Based on my own thinking as well as the Chair's emphasis on games in his writings, I have decided to study Game Theory. I have looked into the basics where available online. As with a lot of math, it seems that in an effort to spell out and/or prove their own theories, most authors end up using more mathematical language, more decision trees and etc. than is really needed. I find that after I study a subject, the details aren't really that important to me anymore. What is important, and what sticks, is an overall philosophy learned while studying the subject.

I went through the same thing while studying Bayesian Statistics. After reading hundreds of pages and going over and over mind-numbingly complex mathematical formulae, I still feel Bayesian statistics can be described, not only adequately, but completely, in a paragraph or so. Maybe something like this:

Future probabilities can be directly predicted from past occurrences. What happened the majority of the time in the past should continue to happen. If it stops occurring, to the point that in all recorded occurrences, it turns the corner from happening the majority of the time to happening the minority of the time, then it will continue to be the minority into the future.

Now maybe some of the more learned here can tell me what great parts of Bayes I have left out. But it seems silly to me to give someone a 500 page book to study from when the above paragraph and a bit of common sense serve to be just as well, unless of course, that someone is planning to become a professor of statistics on a college level.

Anyway, I came across this book while researching game theory: Game Theory: Analysis of Conflict by Roger B. Myerson

It garnered rave reviews on Amazon.com:

To find the best way to present various materials, I went through virtually every game theory book in existence. For the presentation of the basic material on normal and extensive form games, nothing even came close to this book in clarity of presentation and depth of understanding of the issues. Most textbooks, even highly touted ones that are mathematically challenging, do not even come close, and rarely even present the material in a coherent form at all.

This sounds promising, has anyone read this? But still, I wonder if there is a book out there that covers, instead of how to dissect all possible games and create the most intricate strategies, something more like "lessons learned" from game theory, and something that covers the "philosophy" more than the math per se.

Does anyone know of anything like this?





Speak your mind

2 Comments so far

  1. Alan Millhone on January 14, 2007 8:58 pm

    Dear Chair:

    Which games are Ryan referring to in his aricle ? Many feel Checkers is math oriented because of the numbered board, and feel if one ‘knows the numbers’ they can’t be beat in actual play ! Is he wanting a ‘tie’ between board games and the market and use the theory involved in some board games to show a correlation to the theory of investing in the Market?

    Alan Millhone

  2. Lineu Vargas on January 15, 2007 11:52 pm

    In my experience (I did quite a lot of postgraduate level work on Game Theory), the mathematical aspects of Game Theory are inextricably linked with the intuitions derived from the results, so one cannot really have a serious “lessons learned” book. Of course one can write a lot of flowery prose using game theoretic concepts, but this misses completely the point of what Game Theory is about. More fundamentally, one cannot have a “lessons learned” book because there are no lessons learned: Game Theory is not a science, it is an attempt at formally constructing a language to talk about strategic interaction between rational agents when their interests (or payoffs) are interdependent. Also, more importantly, it is not about games in the sense we usually understand the word.
    As books go, Roger Myerson’s is very good, but you might want to check out too “A Course in Game Theory” by Martin Osborne and Ariel Rubinstein.


Resources & Links