The Model Thinker

Data, data, data: It's all one ever hears about these days. Science is all about big data. Our bosses call out for analytics, whatever those might be. And everyone wants to predict what will happen next. Can we accurately predict if a company's stock will rise, whether or not a disease will spread, or who will become the next President of the United States? As anyone who has ever opened up a spreadsheet groaning with weeks, months, or years of data knows, numbers aren't enough: we have to know how to make them talk.

Enter Scott Page and The Model Thinker. A leading professor of quantitative social science at the University of Michigan, he has taken his expertise as both a teacher and researcher and distilled it into the one book anyone will need to master data and turn it to professional use. This is no armchair exercise in imagined understanding, like The Signal and the Noise or The Black Swan or a legion of books on networks, the purposes of which are to make us look good in meetings (or in our own minds) than they are to enable us to do something useful. The Model Thinker is the guide to turning data into understanding. Underneath it all is what Page calls the "many-model paradigm", where the key isn't to just find one related set of statistical tools and work with them over and over, but to test our understanding of things by modeling them from several perspectives. The result is both a deep, quantitative acquaintance with tools ranging from Markov chains to game theory to Taleb-style long-tail statistics to network analysis and complexity theory, and a profound trip through the thought-process of a world-class data modeler. All the major tools of modeling--which readers will have heard of in everything from Wired to The Economist to The New York Times--will finally yield their secrets.

As The Theoretical Minimum showed, readers in quantitative fields aren't just looking for entertainment. They want to change their understanding of, and ability to act, in