How to lie with probability: "there is a 67% chance that God exists. " :D

Discussion in 'Psychology' started by harrytrader, Mar 20, 2004.


    Odds on that God exists, says scientist

    Stewart Maclean, Catherine Bolsover and Polly Curtis
    Monday March 8, 2004

    A scientist has calculated that there is a 67% chance that God exists.

    Dr Stephen Unwin has used a 200-year-old formula to calculate the probability of the existence of an omnipotent being. Bayes' Theory is usually used to work out the likelihood of events, such as nuclear power failure, by balancing the various factors that could affect a situation.

    The Manchester University graduate, who now works as a risk assessor in Ohio, said the theory starts from the assumption that God has a 50/50 chance of existing, and then factors in the evidence both for and against the notion of a higher being.

    Factors that were considered included recognition of goodness, which Dr Unwin said makes the existence of God more likely, countered by things like the existence of natural evil - including earthquakes and cancer.

    The unusual workings - which even take into account the existence of miracles - are set out in his new book, which includes a spreadsheet of the data used so that anyone can make the calculation themselves should they doubt its validity. The book, The Probability of God: A simple calculation that proves the ultimate truth, will be published later this month.

    Dr Unwin said he was interested in bridging the gap between science and religion. He argues that rather than being a theological issue, the question of God's existence is simply a matter of statistics.

    "On arriving in America I was exposed to certain religious outlooks that were somewhat of an assault upon my sensibilities - outlooks in which religion actually competes with science as an explanation of the world," he said.

    "While I could not be sure, having slept through most of the cathedral services I had attended during secondary school, this did not seem like the version of faith I had remembered. In many ways, this project was for me a journey home - a reconciliation of my faith and education."

    Despite his findings, Dr Unwin maintains that he is personally around 95% certain that God exists.

    However, Graham Sharp, media relations director at William Hill, said there were technical problems with giving odds on the existence of God. "The problem is how you confirm the existence of God. With the Loch Ness monster we require confirmation from the Natural History Museum to pay out, but who are we going to ask about God? The church would definitely confirm his existence."

    Mr Sharp said William Hill does take bets on the second coming, which currently stand at 1,000/1. For this confirmation is needed from the Archbishop of Canterbury.

    "We do take bets on the second coming, whether that confirms the existence of God is up to the theologians to argue, most people wouldn't believe that, though."
  2. This guy is not a scientist for sure. If it suffices to use Bayesian theory to be a scientist a monkey can also be a scientist: it suffices to throw a dice and create artificial probability numbers.
  3. This is obviously ridiculous but not as obvious when applied to some other "serious" field :)

  4. Blaise Pascal, Mathematician, Physicist, one of the greatest French writers, is generally recognized as the "inventor" of Probability Theory. (17-18th cetury). He also constructed the first (or second) sophisticated mechanical calculating machine.

    Pascal went into probabilistic arguments about the existence of God and the wisdom of belief expressed in terms of probabilities.
  5. 200-year-old-formula what an argument !
    As for the reference to "Bayes' Theory is usually used to work out the likelihood of events, such as nuclear power failure", yeah I have read a few months ago such a nuclear report by an expert who was comparing estimation based on Bayes Theory (in fact there is no real alternatives than using it in this domain) done by two different parties: government or citizens association the difference can differ by several 0 as all depends of course of the SUBJECTIVE PROBABILITY affected to such or such event.
    One again risk and UNCERTAINTY is two different things, there are some uncertainty that is just unquantifiable although in practice it will be done for example they have just launched in France "Catastrophe bonds" to insure a big enterprise against the risk of great winds France has undergone. In petroleum engineering school we have conference on the subject as it is also a domain of high uncertainty: there ain't just no solution, only nose estimation and high risk taking by assurances companies which will probably count on government subvention ie tax people if ever they can't fullfill their promise. At least in France it works well as Gov or the mentality is socialist but even under non-socialist governement it seems to work also as it implies more debt. BTW after such a catastrophe insurances in France didn't reimburse the victims straight away under the pretext that the money was blocked ... in stock market !!!

    A funny anecdote: for petrol exploration the cost is high and the chance to find something is extremely low. Some US insurance companies refused to deal the petroleum company if they don't take a medium in complement to sismographic sophisticated tools normally used to detect oil :D.

  6. Basically Bayes theory is based on a tree decision since I talked about petroleum field the frequent question is "to drill or not to drill" :D software tool here : sorry not free £695 :)
    <IMG SRC=>
  7. In view of the fact that you published many freightloads of stuff that you don't appear to comprehend anything at all of, let me point out that the above statement is completely nutty referring to Blaise Pascal.

    Do you realize that the mathematics of conical sections as it is still taught to our youth today is practically a faithful copy of how Pascal presented this about 300- years ago?

    What about some of the formulae he derived on binomial series which many mathematicians and engineers use almost daily?

    What about his basic formulae in hydraulics?

    Let me say, your comments are not very bright coming from a compatriot of good old Blaise.

    By the way, if you have any feeling of honor, tell us finally something about your crazy properties you claimed for the number Pi, this in the context of mystical/mathematical hogwash. You still have not published any correction to this. Pascal would not have let stand a mathematical statement in error for that long.

    P.S. Your title "How to lie with probability" is not very bright either. A mathematician would have first established the truth or falsehood of the proposition under discussion before making claims as to somebody attempting to "lie" or deceive using probability theory.

  8. Of course the expert system I was talking about for my model (see is also based on a decision tree and my purpose is to attach inferential probability at each node. But they won't be subjective as in the case of many decision trees since it is based on a mathematical model that can create objective class partitions and not subjective ones which would cause problem and imply fuzzy logic which I want to avoid. I should let people play with that next week or so (I have implemented candlestick also so as people can see the application with something they are more familiar with. I should add camarilla soon). Using a commercial package as the one below is not practical as for the above application: it's too limited in capacity and flexibility.