Introduction to Decision Analysis

Given a few reasonable assumptions, it is possible to prove Bayesian Decision Theory is the best way to make decisions. However, the description of Bayesian Decision Theory is dry, mathematical, and hard to follow for most people.

One approach would be to teach people Bayesian Decision Theory from basic principles, including how to apply it. But this would be even drier, even more mathematical, and, potentially, many people would give up.

Imagine the kind of people who would pay good money to learn how to make better decisions. C-level executive or other people in senior management positions. These are people who regularly make decisions that have millions, sometimes even billions, of dollars at stake.

They often hire consultants for help, but if you do not understand the methodology at all, it would be hard for the consultant to explain how they reached their conclusions. These consultants need an easy resource for them.

However, this resource cannot look like “a book for C-level executives to understand what the decision process consultant is saying”. This would look bad.

Instead, what you would need is a book that looks like it is teaching you how to make better decisions. The consultant would recommend this book as “explaining what I do”, not as “explaining how to understand me”.

Crucially, at no point should the book break character. Not in the book itself, not in the back-cover summary, and not in the marketing materials.

In fact, better if the book just looks like it is in general about decisions, not just in business. Include a few examples from people’s personal lives, to make it more relatable.

Unfortunately, when you never break character, you fool some innocents. I read this book assuming it would teach me how to improve my decision making. It is not designed for that. It tries to make the math “accessible”. It aggressively avoids suggesting learning how to program, instead recommending how to abuse Excel or, sometimes, how to buy overpriced, hard to use, software – that does not run on mobile or on the web, just install it directly on your Windows machine.

There are Excel abuses. There are ways to avoid Jupyter and Graphviz. There is a lot of dumbing down of the math.

I have stopped reading books in the middle. But this one had the fascination of a train-wreck. Even as I grew frustrated with the book, I grew impressed with the author’s commitment to never break character.

But I am glad that I have finished it.