Library:The Signal and the Noise
| The Signal and the Noise: Why So Many Predictions Fail—but Some Don't | |
|---|---|
| Author | Nate Silver |
| First published | 2012 |
| ISBN | — |
| Role | Analysis |
| Domains | Science & Space Politics & Elections |
| Tactics | Statistical & Evidentiary Distortion |
The Signal and the Noise: Why So Many Predictions Fail—but Some Don't is a 2012 book by Nate Silver.
Summary
Nate Silver, the statistician who founded the election forecasting site FiveThirtyEight, examines why forecasts fail in fields from economics and earthquakes to elections, poker and weather. He argues that the explosion of data has made it easier, not harder, to mistake noise for signal.
The book's lesson is that good forecasters think probabilistically, update their beliefs as evidence arrives and are honest about uncertainty.
Key points
- Overfitting. Models that fit past data too closely find patterns that are not real and fail to predict the future.
- Weather as a model. Weather forecasting improved dramatically because forecasters test predictions constantly and admit uncertainty.
- Bayesian thinking. Silver argues for starting with prior estimates and revising them in light of new evidence.
Relevance to Bullshitopedia
Many false claims rest on confident predictions or on patterns found in noise. Silver's framework helps readers judge how much a forecast or statistic should be trusted.