Hurricanes in Torah Codes

I work with CodeFinder software.  Some of the leading researchers in this field work with software that performs Monte Carlo permutations of Torah text.  They derive probabilities by scrambling Torah text thousands of times, then look at how compact the Torah derived matrix is compared with these scrambled (monkey) texts.  If they scramble the Torah 1,000 times, and find 100 matrices as compact or more compact than the Torah, then they rate the initial probability of the Torah Codes matrix at 100/1,000 = .1 (not significant).  But if only 10 out of 1,000 matrices do as well as or better than Torah, then initial p = 10/1,000 = .01 (significant).  HOWEVER, the final probability assigned is not so simple.  The matrix shown on the left below with Andrew, in 5752 (1992 CE) and in (on) the 25^th of Av (August 24), by both Monte Carlo and my own method (Roffman Skip Formula) initially looks like it has a p value of about .0035.  However, this figure is greatly overvalued.  In reality, when this matrix was found by Dr. Haralick, after using my transliteration for Andrew, his computer search was for the terms in the table on the right below.  It allowed for the U.S. (not found on this matrix), four spellings of hurricane (none found on this matrix), four spelling for the year 5752, and three spellings for in the 25^th of Av (the day and month of the storm).  This greatly complicates any realistic probability calculation.  Stay tuned.  Dr. Haralick may soon have a lot to add about his Monte Carlo experiment.  I will update this page accordingly.