This following quotation is from a long article that I decided not to publish about figuring out Falcon’s identity at Birther Report:
There are two approaches to discovery: One is the conspiracist method, where one starts at a point and looks for connections and confirmation. Anything remotely related adds to the confirmation pile. This is what led me astray twice. Coincidences and anomalies are the way the world works, and often mean nothing. The other is a more scientific method that starts with a hypothesis, makes predictions, and by experiment tests those predictions.
Birthers start with the initial opinion that something is wrong with Barack Obama and that they only need to figure out what it is. Loren Collins, in his book Bullspotting, develops the idea of anomalies in conspiracist thinking, introducing the topic by saying:
A primary feature of virtually every form of creative misinformation, from denialism to conspiracism to pseudo-scholarship, is anomaly hunting.
… Instead of stumbling across anomalies in the pursuit of positive evidence and then attempting to explain them, denialists, conspiracists, and pseudo-scholars seek out anomalies directly. Having been confronted with a consensus view that they wish to undermine , they gather up anomalies and then attempt to use the collective weight of those individual unanswered questions as part of their argument that the consensus view is wrong.
Collins, Loren (2012-10-30). Bullspotting: Finding Facts in the Age of Misinformation. Prometheus Books. Kindle Edition.
(For example, birthers, using anomalies, arrived at the conclusion that Obama’s birth certificate is a forgery, while a scientific approach leads to the conclusion that it’s a normal scan of a paper document done by a Xerox machine and rotated by Preview on a Mac.)
In the same way that anomalies are sought by birthers, so are connections. Sometimes birthers do bad math to calculate the probability of their anomalies (see my article, “Miscalculating the odds”), although I haven’t yet seen them do this with connections.
Birthers theories are full of crazy connections, for example connecting President Obama’s social-security number from a series usually assigned to Connecticut residents, to the Sandy Hook shooting in Connecticut, to an abandoned psychiatric hospital near Sandy Hook where presumably patients with Connecticut social-security numbers were treated. (See “Was Sandy Hook Used to Bury Obama’s SSN Records?” by “Barry Soetoro, Esq” at the Memory Hole blog.) And what’s this Subud thing?
A birther at Orly’s site once “connected” me to something called the “Davidson Media Group” that, as I recall, owns a group of multicultural radio stations and has an address in a town where an erroneous public record says that I live, but that I never did. I then got “connected” to three individuals who were on the board of directors of Davidson Media Group, including one person who was previously at Citigroup. So now Citigroup is backing me and the blog. I’d never even heard of Davidson Media Group, but Orly’s commenter generalized on this “connection” concluding: My point is, that those accusing Obama are commen folk, like Orly, those defending him have big money behind them… [ellipsis, spelling, and grammar errors in the original].
I don’t want to single out birthers for criticism because we all have a tendency to notice connections; however, some of us, either by innate talent or methodology, are better at weeding out things that don’t make sense. I’ve certainly taken the connections route to try to answer a question about who an anonymous person online really is. I might ask the question, how likely is it for two people who post pictures from the same Photobucket account to be the same person, or how likely is it for two people using the screen name “Patriot1776” at different forums to be the same person? As often as not, that approach leads to the wrong answer. Emil Karlsson writes about the statistical fallacy of trying to assign a probability to things that have already happened at the Debunking Denialism blog in his article, “The Top Five Most Annoying Statistical Fallacies.” He gives this example:
Imagine the silliness in getting a particular bridge hand, then exclaim that you could not possibly have gotten that particular hand that you just got since the probability is astronomically low.
The more scientific approach is to state a hypothesis that makes predictions, and then make observations that have the chance of confirming or disproving it. And very importantly, don’t ignore those disproving observations1! I’m reading a delightful book right now titled “The Mapmaker’s Wife” in which the testing of a scientific prediction, the bulging of the Earth predicted by Newton’s theories in the 18th century, is a major subtopic. They measured a degree of latitude at the equator (in what is now Ecuador) and compared it to a similar measurement in France. It’s a fascinating narrative comparing good and bad science.
1The issue of deciding when to label something as “bad data” is difficult. The confirmation of Falcon’s name was overwhelming except for one item: Falcon said more than once that his grandfather died in 2002, but the grandfather of the person I identified died in 2001. I concluded that in face of so much solid confirmation, Falcon had just gotten the year wrong.
2I would have liked to illustrate the scientific method by the hypothesis made for the Falcon identification, but to do so in any detail would be to give hints about his identity, which I don’t intend to do. Let me just say that the hypothesis predicted that a certain document could be found, and that when it was, it would say two very specific things. The document was found, and it said what was predicted.