In a couple of e-mails, Walter laid out his proposal to me, about narrow bands of terms in a codes matrix. The central thesis is whether the apparent banding of terms constitutes the means by which we can judge whether a term is valid or not? Therefore it is a worthy proposal to discuss.
What is not apparent is whether we should build constraints in a Bible code program that in effect places training wheels, by limiting displayable terms to those in a narrow band. Walter proposes calling the center or main term of the matrix, the theme or thematic term, a term about which the whole matrix pivots.
Another issue raised by the idea of banding, is whether we are understating the real statistics of the matrix by calculating them as if any ELS of a term is usable? This is a real issue if we know that only terms in very narrow ELS bands are in fact valid terms in that matrix.
Walter originally wrote:
As you know, I believe that within certain ELS bands of a matix, the terms are found that apply to that matix. This may not be altogether true, but it is my contention.
I saw that you mentioned bands in a previous e-mails and I wasn't sure what you were talking about. Let me give an example. When I did the Edison life matrix long ago to prove to myself whether the Bible code was true or not; the center term, Edison, was found at -1231, and I saw that there were clumps of data at 205 row letters distance from the center term. That's what taught me about row-splits. A row-split of 6 was optimal for the Edison matrix, therefore the matrix ELS was 205 (1231/6=205). The terms in my matrix were in the following bands.
All the terms in the matrix are in these narrow ELS bands. Is that what you meant by bands? Actually, I found 90% in the band of plus or minus 7 of the center point of the band, but expanded it to plus or minus 10 for a few of Edison's inventions. The narrow band is the way I have been doing the Bible code from the beginning.
Yes, Roy, this is exactlty what I mean by ELS bands -- people incorporate these whether they realize it or not Also, in my matrixes, I have adopted a means of identifying these bands by using word pad to edit the matrix report. Here, I will do this for you. It might be a good idea if a system was used so that the group could refer to a specific band with mutual understanding.
1-10 LT W11 (lateral band width 11)
195-215 (L5 W21)
400-420 (L4 W21)
605-625 (L3 W21)
810-830 (L2 W21)
1015-1035 (L W21) Lower band width 22
1221-1241 (THM W21) Theme band width 21
1425-1445 (U W21) Upper band width 21
This is an example of how this matrix might be defined. Often times I call the row split 2 band the L band. See? But this is a higher skip matrix. All I'm saying is that the bands are there. With this example, you have confirmed it. Therefore, I believe that a system should be agreed upon so that we could discuss bands without stumbling over definitions. I don't care how we label them.
One time I got down to a matrix that was six characters wide. It had a new theme term -- that was related to the original theme term. And, of course, new terms appear in the lower bands that can't be seen when you start out. Also, "let skip wrap" has to match matrix width in the lower bands. See?
This is interesting stuff to me Roy -- it is pure coding.
CF heaps evaluation on to the lateral band, but the messages found in the higher bands are equally important. See? This is your software telling you how to code -- which brings us to another subject -- "band evaluation" based on band width.
Sometimes a new theme term will appear between the letters of the theme band in one of the L bands (as I mentioned above). I like to use "theme term," instead of "main term" because it is more descriptive of what it actually is.
There are times when the original theme term is not as "generally" descriptive as a term that is found subsequently. It has happened to me that, in these instances, switching the theme term can lead you to a smaller box. See?
For further research and discussion:
From Roy: What Walter discussed in banding of terms, is also the way that I have done Bible code work since I began. However, I have never stated that only terms within the bands are valid for that matrix (even though that is what I believed to be true). I happen to dislike the idea of building training wheels into a program that would limit the displayable terms only to those found within the narrow bands. I like seeing all of the "found" terms. However, if it can be proven that the only valid terms in a matrix are those within the bands, then certainly training wheels could easily be built into the software. This would greatly ease the analysis of terms for people new to the Bible code. If software had these training wheels, it would need to have an "override" or "turn-off" feature for experienced coders.
If bands are in fact the limits for valid terms in a matrix, then it spills over also into the way that statistics are calculated. Instead of calculating Expected Occurrences as if the term could randomly be found within the matrix at any ELS, should the expected occurrences be calculated for the possibility of occurence only with the narrow bands defined by the center/main term, or as Walter calls it, the thematic term?