I am reading (on a free Kindle reader app) a great book on the power (and frustration) of Bayes Theorem. The book is: "The Theory that would not die" by Sharon Bertsch Mcgrayne
Bayes is the guy--from the 18th century--who told us that given some data (actual observations) we can reverse engineer the underlying probabilities, or at least the parameters like mean and deviation. One catch is that we are required to guess a starting point. Oops! Guessing is not what we do in project management, or mathematics for that matter.
This idea (guessing to start, and then improving the guess with real data) is an anathema to the 'frequentists' who go at it the other way 'round: given parameters, we can predict data. Oops! if the event has never happened, or happens infrequently, or has never been observed, where do the parameters come from? How can we use situations like these to drive decision making? If we can't make a decision with it, then does anybody care? Spending time observing such stuff is not what we get paid to do in project management.
I was searching for some point-counter point stuff to go along with the book. I came upon this presentation from a guy (Louis Lyons) at Oxford. You gotta love the last page! (as below)
