It’s an intriguing title!
How Not to be Wrong has a lot to say about the way Maths and statistics are used and abused. On the way it covers various mathematical issues and theories, making them simple and pretty non-mathematical for ordinary readers. Its writer is clearly a leftie leaning liberal – so that was good – it’s always nice to have your own prejudices confirmed.
One of these prejudices was with respect to the tax rate / revenue curve – known as the Laffer Curve, which makes an appearance in Ferris Bueller’s Day Off. It’s a common claim of right wing politicians that increasing taxes results in a lowering of tax income. Eilenberg debunks this myth in chapter 1 – page 28 in my Penguin Edition, to be exact. Look it up all you tories and republicans! It was worth reading this book for that fact alone – handy ammunition for my next rambling conversation with my English accountant friend.
There are other real life issues where the truth of maths (or math as they call it over the pond) impinges on prejudice and leads to misapprehension. Ellenberg deals with some basic points about statistics – such as the need to use common sense to interpret stats, or the fact that not all graphs are straight lines so that extrapolating from data can be a dangerous occupation. This leads to quite a bit of detail about the confidence scientists can have in statistics and the range of error that is acceptable in experiments – a bit more technical, but explained very clearly.
Ellenberg looks at issues such as Bible Codes, where the names of rabbis can be found in mathematical analyses of the first five books of the Bible. He points out all kinds of mistakes in this kind of analysis, including examples of a similar occurrence of rabbinical names in War and Peace, and strange coincidences about assassinations predicted in Moby Dick. Ellenberg puts all this down to chance and the effect of large data groups where random analysis can pick up strange coincidences. The most amusing example he uses is of the dead salmon: apparently tests of brain activity in dead salmon show that they can recognise human facial expressions. Ellenberg attributes this to the random firing of the millions of dead brain cells, arguing that with so much data to look through you can find evidence of almost anything you want to find.
Other chapters in this book deal with chance, gambling, lotteries and the existence of God. There is a lot of detail about lotteries explaining the odds against winning and the actual value of a ticket – its true worth dependent on its chances of winning. Later chapters deal with multidimensional geometries in the clearest way that I have ever read.
As an aside, there is an interesting comment about stock market funds and the way they build in performance to impress:Ellenberg claims that dealers start many different funds in-house, allowing them to develop, and then selecting only the ones that perform well to sell to the public. In this case the past record is of no value in predicting future performance, as the dealers simply kill off badly performing funds before launching them, selling only those with an apparently good track record – even though this was achieved by chance and not the good judgement of the dealer. The book is full of these kind of gems.
This is an American book, and that’s especially obvious in the opening chapters where nearly all the references are to American topics. And of course as this is a book that says quite a lot about statistics there’s baseball too. Some of the phrases here mystify me – first up at the bottom the ninth or something!! Very confusing.
Later it gets more culturally diverse, and even touches on Pascal and Voltaire, much more European.
If you want to know about Maths and statistics I can certainly recommend this book to you.