Nate Silver's The Sound and the Noise is an overview of the art and
science of making predictions -- why we're so often bad at it, how we're
learning to get better, and why it's such a hard thing to do in the first
place.
There is less here than you might expect about politics and polling, the fields in which Silver has become famous. Instead, Silver looks at what we know and have learned about predictions in areas like climate change, baseball, earthquakes, poker, the stock market, and weather forecasts.
In some areas, our ability to make good predictions has gotten much better in recent decades; we are able to predict the weather (for the next week or so, anyway) far more accurately. In other areas -- earthquake prediction, for instance -- we are least learning to acknowledge that we are very far away from being able to make accurate predictions, and accepting that fact is a sort of progress in its own right. And in some areas -- the stock market, or political punditry -- even if some experts continue to claim that they can predict the future, the evidence suggests that they'd be just as well off flipping a coin.
Silver's style is smart but accessible, and when he does dive deeper into statistics than readers might be comfortable with, he's very good at explaining the concepts and tests he's using. This is an entertaining and useful book on a fascinating topic.
There is less here than you might expect about politics and polling, the fields in which Silver has become famous. Instead, Silver looks at what we know and have learned about predictions in areas like climate change, baseball, earthquakes, poker, the stock market, and weather forecasts.
In some areas, our ability to make good predictions has gotten much better in recent decades; we are able to predict the weather (for the next week or so, anyway) far more accurately. In other areas -- earthquake prediction, for instance -- we are least learning to acknowledge that we are very far away from being able to make accurate predictions, and accepting that fact is a sort of progress in its own right. And in some areas -- the stock market, or political punditry -- even if some experts continue to claim that they can predict the future, the evidence suggests that they'd be just as well off flipping a coin.
Silver's style is smart but accessible, and when he does dive deeper into statistics than readers might be comfortable with, he's very good at explaining the concepts and tests he's using. This is an entertaining and useful book on a fascinating topic.
No comments:
Post a Comment