Re: Comments on Snoke's approach

From: Iain Strachan <>
Date: Fri Sep 23 2005 - 03:35:49 EDT


You wrote:

> Let me start with the well-worn, oft-used analogy of dealing a hand of
> bridge. Pick up your cards and no matter what cards you have, you could
> truthfully exclaim that the probability of your being dealt that particular
> hand is infinitesimally small. But you wouldn't be justified on that basis
> in accusing the dealer of cheating and manipulating the cards. However, if
> prior to dealing the cards, someone had written down a possible hand and if
> after the hand is dealt the cards match that specific pattern, you would
> indeed be justified in suspecting foul play. The point is that merely
> having an extremely low probability of occurrence is not an argument for
> cheating--or for design. Consideration must be given to the bigger picture
> such as the number of combinations possible. For a hand of cards, the
> number of possibilities is also vast so that the probability of having a
> low-probability hand is actually one hundred percent.
> When applied to your numero/geometrical findings, it isn't nearly as
> easy to calculate the number of possibilities as it is in a deck of cards.
> But it is fair to say that the total number of possible geometric or
> numerical results is incredibly vast and that every one of them has a low
> probability of occurring. As in the deck of cards, whatever combination
> arises, it will be a low-probability combination. Even if the combination
> has some degree of interest, there is no significance whatsoever unless
> there is a specific prior detailed articulation of the pattern to be
> expected. No, I'm sorry but Rev. 13:18 doesn't even come close to such an
> articulation.

Your hand of cards example is a good one, but I think the same
arguments you make about pre-specifying the hand of cards and then
getting it being "foul play" can reasonably be applied to Vernon's
findings. I don't have definitive answers yet, but I think one can
find a genuine low-probability figure that actually means something,
by using the concept of description length and Kolmogorov complexity
(indeed I shall shortly be using the concepts of Minimum Message
Length/Minimum Description length in my professional work).

The key point is that it takes a certain length of description to
describe 13 cards. If there were another random hand, to describe the
entire sequence of 26 would take twice as much information to describe
it. But if, as in your case, a pre-specified hand then turns up in
your hand, then the entire sequence of 26 can be described in not much
longer than the 13 because you give the 13 and then say "same again".
Given this you can then calculate your very low probability and claim

A similar example can be given with coin-tossing. Toss a coin 100
times and the sequence probability means nothing, even though it is
2^-100. But toss it again and get a repeat of the same sequence, and
it does mean something because the entire sequence of 200 can be
described in a lot less than 200 bits. If a sequence of N bits can be
described in M bits where M<N then the probability of that happening
is 2^(N-M) and that will be a meaningful figure.

You say that the number of possible geometric codings of Vernon's data
must be vast, but each time a symmetry is produced, in principle it
means that a reduced description length can be found, and a meaningful
probability ascribed to it. I only looked briefly at Vernon's data to
see how this could be done, and achieved some low probability
description-length based answers. They weren't as low probability as
some of the figures Vernon is claiming, but they were still
sufficiently low to indicate that some form of numerical design is
present in the Gen 1:1 sequence. All this does not speculate on who
the designer might be. A history of Maths professor I know who is an
atheist was equally convinced that Gen 1:1 was designed when I showed
him the patterns, but he was convinced that it was human and not
divine design and that this sort of thing went on a great deal in the

Received on Fri Sep 23 03:37:08 2005

This archive was generated by hypermail 2.1.8 : Fri Sep 23 2005 - 03:37:08 EDT