AgTalk Home
AgTalk Home
Search Forums | Classifieds (31) | Skins | Language
You are logged in as a guest. ( logon | register )

LandLord Request - Biblical, Rest land every 7 yrs? No crop ~
View previous thread :: View next thread
   Forums List -> Crop TalkMessage format
 
Pofarmer
Posted 3/4/2013 12:47 (#2941258 - in reply to #2940192)
Subject: Re: LandLord Request - Biblical, Rest land every 7 yrs? No crop ~



Or maybe not.

Criticism of the original paper

In 1999, four authors, the Australian mathematician Brendan McKay, the Israeli mathematicians Dror Bar-Natan and Gil Kalai, and the Israeli psychologist Maya Bar-Hillel (collectively known as "MBBK") published a paper in Statistical Science, in which they argue that the case of Witztum, Rips and Rosenberg (WRR) is "fatally defective, indeed that their result merely reflects on the choices made in designing their experiment and collecting the data for it."[15] The MBBK paper was reviewed anonymously by four professional statisticians prior to publication. In the introduction to the paper, Robert Kass, the Editor of the Journal who previously had described the WRR paper as a "challenging puzzle" wrote that "considering the work of McKay, Bar-Natan, Kalai and Bar-Hillel as a whole it indeed appears, as they conclude, that the puzzle has been solved".[7]

In the MBBK paper, the authors present the following arguments:

that because of problems in WRR's test method, the results of WRR's 1994 paper "may reflect (at least to some extent) uninteresting properties of the word list [the appellation-date word pairs] rather than an extraordinary property of Genesis,"[16] and that the test method used by WRR has properties that make it "exceptionally susceptible to systematic bias."[17]
that WRR had many choices available when selecting the appellations, the dates, and the date forms.[18]
that, despite the claims by WRR and S. Z. Havlin that Havlin prepared the appellations independently, the earliest available documents on the experiments do not state that the lists of appellations were prepared by an independent expert. Similarly, the authors quote a 1985 lecture by Eliyahu Rips, in which he describes the appellation selection process as taking "every possible variant that we considered reasonable", and makes no mention of Havlin or an independent expert.[19]
that, by adding some appellations and removing some appellations from WRR's list 2, and then repeating the test on the initial 78,064 letters (the length of Genesis) of a Hebrew translation of War and Peace, they achieved a significance level of one in a million. The authors say this shows that "the freedom provided just in the selection of appellations is sufficient to explain the strong result" in WRR's 1994 paper.[20]
The authors present a "study of variations", in which they repeated the experiments many times, each time making one or more minimal changes to the test method, the dates, or the date forms.[21] The authors conclude that "only a small fraction of variations made WRR's result stronger and then usually by only a small amount ... we believe that these observations are strong evidence for tuning ... "[22]
The authors present experiments they conducted which they say show that some results of experiments by WRR and Harold Gans are "too good to be true." That is, some of the results are statistically improbable even if one accepts that WRR's hypothesis is true. The authors say that these studies "give support to the theory that WRR's experiments were tuned toward an overly idealized result consistent with the common expectations of statistically naïve researchers."[23]
The authors present multiple experiments they conducted in which they attempted to replicate WRR's experiments. The authors used an independent expert to prepare the appellations and dates for each of these experiments. The authors report that the results of these attempted replications were negative.[24]
that the available evidence indicates that the text of Genesis used by WRR is substantially different from its original form, and that ELSs with large skips (which WRR's experiments rely on) could not survive such changes.[25]

From these observations, MBBK created an alternative hypothesis to explain the "puzzle" of how the codes were discovered. MBBK's claim, in essence, was that the WRR authors had cheated[26][27] MBBK went on to describe the means by which the cheating might have occurred, and demonstrate the tactic as presumed.

MBBK's refutation was not strictly mathematical in nature, rather it asserted that the WRR authors and contributors had intentionally or unintentionally (a) selected the names and/or dates in advance and (b) designed their experiments to match their selection and thereby achieved their "desired" result. The MBBK paper argued that the ELS experiment is extraordinarily sensitive to very small changes in the spellings of appellations, and that the WRR result "merely reflects on the choices made in designing their experiment and collecting the data for it."

The MBBK paper demonstrated that this "tuning", when combined with what MBBK asserted was available "wiggle" room, was capable of generating a result similar to WRR's Genesis result in a Hebrew translation of War and Peace. Psychologist and MBBK co-author Maya Bar-Hillel subsequently summarized the MBBK view that the WRR paper was a hoax, an intentionally and a carefully designed "magic trick".[28]

The Bible codes (together with similar arguments concerning hidden prophecies in the writings of Shakespeare) have been quoted as examples of the Texas sharpshooter fallacy.
Top of the page Bottom of the page


Jump to forum :
Search this forum
Printer friendly version
E-mail a link to this thread

(Delete cookies)