Mourning The Yeshiva Slaughter

On the killing of our precious holy and courageous students, refined souls, loving of our holy land, who were massacred in their zealousness for studying with Torah books in their hands. The Martyrs: there is no one in creation who can stand in their exalted place.

The Slaughtered Students
The Slaughtered Students
The Funeral
The Funeral (AP Photo)
We Mourn
Mourning Throughout the Holy Land

We now explore whether there is any relationship between each student's name with the date of death. For the names of the students we use their last name or last name prefixed by the initial of their first name. If they have a middle name then we include in our search the initial of their first name and the initial of their middle name as prefixed to their last name. For the date we use three forms: Adar I 30, 30th of Adar I, on the 30th of Adar I. For each student we show the best table and its associated p-value, and the p-value of the combined experiment using the 6 or 9 combination pairs. Not all the p-values are small, but there is certainly some hint that there are more close meetings than would be expected by chance.

Doron Meherete

For the key words Meherete and on the 30th of Adar I, the probability that a text from the ELS random placement text population would produce a table as compact a table as that produced by the Torah text is 38/1,000.

student 1
With expected number of ELSs set to 10, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 38/1,000. The cylinder size produced by the search was 100 columns.
Finding by Professor Haralick

For the combined experiment using the appellations D. Meherete or Meherete with the three date forms Adar I 30, 30th of Adar I, on the 30th of Adar I we use our standard methodology to combine results on a trial by trial basis. The probability that a text from the ELS random placement text population would produce as compact combined results as that produced by the Torah text is 206/1,000.

Yehonadav Chaim Hirschfeld

For the key words Y. Ch. Hirschfeld, Y. Hirschfeld or Hirschfeld with Adar I 30, 30th of Adar I, or on the 30th of Adar I there was no statistically significant table and there were no ELSs for Y. Ch. Hirschfeld. For the key word pair Y. Hirschfeld and Adar I 30, with expected number of ELSs set to 50, the probability that a text from the ELS text population would produce as compact a table as that produced by the Torah text is 138/1,000.

student 2
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 138/1,000. The cylinder size produced by the search was 4320 columns.
Finding by Professor Haralick

For the combined experiment using the appellations Y. Hirschfeld or Hirschfeld with the three date forms Adar I 30, 30th of Adar I, on the 30th of Adar I we use our standard methodology to combine results on a trial by trial basis. The probability that a text from the ELS random placement text population would produce as compact combined results as that produced by the Torah text is 587/1,000.

Roee Aharon Roth

For the key words Roth and Adar I 30 , with expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce as compact a table as that produced by the Torah text is 2.5/100,000. For the combined experiment using 9 pairs of appellations and dates, we use our standard methodology to combine results on a trial by trial basis. The probability that a text from the ELS random placement text population would produce as compact combined results as that produced by the Torah text is 11/100,000.

student 3
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 2.5/100,000. The cylinder size produced by the search was 88 columns.
Finding by Professor Haralick

For the combined experiment using the appellations R. Roth or Roth with the three date forms Adar I 30, 30th of Adar I, on the 30th of Adar I we use our standard methodology to combine results on a trial by trial basis. The probability that a text from the ELS random placement text population would produce as compact combined results as that produced by the Torah text is 11/100,000.

Segev Pniel Avichail

For the key words S. Avichail and Adar I 30, with expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce as compact a table as that produced by the Torah text is 18.5/1,000.

student 4
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 18.5/1,000. The cylinder size produced by the search was 6868 columns.
Finding by Professor Haralick

Using our standard methodology for combining results on a trial by trial basis, with 9 combination pairs of three appellations S. P. Avichail, S. Avichail, and Avichail with three date forms Adar I 30, 30th of Adar I, and on the 30th of Adar I, the probability that a text from the random placement ELS text population would produce as combined compact result as that produced by the Torah text is 140.5/1,000.

Yonatan Yitzchak Eldar

For the key words Y. Y. Elder and Adar I 30, with expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce as compact a table as that produced by the Torah text is 64.5/1,000.

student 5
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 64.5/1,000. The cylinder size produced by the search was 5,333 columns.
Finding by Professor Haralick

Using our standard methodology for combining results on a trial by trial basis, with 9 combination pairs of three appellations Y. Y. Elder, Y. Elder, and Elder with three date forms Adar I 30, 30th of Adar I, and on the 30th of Adar I, the probability that a text from the random placement ELS text population would produce as combined compact result as that produced by the Torah text is 457/1,000.

Neria Cohen

For the key words N. Cohen or Cohen with Adar I 30, 30th of Adar I, or on the 30th of Adar I there was no statistically significant table. In this case we thought that perhaps the four Hebrew letters that go with N. Cohen are not long enough and that in this case the first and last name should be put together. Therefore, we also searched the key words Neria Cohen with Adar I 30, 30th of Adar I, or on the 30th of Adar I. This search did produce an interesting result for the pair Neria Cohen with Adar I 30. With the expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 15.5/1,000.

student 6
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 15.5/1,000. The cylinder size produced by the search was 17,824 columns.
Finding by Professor Haralick

Using our standard methodology for combining results on a trial by trial basis, with 9 combination pairs of three appellations Neria Cohen, N. Cohen, and Cohen with three date forms Adar I 30, 30th of Adar I, and on the 30th of Adar I, the probability that a text from the random placement ELS text population would produce as combined compact result as that produced by the Torah text is 99/1,000.

Avraham David Moses

For the key words A. Moses with 30th of Adar I,, with expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce as compact a table as that produced by the Torah text is 50.5/1,000.

student 7
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 50.5/1,000. The cylinder size produced by the search was 320 columns.
Finding by Professor Haralick

In this case there is another relatively compact table with the key words A. Moses with Adar I 30,. With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce as compact as table as this one produced by the Torah text is 77/1,000. This table is smaller in area than the previous table but has a higher p-value. The reason is that the ELS for the date does not have a Beth which the previous table does.

student 7
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 77/1,000. The cylinder size produced by the search was 119 columns.
Finding by Professor Haralick

There is one other table associated with this student which is interesting, although not so much statistically. It is a table in which the ELS for the name and the ELS for the date have the same absolute skip. It has been noticed that when this happens, there is often more related material in or nearby the table. Such a search for related key words needs to be done for this case.

student 7
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 77/1,000. The cylinder size produced by the search was 47 columns.
Finding by Professor Haralick

Using our standard methodology for combining results on a trial by trial basis, with 9 combination pairs of three appellations A. D. Moses, A. Moses, and Moses with three date forms Adar I 30, 30th of Adar I, and on the 30th of Adar I, the probability that a text from the random placement ELS text population would produce as combined compact result as that produced by the Torah text is 369.5/1,000.

Yohai Lifshitz

For the key words Lifshitz and 30th of Adar I,, with expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce as compact a table as that produced by the Torah text is 27/1,000.

student 8
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 27/1,000. The cylinder size produced by the search was 89 columns.
Finding by Professor Haralick

In this case there is another relatively compact table with the key words Y. Lifshitz with Adar I 30,. With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce as compact as table as this one produced by the Torah text is 49/1,000. This table is smaller in area than the previous table but has a higher p-value. The reason is that the ELS for the date does not have a Beth which the previous table does.

student 8
With expected number of ELSs set to 50, the probability that a text from the ELS random placement text population would produce a table as compact as that produced by the Torah text is 49/1,000. The cylinder size produced by the search was 8948 columns.
Finding by Professor Haralick

Using our standard methodology for combining results on a trial by trial basis, with 6 combination pairs of two appellations L. Lifshitz, and Lifshitz with three date forms Adar I 30, 30th of Adar I, and on the 30th of Adar I, the probability that a text from the random placement ELS text population would produce as combined compact result as that produced by the Torah text is 149.5/1,000.

Total Combined Results

We have done a total of 8 experiments. Each experiment itself had 6 or 9 subexperiments. So far we have reported on the results for each of the 8 experiments. The p-values observed are the probabilities that the tables produced by the Torah text would have happened by chance. Formally, for each of the 8 students, we have tested the Null hypothesis of No Torah Code Effect against the alternative hypothesis that the Torah text produces more compact tables than expected by chance. For some of the 8 experiments the p-values were statistically significant. For others not so. Now we report on the p-value for the total of 8 experiments.

We use a combining methodology on a trial by trial basis. The results for each of the subexperiments of each experiment are combined by taking the minimum of the rank normalized compactness values. This gives a score for each experiment. Then we take the geometric means of the 8 scores. The normalized rank of the Torah text is the p-value of the experiment.

Setting the expected number of ELSs to 50 the probability that a text from the ELS random placement text population would produce a total combined result as or more compact than the Torah text is 490.5/100,000.