The Signal and the Noise is a cool book by Nate Silver. Silver, who is a statistician, talks a lot about predictions and the many things that relate to them. The book is centered around finding the useful information that allows people to make accurate predictions (the signal) and everything else that can mislead them or distract them from the important stuff (the noise). This has applications for a wide range of disciplines, including economics, sports, gambling, and weather forecasting.
Silver has several good points for predictors, whether professional or hobbyist, to keep in mind. One of them has to do with their ability to make predictions. Some predictors think of themselves are more skilled than they actually are, which can result in them making extremely bold predictions that completely miss the mark. Even when they’re making predictions about common events and have a great deal of information about those events, they still deal with a lot of uncertainty. They often fail because they either overestimated or underestimated the probably of something occurring. For example, if someone regularly sees a lot of news reports about crime, they may think that crime is more common than it actually is. On the other hand, they may be totally unaware that something else ever occurs until it breaks into their world and takes them completely by surprise; in this case, they would be unlikely even to make a prediction about it in the first place since it would be an unknown unknown to them.
Something else that greatly affects the accuracy of someone’s predictions is the quality of the models they develop for making predictions. The more information someone has, the more accurate their predictions should be; bad models make worse predictions when they get more information. So if someone has access to a lot of information but they continually make incorrect predictions, they should either improve their existing models or abandon them in favor of better models. On the subject of accuracy, Silver encourages predictors to make the most accurate predictions they can based on the information they have and the models they use. They shouldn’t adjust their predictions in the hopes of becoming famous, advancing their careers, making a name for themselves, or otherwise trying to win favor from other people. Valuing accuracy over notoriety may not get them as much attention, but it sharpens their skills and allows them to better understand and navigate the world, which is much more useful and important than fame.
One of my favorite parts of the book dealt with Bayes’ theorem, which is a way to determine the probability of a given event. This requires having information about past occurrences of that event (if there are any) and making reasonable estimations of a few related pieces of information before plugging those numbers into a formula and getting the answer. I particular enjoyed this subject because it comes the closest I’ve ever seen to describing a formula for accurately predicting the future. This also ties in with an point from Pierre-Simon Laplace that Silver discusses earlier in the book: if we had perfect knowledge of all particles in the universe and all the ways that they’re affected by universal laws, we could make perfect predictions. Until that point, we can strive to get our predictions as close to perfect as we can even if we never achieve perfection.
The Signal and the Noise was a fun read for me since I spend a great deal of time thinking about the future and enjoy finding things that increase my chances of being able to correctly predict it. One of my all-time favorite books is The Fourth Turning, but that book is more about several predictions that the authors made and less about the different elements that are involved in making predictions. As far as I can remember, this is the first book I’ve read that explores what goes into making predictions in great detail, and learning this stuff was a lot more enjoyable than I initially thought it would be. Since I started reading the book, I’ve noticed myself trying to read a variety of situations and correctly predict their outcomes to an even greater degree than I usually do. I don’t know if I’m any better at this than I was before I read the book, but I’m having fun doing it. And, as Silver says toward the end of the book, the only way to get better at making predictions is to make a lot of them and hone in on what works while ignoring the rest. So if you’d like to learn more about this stuff and see how it affects your view of the world, I recommend checking out The Signal and the Noise.