on 1 April 2013
Math On Trial is a study of several criminal cases where flaws in mathematical and statistical calculations and their analysis led to incorrect verdicts of guilt or innocence. I was surprised how much I enjoyed this book!
Taking a series of riveting cases, including murders, rapes and sex discrimination, Schneps and Colmez both chart the history of the use of math in court, and clearly reveal common fallacies in its use. And it is un-put-down-able stuff.
Never have I been this enlivened by mathematics! I found myself tearing through the book, finishing within a few hours, and was particularly struck by several landmark cases that I had glimpsed in the news but not paid much attention to before.
An accusation of murder by Munchausens Syndrome by Proxy leveled against Sally Clark, whose first two infant children died seemingly of SIDS (cot death), is quite irrefutably repudiated by the authors, and the chapter, like all those in the book, is presented in a style that paints a clear picture of the context of the case, so as a reader I felt I had enough knowledge of the entire situation to be able to draw my own conclusions about the verdict.
The book, however, is not unbiased, and does not pretend to be so, highlighting instead bias in the original cases and statistics. The point of view of Math on Trial then, could be said to be biased towards unbias! With this in mind, one of the strengths of the book is how gripping and emotional it is, for a work of non-fiction. I was saddened by the impossibly heartbreaking situation of Sally Clark; I was horrified by the mistakes made in the case of the several rape/murder trials discussed in the book; throughout the reading, I was induced to continue because of my attachment to protagonists and horror at the injustice.
Undoubtedly the most famous case Math On Trial studies, and certainly the one that may well attract the most attention, is the death of Meredith Kercher, and the overturned verdict of guilty for Raffaele Sollecito and Amanda Knox. Using simple probability calculations, the authors make a brilliant case for allowing specific DNA testing into evidence, which had previously rejected by a Judge in appeal court in 2011.
On an entirely separate note, I must mention the design of the book: a lovely hardback, with beautiful paper stock and a gorgeous cover designed to look like a retro thriller. `Props' to the publishers at Basic Books for several extremely good choices!
I would recommend too for parents trying to support teenagers in their studies of mathematics - or in fact, law - as the book, although written at an adult level, puts complex concepts very simply and does not show anything brutal in pictures or go into more disturbing details than you might find on the news.
on 27 March 2013
I approached this book not really knowing what to expect.I'm not a mathematician and so was worried it would be a little bit over my head. However I was very pleasantly surprised when I found that it explained the theories in a user friendly way that really brought the maths to life, and gave me some law knowledge too.
To be able to take such complex topics and distil them in such an interesting way is a real skill. It reminds me of Alain De Botton's work.
I'm looking forward to seeing what these guys come up with next..
on 16 April 2013
I think this is one of the best books I have read in decades. Not only is the math(s) absolutely riveting, but the stories themselves combine excitement with fantastic detail. Other commentators have stated how the book is impossible to put down - I totally agree - better than many a crime thriller. I spent an entire Sunday reading the first 9 chapters and finished the book the next day. I then read about how the authors had researched their material; their dedication seems to know no bounds. Just as an example, I had a rough idea of the story of the trial of Alfred Dreyfus but had no idea of all the twists and turns until I read this book. Yet this was researched purely for the joy of knowing the whole story - just to recount the mathematical aspects would have required a much less through job. This writing duo are wonderful and I eagerly await their promised next book!
on 16 May 2013
I loved this book! It was interesting to learn how maths can play an important part in our courts, provided it is used correctly. This book includes several real-life examples of when it hasn't been applied correctly, and for some people it has resulted in them being found guilty for appalling crimes when in fact they were completely innocent. Don't be put off if you are a non-mathematician (like me!), it's written in plain English and flows incredibly well.
Definitely one I'd recommend!!
on 29 June 2014
I'm not a mathematician, but I enjoy reading about maths, and trying to understand how results are obtained, even if the computation is beyond me.
The ten chapters here describe ten ways in which maths has been used and misused in criminal trials. I remember the case of Sally Clark well, and I remember being disturbed when reading the newspaper reports of her trial; but I didn't know exactly why I should be so disturbed. This case is the first chapter; it shows how it is quite possible to provide totally deceptive statistics when the concepts are wrongly applied. It's about a mother and two of her children who died; did she murder them, or were they both victims of 'cot death' or SIDS? (At which point I must point out that one of my kids died of cot death, so I may not be totally impartial. Anyhow, maths were used to show that the probability of two cot deaths was next to impossible, therefore it must be murder. The mathematical process was flawed, but her eventual aquittal was due to the finding of significant bacterial infection in the second child, and not because the court accepted that the maths was wrong. Indeed, the problem is simply stated; the figure of 1:8000 or so as the probability of a single cot death was squared to for the probability of a second. The experts and the court clearly could not see that a statistical probability is not a cause, it's just a description. And if the cause of cot death is unknown, this doesn't mean that there is no cause; from which it could be argued that multiple cot deaths in one family could be more likely than not, as whatever genetic and environmental factors are involved in the first could equally be involved in the second. And that's before entering into why the statistics were flawed.
Other chapters aren't quite so mathematically challenging, though it can be hard to follow the arguments at times; the book isn't always an easy read for the non-mathematical.
I don't really remember the case of Lucia de Berk, the nurse in Holland accused of murdering several of her charges. Here we are shown the mathematical failures that led to her being found guilty. One of these failures was looking for 'suspicious' deaths in the hospitals where she worked, but this was really looking for deaths where she was on duty. All deaths, suspicious or not, were not looked into; it's clear that the investigators had a mind-set before the detailed scrutiny of the records; if Lucia was present, the death was 'suspicious', if she wasn't there, the death wasn't suspicious.
In other circumstances this would be called 'confirmation bias', looking for things that fit your pre-conceived notions and ignoring the rest. As this is a book of maths, such psychological problems aren't really dealt with in any depth, though it might be an idea for a collaborator to try to flesh this out in any future edition.
And the authors hope that there will be future volumes, examining other cases where the mis-use of maths has produced an incorrect verdict.
There's a short tail-piece about the future of maths in court; either it's too complicated for the court and the jurors to understand, or the correct application of statistical techniques will be very useful—specially in cases of DNA matching. One message that does come through, sadly, is that judges do ignore statistical evidence at times; perhaps it's too hard for them, perhaps maths isn't somehow appropriate to the majesty of the law. Their reasons for ignoring what may be compelling mathematical evidence seem at times to be quite arbitrary; the arbitrariness of what they will and won't allow to be discussed or to enter their thoughts is yet another problem.