Debunking a JFK Mysterious Witness Death Lone Nutter
Aug. 6, 2013
Updated: March 8, 2014
Click Reclaiming Science:The JFK Conspiracy to look inside the book.
The author writes:
As something of a footnote, conspiracists have long played up the number of “mysterious deaths” associated with the JFK assassination, though an inspection of the list makes it seem substantially less mysterious. In any case, the bottom line of a half-century’s investigation of the assassination is that we are left with the conclusion that was apparent from the start: JFK was killed by a lone assassin named Lee Harvey Oswald.
One of the most preposterous claims of the conspiracists is that there have been large numbers of “mysterious deaths” of witnesses relevant to the JFK assassination. Jim Marrs, in his 1993 book (sic 1989) CROSSFIRE, cited 103 “mysterious deaths” up to 1984.
Conspiracists assert that odds of such groupings of deaths are so low that it is impossible to believe they couldn’t have been part of a plan, in particular citing a 1967 LONDON SUNDAY TIMES article with an early “mysterious deaths” list accompanied by a claim that an actuary had calculated the odds to be “a hundred thousand trillion to one”.
The HSCA found this citation of odds a little hard to believe. The list of people who could be connected to the JFK assassination was long, easily running to thousands, and the idea that a portion of them might have died over some period of years hardly seemed that improbable. The HSCA contacted the TIMES and got back a sheepish answer. It turned out that the question the paper had asked of an actuary was effectively:
Name 15 specific adults selected at random from the population of the USA. What are the odds that all 15 of these named people will be dead within a few years?
The odds are not at all good. Assume that adults in a population have, on the average, a 1 in 10 probability of dying in some given number of years. If 15 adults are selected at random from that population, the odds of all 15 dying to that time would be 1 in 10^15, a thousand trillion to one.
However, anybody with even a simple understanding of probability would know that was asking the wrong question. The right question was obviously: Given a group of several thousand people, what are the odds that at least 15 unspecified members of that group will be dead in a few years?
The answer was that one could bet on it and easily win. Given 1 in 10 odds of an adult in a population dying in some given number of years, we would expect in that time that, duh, roughly a tenth of the population would be dead. The TIMES apologized to the HSCA for the blunder.
The article is a total fiasco
The Lone Nutter’s ignorance and naivete is confirmed by this utterly false statement: “However, anybody with even a simple understanding of probability would know that was asking the wrong question. The right question was obviously: Given a group of several thousand people, what are the odds that at least 15 unspecified members of that group will be dead in a few years? The answer was that one could bet on it and easily win. Given 1 in 10 odds of an adult in a population dying in some given number of years, we would expect in that time that, duh, roughly a tenth of the population would be dead.”
The author calls legitimate seekers of the truth “conspiracists”. But he does not understand the problem, much less the math. This is the correct definition: Given a group of N people, what is the probability that at least n members of the group will die unnaturally (homicide, accident or suicide) within T years?
He just repeats the usual Warren Commission apologist talking points. The mathematical proof of a conspiracy relegates his screed as pure propaganda. Lone Nutters are shameless and have no regard for the truth.
The author’s lack of specificity is the “tell”. He fails to consider any of the following critical factors: the total number of material witnesses in the group, the number and cause of unnatural deaths, the time interval, unnatural mortality rates. All are necessary input for the probability calculation.
This LN does not even qualify as an amateur. But that’s understandable. After all, he’s a Lone Nutter, who by definition is incapable of rational analysis.
His logical errors and omission of key factors in his “analysis” include:
1) not assuming a specific number of witnesses in the target group,
2) invention of a 1 in 10 probability of dying,
3) not assuming a definitive time period,
4) failure to consider unnatural deaths and related mortality stats,
5) use of a pathetically, unscientific probability calculation,
6) naively states that “roughly a tenth of the population would be dead”.
7) failure to refute the relevance of 100 “convenient” deaths
8) failure to consider the more than 60 deaths of witnesses sought to testify.
9) failure to correctly calculate the expected number of unnatural deaths:
E=N*T*R, where N =total witnesses in the group, T= time period, R= weighted average mortality rate.
10) failure to consider the “paradigm shift”: why the witnesses died is a non-factor. The only relevant factors are how many died unnaturally, the time interval and the universe of material witnesses or the number of witnesses called to testify.
There are 120 dead material witnesses in the JFK Calc spreadsheet database based on a 1400+ total universe. Of the 120, 63 were sought to testify at the Warren Commission, Garrison trial, Church Senate and HSCA.
The author does not consider that the number of UNNATURAL deaths among the 1400+ witnesses is the key factor – not total deaths. There was a STATISTICALLY IMPOSSIBLE 77 OFFICIALLY RULED UNNATURAL DEATHS (34 homicides, 24 accidents, 16 suicides, 3 unknown).
In fact, 25 of the 40 accidents and suicides were actually HOMICIDES – based on the STATISTICAL EXPECTATION of 12 accidents and 3 suicides – so we are up to 59 homicides among the 77 unnatural deaths. But by the same reasoning, there was a statistically impossible number of “natural deaths”: HEART ATTACKS and SUDDEN CANCERS. Therefore, the 34 OFFICIAL homicides UNDERSTATES the true number (estimated as 90+) based on STATISTICAL EXPECTATION.
The Paradigm Shift
But his most fundamental flaw was focusing on the relevance of individual witnesses in Marrs’ list without considering the paradigm shift: WHY the witnesses died is IRRELEVANT.
The relevant factors are how many witnesses were called to testify, how many died, their cause of death and the time period. In fact, from 1964-78, approximately 1100 witnesses were called to testify in four investigations. At least 63 died (38 unnaturally, including 27 homicides).
The author claims there were thousands of witnesses. In fact, the FBI claimed to have interviewed 25,000 (only about 1400 were material). But let’s assume there were 25,000.
There were at least 25 homicides of material witnesses from 1964-66. The probability of at least 25 homicides among the 25,000 is 1 in 38 BILLION (2.6E-11). The average homicide rate for 1964-66 was 0.000061.
There were at least 83 homicides from 1964-78. The probability of 83 homicides among the 25,000 is 1 in 43 TRILLION (2.32E-14). Only 32 homicides would normally be expected. The average homicide rate for 1964-78 was 0.000084.
The data and probabilities are displayed in JFK Calc: A Spreadsheet/Database of Mysterious Witness Deaths.
Statistically expected number of unnatural deaths
Expected unnatural deaths: E = N*T*R, where
N = 1400 material witnesses
T = 15 years (1964-78)
R = .000818 average unweighted unnatural mortality rate
JFK Material witnesses unnatural mortality
Among 1400 material witnesses from 1964-78, 77 deaths were officially ruled as unnatural – but only 17 were statistically expected: 34 homicides (2 expected); 24 accidents (12 expected); 16 suicides (3 expected) and 3 unknown (0.2 expected). There were 40 deaths officially ruled as accidental or suicide – but only 15 were expected. Therefore it is likely that approximately 25 (40-15) accidents and suicides were actually homicides.
Expected vs. Official Unnatural Death (1964-78)
Cause Expected Official Mortality Rate
homicide 1.76 34 0.000084 44%
suicide. 2.91 16 0.000130 21%
accident 12.47 24 0.000594 31%
unknown. 0.21 3 0.000010 4%
Total 17.35 77 0.000818 100%
According to the CIA, N= 418 witnesses testified, but the total was 552 including affidavits and depositions.
There were at least n= 18 unnatural deaths over T= 15 years (1964-78). The probability of at least 10 unnatural deaths among the witnesses in 3 years is:
P = 1 – poisson(9, 1.06, true) = 1.81E-07 (1 in 5,509,693)
(based on the 0.000842 national unnatural rate)
P = 1 – poisson(9, 0.31, true) = 1.53E-12 (1 in 652,270,204,558)
(based on the 0.000245 JFK witness-weighted unnatural rate)
The London Sunday Times Actuary
The actuary’s 100,000 trillion to one odds of 18 material witness deaths in three years (13 were unnatural) is matched by assuming 459 witnesses and the 0.000207 weighted unnatural mortality rate. Only one unnatural death would normally be expected among 459 witnesses in the three year period.
The probability is:
P= POISSON (13, 0.29, false) = 9.96E-18 (1 in 100,000 TRILLION)
Convenient deaths spiked in 1964 (Warren Commission) and 1977 (House Select Committee).