The lie detector is a lie, the polygraph is my bitch, and Americans should stop acting like slack-jawed yokels

Last week I found myself in a doctor’s waiting room for a few minutes, and the staff had the TV tuned to one of those daily Dr. Phil/Maury/Jerry/Montel type freak circuses where the host knows everything and fixes all human problems in 30 minutes. I tried to read my book and ignore it, but you know how hard it is not to look at a trainwreck. I was sort of doing okay up until I heard the host use a term that has griped me for years: “lie detector.” Yes, somebody is lying. We’ll find out who right after these messages.


I’d have thought we’d have this polygraph nonsense well behind us by now. There is such an accumulation of hard, scientific evidence demonstrating the unreliability of the process that it’s always a bit of a shock to hear the words spoken aloud.* I know, I know – I’m not exactly talking about a show aimed at the Northeastern Liberal Intelligentsia here. That genre of syndicated afternoon entertainment targets society’s least adept thinkers by and large, so I should expect a medicine show to be a medicine show, complete with Amazing Displays of Alchemy, Phrenology, Astrology, Polygraphy, Parapsychology and Theoretical Economics.

Unfortunately, the polygraph isn’t restricted to the domain of tabloid television. It’s also regularly accorded a measure of credibility in official contexts. It is sometimes allowed in court cases, it is routinely employed by law enforcement and you can be denied employment if you are deemed to have “failed” a test. In other words, we still allow people’s lives to be jacked by charlatans with Ouija boards.

I think I’ve always taken this silliness personally because there was a period of a few days back in the mid-’80s when it was, in fact, personal. I was a bartender for a restaurant in the now-defunct Darryl’s chain, and the geniuses who ran the place concluded that there was some unacceptable behavior going on behind the bar. I was never sure where the suspicion came from, but I have always imagined that their liquor costs spiked a few pennies one week and they freaked out. Anyway, they decided to administer “lie detectors” to all of us so as to root out the perpetrator(s).

You have to understand: nothing illegal was going on. Sure, there was the usual stuff you find at a bar – somebody buys a regular a drink in a way that isn’t sanctioned by management, for instance. But there was no substantive bad behavior. If there had been the other bartenders would have known about it. We worked in exceedingly tight quarters and it would have been damned hard not to be noticed. We wouldn’t have ratted our guilty co-worker out, mind you, but we’d have known what was going on.

We all immediately realized we were the subjects of a witch hunt and I, for one, was terrified. I was innocent of any crime more serious than jaywalking, but all of a sudden they were going to use the scientific equivalent of biorhythm analysis to deprive me of my livelihood.

On the day of the test I was so nervous that it’s amazing the machine didn’t accuse me of lying about my name. But I soldiered through. During the exam I was asked a few “fire questions” – that is, have you ever _________, where that blank is a firing offense. Have you ever given away a free drink without a manager’s approval? Have you ever overpoured a drink? Etc. In three or four cases I had, in fact, committed the offense in question.

Of course I’ve overpoured a drink, dumbass. Every bartender has overpoured a drink. And sure, I’d given away a drink or two. That builds relationships and gets customers back to the bar, and the cost is nonexistent compared to the benefit. On a busy night there may be no time to track down a manager and make the case for approval, so yes, I had made a decision to act in a way that benefited me, the restaurant and the customer.

But if I answered truthfully, I was toast. So I lied. On three or four questions, I lied to the lie detector. Guess what? It couldn’t catch me. And a few days later I found myself being personally congratulated by the Regional Manager for my honesty and integrity.

One of my colleagues, who was a really good, really popular bartender, broke. He couldn’t bring himself to lie, so he admitted his transgressions. The management liked him a lot so they didn’t fire him, but they did kick him back to the wait staff. Could have been worse, I suppose.

In any case, this is why my disgust with the purveyors of polygraphy is so acute. It’s more than an intellectual disdain for people who aren’t smart enough to understand even the rudiments of scientific and statistical analyses. It’s a personal, firsthand understanding that transcends the research and it’s an empathetic response for the anxiety that innocent people suffer.

The polygraph is my bitch. And anytime I hear the term “lie detector,” I know I’m listening to a pseudoscience-pandering huckster’s dupe. I can’t help it. If you use the term, my mind intuitively slots you into the category of “people who watch afternoon TV and believe it.”

That may be unfair, but it’s an observation that has stronger supporting evidence than the polygraph does.


* Forgive my laziness in citing Wikipedia here, but I’m on a time budget this morning. This section does a decent job of pulling together the data. Click the links for more from the original sources.



  • Sam,

    I couldn’t agree more with you about what a load of BS the polygraph is. You figured out its nothing more than a prop, meant to con and intimidate the naive. It appears that the polygraph operators song and dance didn’t let you buy in that it worked. Your post more than proves that it can and has been beaten. The ultimate countermeasure to the polygraph is dis-belief and knowledge of how the scam works. Knowledge overcomes fear, which is the real tool they are trying to use. Thanks for a great post.

  • Barry Cushman

    You probably actually support that polygraph works. The questions to which you lied sound like they were comparison questions. They are questions to which you are either expected to lie or to which you are unsure of your answer. They are compared against the relevant questions.

    In 2003 the NAS looked at polygraph and concluded specific issue testing works, but like any test, not perfectly. They found a median accuracy of .86 and .87 for the lab and field, respectively.

    Employee polygraphs in the 80s were, sadly, a ruse to make a lot of “chart rollers” rich. Thanks to the EPPA, that stopped. If you’re test took less than a couple hours, it was a sham.

    In the interest of full disclosure, I am an examiner. However, I reviewed the evidence (pro and con) before training, and I entered the field because the overwhelming evidence supports the usefulness of polygraph. “Polygraph,” however, means a lot of things, and not all polygraph is equal.

  • Barry: the questions I refer to were the money questions. There were other money questions, but they were equally on point about specific actions. The rest of the stuff was the baselining. (I’ve done a good bit of reading about the process and have a Psych degree as well – so I’m not ignorant about how this and other kinds of testing work).

    Your .86 number is WAY higher than anything else I’ve seen. Most credible research shows that polygraphs are at best only marginally better than chance, and in those cases there are likely extraneous factors related to the subject’s state that more than explain the variance.

    Further, even if I accept the .86 number, that means that for every 100 subjects you interview, you’re going to screw 14 of them. And since you pretend to the mantle of scientific accuracy, you’re going to do so in a way that looks more damning and leaves less recourse than if you’d simply put them in a dark room and shined a bright light in their face.

    I’d accept that not all polygraph is equal. But all the evidence we have before us makes clear that the variance isn’t good-to-bad, it’s bad-to-worse.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s