Naked Statistics: Stripping the Dread from the Data:

 

 

About 50 pages into this book I thought to myself; ‘they should teach this in high school’; the ‘this’ being the general statistical literacy which the book is about. Very shortly thereafter the prescriptiveness and condescension of that statement hit me and and I thought better of sharing it openly with anyone. Also, it made me realize I’m old. My personal feelings aside, Whelan’s clarity of presentation and methodical march through the basics of applied statistics is broadly approachable. And funny, to boot. The standard examples like the Monty Hall problem and laughable reverse causality situations all make an appearance. And so do new ones, like the Schlitz Beer’s taste test ploy in the 1980s, which, though commendable, couldn’t save the brand.

All humor and Stats 101 aside, there is deeper layer, a line of argument that should resonate with data and analytics professionals. It is that incorrect, misleading and malfeasant uses of statistics are not the fault of the math itself, but rather the application. In short, no amount of sophisticated modeling and analysis can overcome bad data, biased sampling, poor choices of variables and a faulty research design. More conceptually and, perhaps, normatively, Whelan argues that our thinking about how to use data has lagged seriously behind our ability to gather and analyze it. Put another way, we can put a lot of effort into building very precise models — getting the data, sampling and variables correct — only to find out later that they are not very ‘accurate’ in the sense of solving the problem we originally intended to.

Whelan equates data to offensive lineman. As such, they are a necessary but not sufficient condition for creating an accurate, useful and honest statistical analysis. Machine learning as a discipline seems to be coming around to this view as well. Modeling improvements still matter, but rarely will they improve performance to same level as greatly increasing the size of the data. Amidst the continuous improvements in computing power, deep learning and AI in general, it can be easy to forget the importance of asking the right question and collecting the right data. This book is an enjoyable read that provides a gentle reminder of why we do analyses in the first place and how to ensure that the fundamentals — the lineman — are both efficiently operating and dutifully respected.


   

How To: Absurd Scientific Advice for Common Real World Problems:

 

 

A decade ago, around a campfire, after many beers, we pondered the horizontal pressure of an above ground swimming pool. How much force does the water exert and was it greater for a large but shallow pool or a tall, but small (in circumference) one. Related, is it more difficult to displace water in a small versus a large pool? While we did not answer those questions that night, we did successfully melt an aluminum can and boil water in a paper cup. If you’ve ever wondered these type of wonders or experimented these type of experiments then this is your kind of book (the pool question is covered, btw). Basically a ‘deeply-researched and somewhat obscene answers to questions you are usually too sober or too embarrassed to ask’ book.

Munroe is the author of the XKCD comic strip; well known in some circles, likely unheard of in most. Here he presents legit science wrapped up in similar dry humor. There are equations, but if you skip them you won’t miss anything. There are also drawings, which shouldn’t be passed over. The formula is: 1) Insert somewhat innocuous question about how to do a standard human thing; 2) Take said question to the extreme; and 3) Supported extreme answer with math and jokes. And it works. Rare is the book that can handle Kant and Evil Knievel in successive chapters.

What else is covered? Popular Culture: Is the dig at Oak Island worth it?; Popular Wisdom: Is ‘red at night’ really a delight?; and Popular Mechanics: How to best disable a flying drone. If you still aren’t convinced, you’ll learn how far you can actually throw a given object…and for that alone you should crack the cover.


   

The Subtle Art of Not Giving a Fuck:

 

 

Online reviews of this book range from “Dude, totally best book ever” to “This book is a poorly done ‘Buddhism for Bros’”. I fall somewhere in the middle. It contains useful wisdom, wrapped, occasionally a bit overly so, in modern context and lexicon. The title — clearly a marketing win — sets the expectations of readers; which depending on how far down the self-help (or Psychology PhD) path one has trod could be a positive or negative push.

I took the following from this book, presented in what I gathered was the author’s line of reasoning:

This is all, of course, aimed at well-to-do, existentially-confused young westerners whose problems do not include access to clean water, malaria, famine, war, despotism and other life threatening circumstances. It is a quick read and, for those of us lucky enough to have some say in what we ‘give a fuck about’, a worthwhile reminder of the importance of setting the right values and in putting in the work necessary to nurture, express and live those values.


Prisoners of Geography:

 

 

Don’t let the title fool you, this is as much or more of a history book than a geography book. While the book is organized into chapters based on ’10 maps that explain the world’, with a few exceptions, there is almost a complete lack of interesting or useful maps to help the narrative. The ’10 maps’ are simply ten countries/regions of focus — each presented with a very standard political geography map. I spent no small amount of time referencing Google Maps during my reading and feel a sizable contribution of this book could have been in clearly locating (mapping) the forces driving each country’s decisions, both past and present.

The tagline here should be ‘selected geopolitics of the 20th and early 21st century’. With that aim in mind, it is a successful book: approachable, yet detailed. An undergraduate modern world history seminar could do well by assigning this as a supplemental reader. A few notable themes that emerge: 1) The disastrous consequences of colonial powers drawing borders without regard for historical settlement patterns; 2) The extractive development of the Southern Hemisphere leading to large coastal capitals and underdeveloped hinterlands; and 3) The continued importance of proximity in diplomacy (Moscow is near, Washington is far).

Many of the (assumed) political motivations are discussed under the umbrella of energy politics. Either I’m naive about the supremacy of energy in driving (geo)politics or this book takes an overly simplistic view. Truthfully, a mass market non-fiction book probably cannot do justice to the full suite of considerations; nonetheless, Marshall leaves the reader feeling as if the sole determinant of a state’s actions are oil and natural gas. Barring the final chapter on the Arctic, climate and ecological issues are given short shrift. Technological capital and labor markets are thinly discussed as well. Again, perhaps state machinations really are primarily driven by the quest for hydrocarbons…and if so, it would explain the world’s reluctance to address climate change and support alternative energy solutions.

All in, it is a quick and worthwhile read. Many factoids and smaller points of information may surprise — ex. the British traded their overseas bases to the Americans for warships in 1940 — while the general themes are broad and, somewhat evident. I was left with a sense of foreboding; a collision between the US and China in the Pacific seems imminent as does Russia’s dominance of the Arctic. Geography may end up being a cruel jailor.