Scan barcode
yash_verma's review against another edition
5.0
Must read if you are living in the Bayesian land... A book showing all the practical aspects of probability in various fields...
bentrevett's review
4.0
Everyone making any predictions is dumb, unless they’re using Bayesian statistics, then they’re a genius
tiarala's review
3.0
Interesting insights bogged down by too much granularity. Left this book feeling like I learned a lot, but I know I zoned out in parts because it just went into too much detail about things that I had no personal context for. Would've appreciated a more general look at predictions rather than specific cases, especially when it came to professional sports chapters.
slichto3's review
3.0
The Signal and the Noise is an ambitious project. Written by professional forecaster Nate Silver, creator of the blog fivethirtyeight.com, it strives to describe the successes and failures of famous predictions, as well as to provide guidance on how to create good predictions in the future. As Silver describes in the book, good predictions are hard. Really, really hard. And history is riddled with failed predictions, even when the people making these predictions are very intelligent (in fact, it seems that sometimes greater intelligence can lead to greater failure). However, there are some basic lessons that we can learn about why predictions fail and how we can make better descriptions. The Signal and the Noise details some of these lessons, but it keeps most of these lessons broad. That is the largest weakness of the book. It entertainingly shows a large number of predictions, successful and unsuccessful, and some lessons that we can learn from those predictions, but it stops short of providing much beyond these broad lessons.
The chapters cover a number of fields that require predictions: the stock market, poker, sports, elections, weather, and earthquakes are just a few. If you are unfamiliar with any of these fields, Silver’s descriptions can be fascinating. I very much enjoyed reading about his analysis of the housing bubble in 2008, as well as his objective exploration of climate change. Silver is a decent writer and appears to be a very good analyst. Silver also covers a set of best practices for predictions, including being a “fox, not a hedgehog,” balancing complexity and simplicity, and, most emphatically, using Bayesian analysis.
All of this is interesting, but that’s about as far as I would go in describing the content of The Signal and the Noise. There’s just not very much that you can actually use. Silver stays fairly broad, for the most part, and this is even more apparent if you already have some basic familiarity with a topic that he’s describing. For example, he has a chapter on gambling on the results of NBA games. He describes a gambler who’s managed to make a lot of money on the NBA, as well as the basics behind Bayesian analysis. Essentially, though, the key to the gambler’s success came down to paying a lot of attention to the NBA. He attempted to make probabilistic guesses about the results of the league. Then, he compared his analyses to the odds in Vegas, and if they were different, he bet. There’s not much else. If you thought about it, you probably already knew this. So what can you do with this knowledge? Pretty much nothing except try to think more probabilistically about the world. And that’s the problem with this book – most of its wisdom can be described in probably a page or two. The rest is interesting stories, which are fun and a little educational, but will not give you any groundbreaking insights.
If you’re looking for a primer to making statistical predictions and some interesting stories, then this book is great for you. But if you’re looking for a more in-depth analysis of how to make good predictions, then you probably will want to look somewhere else. Still, The Signal and the Noise is reasonably enjoyable read for most.
The chapters cover a number of fields that require predictions: the stock market, poker, sports, elections, weather, and earthquakes are just a few. If you are unfamiliar with any of these fields, Silver’s descriptions can be fascinating. I very much enjoyed reading about his analysis of the housing bubble in 2008, as well as his objective exploration of climate change. Silver is a decent writer and appears to be a very good analyst. Silver also covers a set of best practices for predictions, including being a “fox, not a hedgehog,” balancing complexity and simplicity, and, most emphatically, using Bayesian analysis.
All of this is interesting, but that’s about as far as I would go in describing the content of The Signal and the Noise. There’s just not very much that you can actually use. Silver stays fairly broad, for the most part, and this is even more apparent if you already have some basic familiarity with a topic that he’s describing. For example, he has a chapter on gambling on the results of NBA games. He describes a gambler who’s managed to make a lot of money on the NBA, as well as the basics behind Bayesian analysis. Essentially, though, the key to the gambler’s success came down to paying a lot of attention to the NBA. He attempted to make probabilistic guesses about the results of the league. Then, he compared his analyses to the odds in Vegas, and if they were different, he bet. There’s not much else. If you thought about it, you probably already knew this. So what can you do with this knowledge? Pretty much nothing except try to think more probabilistically about the world. And that’s the problem with this book – most of its wisdom can be described in probably a page or two. The rest is interesting stories, which are fun and a little educational, but will not give you any groundbreaking insights.
If you’re looking for a primer to making statistical predictions and some interesting stories, then this book is great for you. But if you’re looking for a more in-depth analysis of how to make good predictions, then you probably will want to look somewhere else. Still, The Signal and the Noise is reasonably enjoyable read for most.
owook's review against another edition
3.0
Interesting overall, though some sections really beat a topic to death. I would give it four stars but it took me months to slog through.
jlbates's review
4.0
Good mix between some math but not so much that it puts you to sleep. I liked how it was broken into sections (poker, economy, earthquakes, weather, 9/11, etc) that make it relevant to the world. Overall I'd recommend it with the understanding that it won't necessarily tell you how to make predictions on your own, but will make you appropriately skeptical (and simultaneously more kind to weather forecasters).
millerj1326's review
informative
fast-paced
4.5
In my opinion, a very good explanation of statistical concepts in a real-life framework!