In the middle of last year I started reading non-fiction books quite regularly. Most of books relevant to computing education have already been covered in previous posts: Teacher Proof by Tom Bennett, Why Don’t Students Like School by Dan Willingham, Mindstorms by Seymour Papert, Visible Learning by John Hattie, and Peer Instruction by Eric Mazur. (Are there many good non-fiction computing books? Seems like the economists and psychologists have managed to produce best-sellers, but not computer scientists.)
Instead, this post features mini-reviews of seven books not relevant to computing education that I’ve read in the past year:
The Signal and the Noise by Nate Silver. At a previous job, we used to play poker at lunchtimes. Once an argument broke out over whether you judged a decision (raise/fold) on the information you had available at the time, or the eventual outcome of the hand. Were you right to fold against a high bet when only holding a pair of eights because it seemed unlikely you would win, or wrong because you later would have won when the third eight appeared? I see Silver’s book, which features a chunk on poker, as a long explanation of why the decision must be judged against the available evidence, not the outcome that unfolded later. Most practical outcomes have an element of uncertainty, and Silver argues that people are bad at recognising and accounting for the uncertainty. We may want to predict the election or the weather, or some other outcome of interest. Realistically, we can only offer a probabilistic prediction: 80% chance of rain, 75% chance of Obama’s victory. If it later turns out there was no rain, that does not mean the prediction was flawed. Moreover, any certain prediction — it will definitely rain today, Obama will win the election — is flawed or useless, even if it appears better and more useful than the hazy probabilistic prediction.
The Gamble by John Sides and Lynn Vavreck. The Gamble concerns the determining factors behind the 2012 US election result. It is chock full of data, and offers a relatively unexciting explanation for the election result: there were no game-changing gaffes, no big swings, no canny campaigning. The election was decided by fundamentals (e.g. the economy) and two juggernaut campaigns that cancelled each other out. I find The Gamble an interesting book in the wider scheme of things. I must admit that I would have preferred to read a dramatic blow-by-blow account of the debates and gaffes (binders full of women! that 47% remark!) and so on, but The Gamble seems to be the correct answer: a slow and steady data-based account of the election that gets to the heart of the matter. I guess evidence-based rationalism means that you have to live without unnecessary excitement.
Liar’s Poker, and The Big Short, both by Michael Lewis. These are excellent as a pair. Liar’s Poker was written in the ’80s by Lewis after he quit his bond trading job and wrote a tell-all book about Wall Street. It explains how, in the ’80s, mortgages went from simple affairs between the building society (or equivalent) and the borrower, to being traded as bonds on Wall Street. It details that a lot of traders were immature, unscrupulous, and often trading on their customers’ ignorance to make fortunes. Of course, repackaging and misrating of these mortgage-backed bonds was a trigger for the recent economic crisis, so it was no surprise that Lewis returned to the same area to write The Big Short, which focuses on the oddball few who correctly predicted the meltdown ahead of time and found clever ways to make money off this (more on this in a moment). Several of them were driven close to a personal meltdown, because they could not tally their evidence-based prediction that the market would crash heavily, with the behaviour and beliefs of all the other traders in the industry, who seemed to be happily sailing their ships over the cliff. It gives you some insight into why when everyone is wrong, it is hard to be right.
This pair of books also leads to an interesting observation. In Liar’s Poker, Lewis did not portray brokers well. They mostly appear in the book as a bunch of conniving (although admittedly canny), greedy, immature louts. The book was a fairly damning exposé, and Lewis’ ex-CEO would later claim that it “destroyed his career”. So there Lewis is, 20 years on, trying to write a second book about Wall Street. Who would dare talk to him, knowing all this? NPR asked him in an interview, as described by Tim O’Reilly:
You have to understand, Lewis replied (more or less), that many of those people got into the financial industry after reading [Liar's Poker]. Their big takeaway was how easy it was to make a lot of money without regard to the niceties of creating much value. He finished with the memorable line, “You never know what book you wrote until you know what book people read.”
Bad Pharma by Ben Goldacre. Bad Pharma is an examination of how medical research and development is set up wrongly, and is providing opportunities and incentives for companies to act against patient’s interests, often with collusion from regulators and doctors. Some of the bad-regulation aspects chime with The Big Short, where regulators seem beholden to the companies they are intended to regulate. It’s an important book and one worth reading, as a general read, and if you are involved in (any kind of) research. The practices to do with authorship on papers and ghost-writing, and the responses by some journal editors, are quite eye-opening.
Marvel Comics: The Untold Story by Sean Howe, and Brick by Brick by David Robertson. Two similar books, one about Marvel Comics and one about Lego. My slight frustration with both books is that I was somewhat hoping for an indulgent read all about the evolution of Technic and Spider-Man, but instead both books focus heavily on the political and business sides of Marvel and Lego, with surprising little detail on the actual products of the company. You read all about the politics that forced a change in writer of the X-Men, but not much about the plotlines, or evolution of the art and colouring. You read about the background to the changing CEOs at Lego, but get little description of the differing sets and bricks beyond their general sales figures. Clearly, I was still picking too much from the serious end of things!
You do come away with a sense of how Marvel shoot themselves in the foot by driving short-term sales (limited edition covers, cross-overs) over quality or long-term loyalty — and they continued to do this even though they knew it hurt long-term sales. Also interesting is how both companies nearly drove themselves to the point of irrelevance and bankruptcy: Lego were saved by refocusing and getting smarter about their business processes; Marvel were largely saved by their recent adventures into blockbuster films. The Marvel book also has a fairly dark undertone: steadily thoughout the book, lots of employees seem to die early, to the point where it no longer seems to be confirmation bias. One illustrator dies at the drawing board of a heart attack, aged just 32. It’s a little like seeing a documentary about the sweatshops where your clothes are made: a hidden human cost to your fun. I suspect there would be similar tales from a book about the computer game industry.
That’s the lot for now! I’ve got a few more books left on my reading pile, but recommendations for others (relevant to computing education or not) are welcome.