Book Recommendation: Thinking In Bets

I'm always looking for ways to improve my decision-making skills. That's why I picked up "Thinking in Bets" by Annie Duke. In this book, Duke, a former professional poker player and cognitive psychologist, applies the principles of poker to decision-making in all areas of life, including data science.

The main idea of the book is that we should think of decisions as bets, with probabilities and outcomes that we can estimate and improve over time. Duke argues that by adopting this mindset, we can avoid common decision-making pitfalls such as hindsight bias, overconfidence, and emotional attachments to our choices.

As a data scientist, I found the book particularly insightful because it emphasizes the importance of probabilistic thinking. In data science, we deal with uncertainty all the time, whether it's in the form of missing data, noisy signals, or unpredictable user behavior. By framing our decisions as bets, we can make more accurate and robust predictions, and we can also quantify the uncertainty and risk associated with our models.

One of the key takeaways from the book is the concept of "resulting," which is the tendency to judge the quality of a decision based on its outcome, rather than its expected value. In data science, this can manifest as overfitting or cherry-picking the data to support our hypothesis. By focusing on the process of decision-making and the probabilities involved, we can avoid this trap and make more informed choices.

When we experience a positive outcome, we tend to assume that our decision-making process was good, regardless of whether we made a good decision. Conversely, when we experience a negative outcome, we tend to assume that our decision-making process was bad, regardless of whether we made a good decision.

It's important to recognize the role of resulting in our work. We can use data to help us make better decisions, but we can't control all the variables that impact the outcome of our decisions. Instead of solely focusing on the outcome, we should pay attention to the quality of our decision-making process, including the data we use, the assumptions we make, and the models we build. By doing so, we can improve our decision-making skills over time and make better use of the data we have available.

Overall, "Thinking in Bets" is a must-read for data scientists who are interested in statistics and probability, and who want to improve their decision-making skills. The book is engaging, insightful, and full of practical advice that can be applied to any field that involves uncertainty and risk. I highly recommend giving this book a try.