BEYOND BIG DATA
A NEW WAY OF UNDERSTANDING AND ANALYSING RECENT DATA FUSILLADE.
IF YOU PICK UP a book on how to make data work for you, it is assumed that you like numbers and statistical terms. The Model Thinker by Scott E. Page is all about models – formal structures presented through mathematics and diagrams to help us understand why things happen the way they happen. In simple terms, models are tools which encapsulate daily-life problems and situations in mathematical terms, try to explain why they are happening, predict possible outcomes and help design potential solutions. The book discusses a wide array of models, how they are formed, their utility and limitations. Which means nothing is random in life. All our decisions and actions happen because certain variables function in a certain way and models teach us to put these variables in a mathematical form. These equations can be simple or complex based on the number of variables involved. Page does a great job of analysing 30-plus models, along with real-life examples. He also focusses on many-model thinking as single models often fail to provide a complete picture in this age of Big Data where the fire-hose of data flow leaves little scope for complex and nuanced analyses. For instance, a single model may point out that the poor state of national health in the US might have been triggered by the opioid crisis, but the current scenario is not the outcome of a single cause.
The real-life examples given here are intriguing. The book explains why the US government decided to bail out the American International Group (AIG) instead of the Lehman Brothers in 2008 – the choice made was based on a model. The author gives another example of how the French authorities used probabilistic models to locate the fuselage of the AF 477 that crashed in 2009. There are many more incidents which show how business-related issues could be accurately predicted by using specific models. But such things do not make the book a page-turner. All too often, readers drown in too many assumptions, logical conclusions and alphabetic symbols while trying to understand the basic premise of a model and what it can achieve. The best approach: Read this book as an academic exercise but certainly not for fun.