Bayes and the Bible
http://home.messiah.edu/~chase/articles/FilteringTheBibleAndSpam.pdf
After the realization of the power of Bayes’ Theorem, many people started using the theorem to predict the actual truth of data based on past results. Using this method, one could use the verify the results of medicine and spam screenings using the success in the past. Emails have used Bayes’ Theorem to see how often the inclusion of a particular keyword (such as Nigerian) would be a spam email versus how many times the email is actually a legitimate email. After applying the rule to an email, the success or failure can be used to update the spam filter’s rate of success.
In a very similar fashion, John Craig used this principle of utilizing the evidence of the past and current statistics to predict the eventual end of belief of the Bible. This prediction was based on the existence of scriptures of Biblical events and the less believable these events become as they are translated and pass in time. Using a couple of constants for rates of less belief and initial conditions, he constructed a model to estimate the current believe-ability and when the belief would fall below zero. Using his factors, he estimated that the current belief in the Bible is around 70% and the time when nobody will believe the Bible (or as he puts, the End of Days) is estimated at 3150AD.
In class, we studied Bayesian probabilities in order to estimate the truth of events based on an observation in current time and basing it on the statistics from the past. Although Craig does not directly use Bayes’ Theorem (partly because it had not been thought of yet), he uses a similar method of basing his prediction on the observations of the past and the current statistics. Also, instead of using a straight probability, instead what is more important to Craig is the predicted decline in belief of the Bible (decrease in the Bayesian probability) to a value of zero. Therefore, Craig uses the difference of the log values to estimate the belief in the Bible to being decaying exponentially. Craig could be seen as predating Bayesian statistics in order to predict truth in spam, but similarly using prior information in order to test the validity of his own statement about the eventual disbelief in the Bible.
-Jset