The Model Thinker

"We confront no end of complex problems: why is inequality on the rise? Why are more and more Americans clinically obese? Does a racially diverse team make better decisions? How can we predict the outcomes of elections? At the same time, we find ourselves awash in data, be it on the opioid crisis, college admissions, genetic correlates of disease, financial transactions, or athletic performance. To confront such complexity and put that data to use, we need models: we can use linear regression to predict sales growth, or a power-law distribution to explain city sizes and book sales. Although each model offers insight, any single model will be wrong--just ask the physicist who, trying to understand barnyard animals, imagined a spherical cow. We must be able to do better. The question is simply how. In [this book], Scott E. Page gives us the answer: many-model thinking. By applying multiple diverse frameworks, we can achieve greater insights--indeed, using many models enables us to scale a hierarchy encompassing data, information, knowledge, and ultimately wisdom. Underpinning this, Page presents twenty-five broad classes of models--including models of growth, random walks, entropy, Markov chains, and many more--in a user-friendly and highly readable format, while teaching us how and when to apply them. Whether you work in science, business, government, or even literary studies, you confront complex problems, and you have more data than ever before.