John Sullivan, in the
introduction to his book,
Gatekeeper: Memoirs of a CIA Polygrapher Examiner (now scheduled for release in April 2007 according to the publisher) writes, among other things:
Quote: Shortly after my book Of Spies and Lies came out, I went back to Long Island, New York, for a class reunion. The brightest guy in my class offered to bet me that he could “beat” me on a polygraph test. My usual answer to such queries is, “Polygraph isn’t a parlor game and running tests as part of a bet trivializes my work.” But this time, I made him the following offer: “I’ll run a test on you, but we will do it the way it is really done. What that means is that you put up some serious money. If I can catch you in a lie, the money is mine.” In a real test an examiner’s ego may be bruised if he is beaten on a test, but the subject has the most to lose.
“Here is the test I will conduct. On a sheet of paper, I will write the numbers from 1 through 10, and ask you to circle one of the numbers between 3 and 8, without telling me which number you circled. I will then ask you ten questions, each beginning with have you circled number 1, 2, 3, 4, 5, etc.? Number 1, 2, 9 and 10 are ‘dead’ numbers. You are to answer each of my questions with a ‘No.’ One answer will be a lie. If I identify the number you circled, you have to pay me the money. That’s the way it is in the real world of polygraph. The pressure is on you, not me.” My classmate declined to take me up on my offer.
I will accept your offer, Mr. Sullivan, provided we
both put up serious money. If you can't divine the number I circled, I take the money. If you succeed, you take it. Your nominal odds of success are 1 in 6, the same as those of taking home the grand prize in Russian roulette.
While you maintain that "polygraph isn't a parlor game," the wager you offered your classmate resembles nothing so much as a parlor game. You were less than honest when you told your classmate that the numbers test you described is "the way it is in the real world." The procedure you described is a seldom-used Peak of Tension test (POT). The great majority of polygraph examinations administered in the "real world" by agencies such as the CIA are of either the Control Question Test (CQT) or Relevant/Irrelevant (R/I) variety. Polygraphers like to use the POT for public demonstrations because 1) it has a better chance of succeeding than an actual lie test of the CQT or R/I variety and 2) it doesn't disclose the trickery on which the CQT depends.
So what say you, Mr. Sullivan? Are you game? I say the pressure would be on
you, not me.