my time wasted listening to this book. In short it's a book that has no added value whatsoever.
This book emphasizes how data should be addressed as the title says. Distinguishing between the signal and the noise that comes along. Though the author involves many endless examples along his personal interest, not many gives concise illustration of how interpretations should be made and how people failed in avoiding them. The poker, baseball, basketball, weather, and other topics give little or no insight on what the reader should be doing, which is not productive after 15+ hours of listening.
It was clear and generally well presented (accessible to a wide audience). I was well aware that people generally are not good at making predictions, relying too heavily on their heuristics and biases (e.g. failure to use prior probabilities as called for by Bayesian statistics); and the importance of putting "band widths" around probability estimates. I also thought that much too much text was devoted to each of the major topics covered (e.g. weather forecasting, political forecasts, economic forecasts, picking stocks, gambling/poker strategies, etc.). These sections could have been considerably shorter.
"Is that all there is?"
Generally disappointing, in that it did not expand my knowledge -- though I don't fault the book, as I've read widely in this area.
A really smart book on why predicting stuff is hard and a really good argument for an open mind. Seeing the world as it really is, rather than how we want it to be, is the hardest thing we will ever do. But we really should try harder.
One of the best audiobooks I've listened to so far
Nate Silver gives an advanced, yet comprehensible lesson in statistics using exciting real world examples of how statistics were used correctly or incorrectly in each case. Topics range from earth quakes to political elections, which he is most recenty famous for.
I haven't read the print version; I'm a listener exclusively.
Good to Great, Jim Collins: The idea, that research and analysis is key, before a conclusion can be drawn is a theme in these books.
He has a very good reading voice. I'm not sure if I've listened to him before, and as a reader he doesn't stand out among the good readers, but he's definitely in that group.
This isn't that kind of book; there is no story for to film.
This was exciting to listen to; I really appreciate good analysis before conclusions are drawn, and I feel like Nate did a great job in applying his claimed principles throughout the book.
Great narration. Nice to find someone who takes a dispassionate view of events.
Why We Make Mistakes
A guide to logical thinking and alalysis of data that should be required reading for everyone. Covers somewhat different territory from that first plowed by Freakonomics and Super Freakonomics, but just as insightful.