The Probability Of… An Interesting Read?

Whenever we wish for something, we should be careful. Our wishes just might be granted. Several times in this series I have stated that one of my goals is to encourage readers to break down the boxes between disciplines and realize that the lessons of statistics, probability, and decision theory apply to everyday life.

The other day I was at our local library (yes, I’m a dead tree junkie) looking for Dawkins’ book to check out a comment someone made about the column I wrote in which Dawkins was mentioned. A shelf away from where that book should have been, my eye fell on a book that I knew was just calling out to be read. The book was written Stephen D. Unwin, who earned his PhD in theoretical physics and is president of his own company with the specialty of assessing risks. His company could be called on to study the plans for a nuclear reactor that is going to be sited 300 feet from your house to assess the probability of an accident. Given that is how he makes his living, the title of his book is a real eye-catcher: The Probability of God!

Yes, Stephen has undertaken no less a Herculean task than using Bayesian probability computation to estimate the probability that God exists! By “God,” he does not mean some wimpy pantheistic spinoff of Spinoza or the Buddha. He means the muscular old man with a long beard who can either zap you real good or give you unlimited treasure if he so wishes. (Note: one of the disquieting things about this book is that Unwin always refers to his deity as masculine, but he stubbornly uses the feminine gender for the indefinite. Our language is in transition, but the juxtaposition of his affectations is jarring.)

The technical part of the book is well-written. The discussions that he has exploring the pitfalls of definitions (Like what does “exist” mean? He devotes a whole chapter to that discussion.) are thought-provoking. All in all, I think that any reader of this column would enjoy reading Unwin’s book. However, the whole thing is colored by the fact that he computes the probability of a person-God (his terminology) existing at 0.67.

What does a probability of 0.67 that God exists mean? In this case, it probably means that Unwin would not have written the book if his probability was either less than 0.5 or suspiciously above 0.99. Too low of a probability would not have supported his declared belief in the Christian faith, and too high of a probability would have turned off potential candidates for conversion. Please understand, I am not accusing him of fudging the numbers. On the contrary, I sadly think he believes what he wrote.

After introducing the Bayesian formulation, he selects various parameters that can be used as evidence in the computation of posterior probabilities. He narrow the universe of data down to six parameters: (1) the recognition of goodness; (2) the existence of moral evil; (3) the existence of natural evil; (4) intra-natural miracles; (5) extra-natural miracles; and (6) religious experiences. Each of these is discussed in detail and then evaluated as to its evidential value in computing a posterior probability that God exists. To give a taste of the spirit of how evidential value is assigned, this quote is from his discussion on moral evil: “In a godless universe, moral evil is inevitable.” Applying this observation to his formulation of probability has the surprising result of slightly lowering the probability of God existing. It’s that sort of mental jujitsu that becomes a saving grace for the book.

The bottom line is GIGO (garbage in, garbage out), you can dress up your beliefs in the nicest mathematical clothes available, but that is not science. To his credit, Unwin admits this and repeatedly indicates that his computations are essentially a way of organizing and presenting his own beliefs in a way that he hopes will enable reader to see why he has faith (and how much). He even has a chapter devoted to showing readers how to program Excel so they can compute their very own probability of God.

My only real complaint about this otherwise enjoyable and informative book is the sub-title on the cover, “A Simple Calculation that Proves the Ultimate Truth.” Come on , Stephen, what are your definitions of “proves” and “ultimate”? His arguments come nowhere near being proofs and ultimate truth remains an elusive grail. I would be extremely nervous if he allowed a “safe” nuclear plant near my house.

In response to the interest my original tutorial generated, I have completely rewritten and expanded it. Check out the tutorial availability through Lockergnome. The new version is over 100 pages long with chapters that alternate between discussion of the theoretical aspects and puzzles just for the fun of it. Puzzle lovers will be glad to know that I included an answers section that includes discussions as to why the answer is correct and how it was obtained. Most of the material has appeared in these columns, but some is new. Most of the discussions are expanded compared to what they were in the original column format.

[tags]The Probability of God, Stephen D. Unwin, decision theory, 0.67[/tags]