Table of Contents
FINAL RESPONSE TO TIM HAYES
On 21st March I send you an email dealing with Tim Hayes' response. I can imagine that you might have expected a few more words from me there and, indeed, I should not have left you with so little as was written in this email. Finally, my wife persuaded me to look into Tim Hayes' web page this morning and to read his response to my recently published test results. So, here is a more comprehensive answer to his statements.
Again, I have the strong feeling that Tim Hayes is missing the basic point of theomatics or that he has deliberately chosen to deny the basic theomatic rule. Anything related to God is at first of matter of the heart and not of the brain. Brilliant mathematical performance is certainly not the number one topic in our relation to God. Since the dispute with Tim Hayes is a mathematical one my following response will also concentrate on mathematics.
To me Tim Hayes has a completely different view of theomatics. He cannot accept its presence if it cannot be absolutely proven with at least a 100% certainty. In consequence he has to apply a completely different method of testing.
The basis of your statistical statements and of my addings in the statistical tests follows the rules of daily statistic practice used commonly in e.g. industry for quality assurance purposes. This method is highly effective and economical by accepting a certain level of error probability. To my understanding the Tim Hayes MOS approach is quite different. Tim's approach is looking for absolute certainty. That is where the difference in the amount of testing between Tim Hayes and theomatics is resulting from.
Let's take Tim's example of the "tallest man" in his "An Illustration" chapter. Tim identifies the tallest man out of a sample but he is not sure, whether this man is really the tallest man in absolute terms. So he has to test another sample and a further sample and another sample ... until he is absolute sure that he has identified his man to be the absolutely the tallest (law of the large numbers, probability to tend to 100%, huge amount of testing).
The theomatics approach is the more practical one in terms of amount of testing and is BY NO MEANS LESS VALID. Theomatics identifies the tallest man out of a sample plus the tallness distribution in the sample. Based on the statistical analysis of the tallness distribution in the sample a statement can be made on the absolute tallness of our tallest man from the sample but not with 100% certainty but with only e.g. 95% probability. The difference of 5% is the error probability that our tallest man from the sample might not be the absolute tallest man due to the randomness of the taken sample. The result of the theomatics testing could alternatively have been, that with a probabilty of 95% our tallest man of the sample is not the absolute tallest man, but also with an error probabilty of 5% that he could nevertheless be the tallest.
The larger our sample size is the more accurate our statement on the absolute tallness of the identified man will be. If our sample size would be as large as the one of Tim Hayes, the result would become identical, if all the other premises were identical too, which they seem not to be. In fact, the premises are the main area of disagreement. There is no dispute on the principles of the MOS method, but to my understanding it is not applicable to theomatics in the way proposed by Tim Hayes for at least the following two reasons:
Following the theomatics approach the company would produce let' s say 10,100 parts and pick a sample of 100 parts randomly out of the 10,100 for testing. If the test results of the sample are such that more than 1,000 operating hours for the parts could be expected the company would deliver the 10,000 parts with a certain error probability that some of the delivered parts might fail before they achieve 1,000 operating hours due to the randomness of the sample picking.
i.e. the theomatics method is as valid as the Tim Hayes method but much more practical and easier to handle for theomatic testing.
It seems to me that Tim Hayes is not aware of this different valid approach which in fact is based on the same statistical principles. Furthermore, I feel that the quest for the "relative truth" is more appropriate in bible analysis than looking for "absolute truth". What would happen to our faith if we had absolute truth? If there would be 100% knowledge there is left only 0% for faith. I cannot imagine that this is the intent of God's word. A further strong point is the difficulty to test the complete bible because of the huge amount of data. Thus it is very reasonable to test samples and apply the methods of conclusive statistics.
Well, I hope I could provide some clarification on the response of Tim Hayes. At least for the story of the prodigal son theomatics has demonstrated non-randomness with a commonly used error probability of 95%. From the statistical point of view, there is absolutely no need to perform a large amount of testing in order to demonstrate a 100% certainty. The statistical world can very much live with 95% certainty.
Please, feel free to use this email as you like.
May God always bless you,