Slashdot Mirror
Math Says Conspiracies Are Prone To Unravel (bbc.com)
An anonymous reader writes: Who doesn't love a good conspiracy theory? Well, I don't — they're usually annoying daydreams from annoying people. Fortunately, an Oxford mathematician seems to feel the same way. Dr. David Grimes just published research in PLOS One establishing a formula for determining the likelihood of a failed conspiracy — in other words, how likely some of its participants are to spill the beans. There are three main factors: number of conspirators, the amount of time passed since it started, and how often we can expect conspiracies to intrinsically fail (a value he derived by studying actual conspiracies that were exposed). From the article: "He then applied his equation to four famous conspiracy theories: The belief that the Moon landing was faked, the belief that climate change is a fraud, the belief that vaccines cause autism, and the belief that pharmaceutical companies have suppressed a cure for cancer. Dr. Grimes's analysis suggests that if these four conspiracies were real, most are very likely to have been revealed as such by now. Specifically, the Moon landing 'hoax' would have been revealed in 3.7 years, the climate change 'fraud' in 3.7 to 26.8 years, the vaccine-autism 'conspiracy' in 3.2 to 34.8 years, and the cancer 'conspiracy' in 3.2 years."
The crucial value he is using how often do conspiracies fail, but then uses failed ones to measure the length of time. Isn't that kinda like asking how long until your car explodes, and only looking at cars that explode as your data. On top of that, looks like he is using only a same size of 3 to determine this metric making it even more questionable. While I applaud the effort, this doesn't seem to convincing.
One problem with this analysis is that it doesn't take into account *successful* conspiracies.
Suppose there are conspiracies which succeeded completely - in that the public was defrauded, suspected nothing, and life went on as normal.
If we are using past performance to predict future trends, shouldn't those conspiracies be counted? There's no realistic way to account for or even detect them.
Take for example the 1968 presidential campaign of Richard Nixon.
During that campaign, [incumbent president] Johnson was negotiating with Vietnam to bring an end to the Vietnam war.
Nixon though that this action would ruin his chances of being elected, so he contacted the Vietnamese government and said that if they obstructed talks, they'd get a better deal when he was elected.
(An example of an American interfering with the political process, prolonging a war for 7 more years, with enforced conscription, and causing the unnecessary deaths of tens of thousands of Americans.)
This action was known to Nixon's campaign manager (Mitchell) and several aides. Johnson knew about it (a tape in the Johnson presidential library has Johnson denouncing Nixon for “treason”)
Neither side wanted to push the issue, so it was dropped.
This was a conspiracy, involved several dozen people (including FBI agents), and was monstrously important at the time. It took 50 years for the documents to be released describing the situation. Johnson's tape was released in 2008, and some other files are still hidden.
I don't have a lot of faith in this paper - it doesn't take into account conspiracies that actually succeed.