Wednesday, May 16, 2012

Book Recomend - Thinking Fast and Slow

 The first GPS I ever encountered had a quirk. If it stopped receiving a signal it would, without any indication or alarm of any kind, switch over to navigation by dead reckoning. It would always give you a position that looked accurate out to three places - even if it was not receiving a signal at all.

 (The cruise ship Royal Majesty went aground for this reason)

 Turns out, the human mind works much the same way. Here is an example from the book
 Likely you read the left box as A B C and the right box as 12 13 14 but A 13 C or 12 B 14 would have been equally valid. Your mind resolved the ambiguity without effort but like the GPS switching to DR it did so with out notification.

People have a strong bias for plausible  narratives and prefer them to more probably  scenario. From chapter 9:

A remarkable aspect of your mental life is that you are rarely stumped......The normal state of your mind is that you have intuitive feelings and opinions about almost everything that comes your way.

The  mind uses a trick the author calls Substituting Questions. Rather then answer a hard question the mind substitutes a simple problems and answers that one instead. Ironically the less information we have about a question the more certain we are that we have found the correct answer.

There is far more in this book then the above. The best review I've seen is this one from the New York Time Book Review. 

Wikipedia has a good summary here. 

 Thinking Fast and Slow has five parts:

Part 1 Two Systems
Part 2 Heuristics and Biases
Part 3 Overconfidence
Part 4 Choices
Part 5 Two Selves

 The difficulty I had with this book is it seems too long. I found parts 1 through 3 to be fascinating but started bogging down a little on parts 4 and 5. When I hit the section about Bernoulli's Errors and  the fourfold pattern  I felt the author was starting to torment me.

 At one point when I came to problems in the book like "Bet A 11/36 to win $160, 25/36 to lose $15 or Bet B 35/36 to win $40, 1/36 to lose $10"  I had to skim, I understand that I don't understand.

 This problem is a little more up my alley - I got it wrong but it illustrates the point.

A bat and ball cost $1.10.
The bat costs one dollar more than the ball.
How much does the ball cost?

Most people choose 10 cents - the correct answer  is 5 cents: an explanation is here.
 A  point of interest to mariners was the section on feedback and practice in the chapter on Expert Intuition. The author compares driving a car which we can master quickly compared to a harbor pilot maneuvering a large ship. Ship handling takes  much longer to learn because of the long delay between  the action and the outcome. The quality and speed of the feedback is much better in the case of the car.

I highly recommend this book if you make high-stakes decisions, every master does. I've read other books on cognitive bias and so forth but this book goes both deeper and covers more ground then any other book I've read. It is the "big picture"  The going gets a bit heavy in places but overall well worth reading.



Michelle said...

OK that's really interesting. Count me among the "more than 50 percent of students at MIT, Harvard and Princeton" and the "more than 80 percent" from the "less selective universities" as I got it wrong and still don't fully understand the logic. Perhaps it's a good thing that I'm not making the sorts of decisions that ship captains are on a daily basis!

K.C. said...

There are two kinds of problems, ones you can quickly solve in you head and ones you need a pencil a paper and work out. This problem is interesting because it appears to be the first type but in fact is the second.

Reid Sprague said...

Interesting! Thanks for the review, I'll be giving it a read.

I'm old enough now not to lose too much confidence in myself once I find out (as it sounds like I'm liable to if I read this) how much freewheeling and guessing I really do!

Although there is something about maritime decisions that drives you to double- and triple-check those intuitions. . .

Thanks again!