Draft of Haralick's Paper
Responds to WRR Skeptics

A landmark study by Professor Robert Haralick of two controversial code experiments that appeared in Statistical Science has been released to code researchers in draft form and is currently under review for publication in an academic journal.

Five years in the making, Testing the Torah Code Hypothesis: The Torah Code Effect is Real analyzes the famous Witztum, Rips and Rosenberg (WRR) paper, Equidistant Letter Sequences in the Book of Genesis, which was published by Statistical Science in 1994, and the response to the WRR paper, Solving the Bible Code Puzzle, by McKay, Bar-Natan, Bar-Hillel and Kalai (McKay et. al.) that appeared in Statistical Science in 1999.

Distinguished Professor of Computer Science at the City University of New York's Graduate Center, Haralick has co-authored or edited nine books and has well over five hundred papers listed in his resume.

WRR's experiment was given worldwide attention with the success of the Michael Drosnin book, The Bible Code (Simon & Schuster, 1997). It involved looking for the names, birth dates or dates of death of an a priori set of 32 famous rabbis. ELSs of their names and dates formed compact arrangements in the Hebrew text of Genesis. By their Monte Carlo experiment, the odds against this happening by chance were less than 16 out of 1,000,000.

But skeptics McKay et. al. countered with a paper accusing WRR of rigging the names to improve results. They supported this hypothesis with an experiment using a Hebrew text of Tolstoy's War and Peace. With a little tweaking here and some tinkering there, their Monte Carlo procedure "proved" that any book could produce the same results if you are willing to bend the rules of good research. And Statistical Science, under pressure from scientists who did not cotton to the idea of one of their most important journals confirming special attributes of the Bible, actually printed their paper.

Now Haralick brings his impressive knowledge to bear on the McKay et. al. research with an experiment "which proves the McKay et. al. argument to be fallacious." Furthermore, he writes that not only did WRR not fudge the names and dates they used, but that McKay et. al. were guilty of the very thing they accused WRR of doing—they "cooked" their experiment with names and dates that they selected in private to improve the results of their War and Peace research.

One of McKay’s inferences was that a more complete list of names would show the WRR results to be faulty. Setting up his experiment to use both WRR's and McKay et. al.'s lists of names, Haralick uses better statistical protocols to test them, including the ELS Random Placement Model, and four new compactness measures, and finds that the combined list does better in the Genesis text than the original WRR list and that the combined list does worse than the original McKay list in War and Peace. The study's conclusion is that WRR's Torah code effect in Genesis is real and that there is no similar effect in the War and Peace text.

But Haralick's experiment showed that McKay et. al. had part of their appellation list right: some of the names and dates on their list were valid.



Copyright © 2017 BibleCodeDigest.com