Learn more Shop now Shop now Shop now Shop now Shop now Shop now Learn More Shop now Learn more Click Here Shop Kindle Amazon Music Unlimited for Family Shop now Shop now



There was a problem filtering reviews right now. Please try again later.

on 24 December 2016
There were interesting nuggets in this book but I was left with an uneasy feeling that at the end I didn't understand what is the Principle of improbability nor what some of the various Laws actually stated.
So far as the main principle is concerned, does it actually say anything more than that some things are less improbable than they seem at first sight? This pales into insignificance compared with a principle such as the Uncertainty Principle.
The Laws are quite a rag bag. The Laws of Large Numbers and of Truly Large Numbers gave real insights derived from his expertise in probability, though a more thorough discussion, of why truly large numbers may be smaller than large numbers, would have helped. On the other hand the Law of Selection seems to be just a heuristic for weeding out false claims. Its relationship to Darwinian Natural Selection is tenuous; and his first example, explaining why some hypothetical insects migrate in response to a changing climate, is not an example of evolution at all.
Actually, I thought he was a bit short of material, and his subject does not justify a whole book. As a whole, it is a bit lightweight. But if you don't expect too much, it is a good read.
0Comment| One person found this helpful. Was this review helpful to you?YesNoReport abuse
on 21 September 2017
Love this book so much in exploring probability and how your assumptions and initial thoughts can be challenged. Recommend highly! Covers many interesting concepts.
0Comment|Was this review helpful to you?YesNoReport abuse
on 12 July 2015
If you can't stretch yourself to reading 'Thinking Fast Thinking Slow' you could do worse than this.Its basically an idiots guide to statistical analysis and dispels the curse of magical thinking effectively, by combining several probability theories.
0Comment| One person found this helpful. Was this review helpful to you?YesNoReport abuse
on 29 April 2017
Many of us, including me, are puzzled by the occurrence of "rare events". This book shows why most of them may be rare, even "unbelievably rare" but were to be expected, nonetheless. Very interesting subject. Very authoritative treatment.
0Comment|Was this review helpful to you?YesNoReport abuse
on 30 July 2015
Clearly written, very accessible account of what should really be considered crucial basic information for understanding the world, i.e., why many improbable-seeming events are in fact not terribly unlikely and may even be inevitable -- as well as why you can pretty much forget certain future possibilities because you're so unlikely to see them that they are effectively impossible.

I have a couple of minor beefs, one of them probably overly pedantic and the other one perhaps more substantive. The first is that, for a guy who's trying to introduce some intellectual rigor into our day-to-day analysis of what's going on, he's a little bit loose with his language sometimes. Possibly the most egregious example is that he uses the Law of Selection in two completely different ways. This law is introduced as a kind of hindsight bias, where we remember everything that fits our theory and conveniently forget things that contradict it. For example: "it always rains when I forget my umbrella" is based on only the times when it rained and you forgot your umbrella, ignoring all the times you either didn't forget your umbrella or it didn't rain. It is a retroactive and psychological law, a subjective failure in looking back at history. Then suddenly he uses the same phrase, the Law of Selection to apply to phenomena like Darwin's natural selection, which is a forward-looking, future-determining practice. Both are real phenomena, but it seems to me different concepts ought to have different names, otherwise we begin to lose faith in his precision, which strikes me as important in this subject area.

Secondly, and perhaps more importantly, while he gives us plenty of reasons why things like hindsight bias and the law of selection tend to make us think things are stranger than they are, I'd have appreciated a little bit more thought on when statistical explanations do not actually suffice, and how to tell when they explain things and when they don't. To be fair, he does address this issue to a certain extent, I'm only saying I'd like more of it, and there are times his arguments beg for a bit more of it. For example, he argues that while it may seem that the attacks of September 11 should have been predicted, since there were plentiful warning signs, actually that's not fair because there are so many warning signs of all kinds at all times that it is only retrospectively that any kind of pattern could really be discerned. My problem with that -- apart from the questions it raises about the general utility of intelligence services -- is that his explanation fails entirely to deal with the question of the quality of the warning signs received. Possibly he is just trying to keep things simple, and maybe that's fair enough. However, if you're going to pick an example like that, you've got to do it justice. Numerous warning signs about 9/11 were reportedly rather specific about the nature, timing, and location of the attacks. Now, unless there were a similar number of similarly specific warnings about other types of attacks and their targets, then saying the attacks should have been anticipated (and prevented) is not mere 20/20 hindsight. His essential and simple point remains valid, but there was a missed opportunity here to take the analysis to a higher and more useful level.

Anyway, these are minor quibbles and specific to my own biases, and overall this book serves as an excellent introduction to some essential concepts about statistics and probabilities that help make sense of the world. And, incidentally, acknowledging that an occurrence did not qualify as a miracle does not have to take the magic out of life -- if you win the lottery, I think you can still feel a bit lucky even though you know the odds favored somebody winning it eventually.
22 Comments| 2 people found this helpful. Was this review helpful to you?YesNoReport abuse
on 7 December 2015
This book will 'do yer 'ed in' - but in the best possible way. I feel it has been written with both the academic professional and the layman in mind. Being a very definite layman I struggled at times to grasp at first some of the ideas and principles involved. But at the end felt that I had an understanding of why my toast lands butter side down, and other such improbably coincidences. Well worth both the money to buy it and the effort to read it.
0Comment| One person found this helpful. Was this review helpful to you?YesNoReport abuse
on 7 January 2017
The odds of winning the UK lottery are 14 million to 1 ... yet it happens every week.
Excellent book, very accessible. I read it twice, the second time to try to remember the multiple examples used to illustrate the various principles (the laws of very large numbers, the laws of selection etc). I am very familiar with statistics, but I think this is written in a way that allows most people to understand what is being discussed. I especially liked (and could definitely relate to) the dangers associated with trying to explain relationships in data, rather than starting with a theory and testing it against the data. Thoroughly reccommend this book.
0Comment|Was this review helpful to you?YesNoReport abuse
on 15 July 2015
The message in this book is very short.

But the book isn't very short; it's actually quite long.

The constantly underlined and highlighted idea is this:

1 - Borel's Law says (paraphrased) "sufficiently unlikely events are impossible" (in practical, macroscopic or human terms)

2 - Things that LOOK impossible KEEP HAPPENING every day somewhere on Earth

3 - 'Impossible' things keep happening not because Borel's Law is necessarily wrong, but because the things that LOOKED impossible actually were NOT impossible

The events ('things') seemed to be impossible because they were not correctly understood (the probabilities against or for). Once other established mathematical or statistical laws were brought to bear on the 'impossible' event, the impossibility became not just infinitesimally possible, but actually INEVITABLE given enough 'runs' or time -- that is the law of very large numbers. Other laws, when applied correctly, did the same thing.

An obvious, and discussed, event is lottery numbers being the same in successive draws. It seems, in a human time frame, impossible. But consider how many lotteries there are in the world and how many draws and the sheer number of opportunities that this has to occur means that, if not inevitable, it nearly is.

I list some of the quoted and used laws in the next paragraph.

Hand seems to want to coin a new phrase in mathematics or statistics -- "the Improbability Principle" -- whereas what he has actually done is to use that term as an umbrella term for several mathematical laws that are already established. Such as [the laws of] inevitability, truly large numbers, selection (including anthropic principle), the probability lever, and near enough.

As such, he seems to have a weak claim.

Hand uses these established laws to show how a layman (not a mathematician or statistician) typically overestimates the improbability of something happening (or underestimating the probability of something happening) and hence thinks it is impossible. Yet the event happens more than once somewhere on this planet at least.

Like I alluded to earlier, this book is 'impossibly' repetitious. Hand seems to be labouring under the mistaken belief or illusion that TEN examples of one idea are better than one, two or three.

The book IS interesting, but mostly at the introductory and summary pages.

The bulk of this book is too tedious.
0Comment| 5 people found this helpful. Was this review helpful to you?YesNoReport abuse
on 9 December 2016
Really good read with lots of interesting details.
0Comment|Was this review helpful to you?YesNoReport abuse
on 21 February 2017
Easy to read, to understand and very enjoyable.
0Comment|Was this review helpful to you?YesNoReport abuse

Sponsored Links

  (What is this?)