Statistical God

The Probability That God Exists.

I have been reading Richard Dawkins’ book, The God Delusion, about which I will post in due course. This article is about a specific item in that book, where Dawkins quotes Stephen Unwin’s book, The Probability Of God. In fairness I will say that I have not read Unwin’s book (though I did subsequently check the quote), I based this post on what is quoted in Dawkins’ book. It is not a review of the book, only my thoughts on the particular risk management technique in relation to the subject.

Unwin is by profession a risk management consultant who used a branch of statistics, called Bayesian statistics, to try to demonstrate the existance of God. Bayes theorem is used in business as a risk management tool, as a way of combining estimated liklihoods to arrive at a final probability; note “estimated”. For once, I am on slightly familiar ground, though I claim no expertise in Bayesian statistics. I have used risk management techniques in my former employment, and have a rudimentary understanding of Bayesian theory. Even as a Christian believer in God, I have trouble agreeing Unwins’s verdict, as I don’t believe He can be subjected to a human derived method of analysis by estimation.

Unwin’s basis is to start with complete uncertainty, which he defines as 50/50, either way. 50% for the existence of God and 50% for non-existence. Then he states six facts that he believes bear on the matter, and this is where I disagree; his so called facts. He says:

1. People have a sense of goodness.
2. People do evil things (Hitler, Stalin, Hussein).
3. Nature does evil things (earthquakes, tsunamis, hurricanes).
4. There might be minor miracles.
5. There might be major miracles.
6. People have religious experiences.

Unfortunately, to me none of Unwin’s proposals stand on their own as an objective fact.

1. Some (perhaps most) people have a sense of goodness, is a fact. The original statement is not.
2. Some people do evil things is a fact, but again his original statement is not.
3. Nature does not do evil things. It sometimes does what humans think are bad, but nature is not evil. Nature has no malice or intent to harm. It is a sometimes localised way of trying to keep a wider balance.
4/5. Why separate these? There are or are not miracles, the perceived degree is irrelevant in this context.
6. Some people have spiritual experiences. Not everyone who does defines it as religious.

Unwin’s estimated probability of the existence of God, if you follow through his calculations, is 67% which he seems to think is not high enough. To push the probability up, in Dawkins’ words, “he takes the bizarre step of boosting it to 95% by an emergency injection of ‘faith'”.

Much as I hate to say it, and I still believe in God, I’ve found something I have to agree with Dawkins opinion on.

2 thoughts on “Statistical God”

1. Jan Weir

Jim:
I don’t know you and I hope you’ll forgive my bluntness. To have an opinion on a book you’ve never read is bad enough – but to post that opinion is beyond the pale. I do understand that that’s the culture we inhabit, but I reserve the right to be disappointed.

1. JandWs Post author

I did not post an opinion on the book, only on the specific risk management technique, which I did take the trouble to verify. Although I have not read the whole book, I am experienced in risk management practices.
I have nevertheless edited the post slightly to make this clear. I am sorry you felt disappointed.

This site uses Akismet to reduce spam. Learn how your comment data is processed.