🚀 go-pugleaf

On 8 Jun 1997 03:44:33 GMT, da...@taic.net wrote:
>	Here is my proposition: The ability of a text to have predictive
>value goes hand-in-hand with statistical relevancy. If I cannot look at a
>text *before* an event, and find statistically meaningful and relevant
>information with predictive value, then there can be no statistically
>relevant "investigation" of such text after the fact to find hidden
>meanings. Not with any degree of intellectual honesty.

No, you have entirely missed the point, I'm afraid. Let's say I find
"datataic" encoded near the word "Moshiach" (Messiah), and the date
9Sivan5757. Now, does that mean you are the Messiah, and will be
revealed in about 6 days? Does it mean that you will open Moshiach's
Bar and Grill in 6 days? Or is it not talking about you at all??
Answer: we don't know, and we can't know, and the words have no
predictive value. Please read those releases I posted more carefully,
and you will see this explained in detail. As Doron Witztum wrote, he
found something he _could_ have presented as a code predicting
Churchill's assassination!

But after the fact, it is certainly appropriate to go back and do
research as was done by Rips, Witztum and Rosenberg. What they
discovered was a pattern whose likelihood of chance appearance was
vanishingly small. This wasn't presented through hokus-pocus
blind-my-eyes presentations of codes without statistical value, but
through serious research presented in the statistics community. This
sort of research certainly does not depend upon the ability to predict
the future.

YM