Customer Reviews

Guesstimation: Solving the World's Problems on the Back of a Cocktail Napkin

5 star:		(5)
4 star:		(4)
3 star:		(4)
2 star:		(0)
1 star:		(1)

 

 

This book was entertaining reading...except that most of the more interesting examples described in this book were so familiar. This is one of those books where I might rate it better if it weren't for the fact that there is a far better book for anyone interested in this topic. Like Guesstimation, How to Measure Anything: Finding the Value of "Intangibles" in Business by Douglas Hubbard also discusses the Fermi approach and how ancient Greeks estimated the circumfrance of the Earth (Hubbard's book uses these same examples even though it came out almost a year before Guesstimation...curious). But Hubbard picks up where Guestimation, as another reviewer puts it, "falls short". After a bit of "Fermi decomposition", Hubbard discusses how we can learn to excel at subjectively assessing odds and ranges and how we can compute the value of further measurement. Then he gets into a fascinating array of practical methods of observation to further reduce uncertainty. Although the techniques in Hubbard's book are based on sophisticated mathematical methods, he is able to reach a much broader audience by distilling the math into simple charts, tables and procedures. In short, if you owned both of these books, Guesstimation would be redundant and wouldn't cover nearly as much.

Somehow, guessing at numbers is unsettling, even though I've done it all my life. John Adam is a professor of applied mathematics, with a degree in physics. Larry Weinstein is a nuclear physicist. Their book is devoted to proving that intelligent guessing is useful and fun.

The book lays out some general principles but its great strength lies in the interesting problems, a series of hints to help you solve each problem, and an interesting discussion of the pitfalls and triumphs involved. Three key points: estimate by powers of ten, break complex problems into simple steps and consider alternative approaches.

The book includes an excellent appendix containing a few formulas and scientific concepts, together with some useful statistics. The pen-and-ink sketches are funny and to the point.

Best of all, the Princeton Press maintains a site with new problems on a weekly basis; a recent question was how many golf balls would it take to encircle the earth at the equator. Hints included:

a. What is the diameter of a golf ball?

b. What is the circumference of the earth?

The authors give several interesting hints at determining the circumference of the earth (if you don't know it), including the 24 time zones, the number of time zones in the US, the time it takes to fly from New York City to Los Angeles, etc.

Brain stretching stuff, which is always good for you, and the publishers claim job applicants should be prepared for tests of their estimating abilities. The Chinese rights have just been sold, and we may face even more competition on that front as well. If you learn by doing, this book is a great way to improve your skills and have fun doing it.

Robert C. Ross 2008

I was somehow expecting to like this book more. And even having said that, I am not sure whether three stars is a little mean, since I did learn real stuff from it, and anything from which you learn real stuff is better than average.

Why do I not like it better? First, the style the authors have decided to adopt, of flippant, cheerful informality ('hey, yay, who would have believed that', 'that is a _really_ big number', 'Gravity sucks!'), that seems to be aimed at frat-boy undergraduate physics nerds - a set of very small cardinality, I would have thought. The effect, at least for me, is the prose equivalent of someone dragging nails down a blackboard. Second, the examples are a bit samey - you do not need to read a lot of them to get the general idea. Third, the examples are very disjointed: question, hints, workout; question, hints, workout, question...

This is all very critical, and I should emphasise that I _did_ learn some useful tips about how to think about certain problems, esp. in the more 'physics-y' questions that come up later; i.e. the more authentically Fermi questions. If you don't mind the prose, you might easily give this a four. Otherwise, you might be better advised to track down Sanjoy Mahajan's work in progress (the name of which escapes me at the moment), which looks like it will be a more serious, and structured, attempt to explore the same sorts of methods.

If you can do basic arithmetic with one-digit numbers - add, subtract, multiply, divide - you can be a math genius. Maybe not genius, but you can still blow the doors off most people, and build up a healthy amount of BS-proofing, just by learning how to apply the skills you already have. This book offers dozens of worked examples, using pretty much just the math you learned by sixth grade.

Weinstein and Adam chose a format that's easy to pick up and thumb through. They present each poser on one page, with hints to help you get started. A few extra facts "to hang things on" appear at the back of the book: the sun is about 10^11 meters away, a billion seconds is about 30 years, things like that. Then, the next page or two after the problem works out its answer, often more than one way.

For example: could we create a human chain from Earth to the sun? Well, the sun is about 10^11 meters away, and a person is about 10^0 meters from fingertip to fingertip with arms stretched out. (For back-of-the-envelope purposes, you can often skip the leading digits of numbers.) So, the distance from earth to sun is about 10^11 people-widths, but the Earth's human population is just under 10^10. Answer: We'd certainly come up short.

Some questions, like that one, are silly factoids. Others have more pressing social importance. How much funding does a subsidized school lunch program need per year? How many acres of farmland would it take to fuel your car with ethanol? How much landfill area does your town need for the next decade? When political special interests start throwing numbers around to answer these questions, are they lying to you? Even if you don't have exact numbers to work with, the way you get the answer is what matters, and you know exactly what assumptions you've made. Then, when you get more facts, you can refine your answer.

You don't have to be a nerd to command a lot of nerd power. Grade-school arithmetic (which the authors review), a few basic facts, and a bucket of common sense go a long way. This book, with its puzzle-solving format, can help you develop that skill.

-- wiredweird

I love the samples in this book and they really helped me hone my estimations

Thousands of prices, hundreds of distances, millions of digits, without knowing it we encounter numbers in our daily lives. We use these numbers subconsciously to make simple decisions on purchases, traveling, or even sports. When numbers get larger and the problems become more complex most people become overwhelmed. Guesstimation by Lawrence Weinstein and John Adams attempts to bring order to this chaos. With some basic facts, the authors show the readers how almost every problem can be estimated with some level of accuracy.
Topics are introduced on a very basic level and then slowly increase in complexity. My favorite part of the book is the basic introduction of facts. Small facts can help you infer significantly more difficult facts. For example, most people don't know the circumference of the earth at the Equator. However, broken down into smaller steps you can estimate the distance quite easily. Using the distance of New York to Los Angeles of 3000 miles (which is 3 time zones), you can extrapolate this number to the entire earth (24 time zones) by multiplying by 8. Since 3 times 8 equals 24, we can estimate the circumference of the Earth to be 24,000 miles. The actual circumference of the Earth is 24,859.82 Miles which is quite close to our estimation. Cool little tricks like this make this book absolutely fascinating.
Estimating the total length of all the hair on an average woman's head or how long a running faucet would take to fill the capital building may seem impossible at first glance. However, many high end employers (including consulting firms, software innovators, and investment banks) use similar questions for interviews. This book will give you the tools to solve these problems and train your thought process. For every avid learner or any eager job applicant, this book is a must have.

There is so much in this book...I walk away from this book wanting more. Not because they left anything out, quite the contrary. I have a desire to do the 'homework' that is in the last chapter!!

After the first one or two questions you quickly figure out how to do the 'guesstimation' and the rest is just math.

The authors made this book simple to read and enjoyable...you didn't always see the punch line coming.

This is a great "hands on" book that teaches the art of making a quantitative "educated" guess based on just a few basic facts most people know (or should know). I found this book great reading and very educational. Recommended for anyone.

The idea of this book is great, and I had hoped to learn new principles for estimation. Unfortunately, the execution falls short. I expected the authors to teach methods and techniques, but instead they rely almost exclusively on examples that try to teach by illustration. I don't find that style effective and I had hoped for a deeper presentation.

a great addition to any intro physics course - training students to solve these type of problems provides them with a really useful quantitative skill that can be applied to many types of real world problems.