Hacker News new | past | comments | ask | show | jobs | submit login
An Ex-Cop's War on Lie Detectors (bloomberg.com)
144 points by new1234567 on Aug 4, 2015 | hide | past | web | favorite | 64 comments

The most despicable use of lie detectors I've ever seen was on the Dr. Phil show. Dr. Phil asked a man who cheated on his wife whether he diddled his kids while he was hooked up to a lie detector. He answered no but obviously a question that is as emotionally charged at that will appear as a lie. This man's relationship with his family, friends and probably any future work relationships were all ruined by a stupid magic trick. The man turned to Dr. Phil and the the former FBI agent who administered the polygraph and start pleading and begging for any reason why the lie detector would give a false positive. Dr. Phil and the former FBI agent were smart enough to know why it wasn't working but they continued be evasive as the man's past and future relationships with his children, family, friend, and coworkers went up in flames.

I hate to say it, but that's what you get for seeking help from a TV psychologist.

You're insane if you think a TV show is there to help you.

He's not a Psychologist or any sort of a doctor at all. Just another modern day Jerry Springer act.

In a real polygraph exam a baseline is established and you review all the questions "before" the test is run so that there is no surprise when they are asked the second time. Even then the nature of the questions can elicit a measurable response.

The best advice I ever got from a polygraph examiner was to never voluntarily take one. Advice given while I was strapped to his machine. Never could be sure if he was being sincere or just manipulating me.

> The best advice I ever got from a polygraph examiner was to never voluntarily take one. Advice given while I was strapped to his machine. Never could be sure if he was being sincere or just manipulating me.

I believe he was being sincere. If you think about it, if the best it can do is simply not provide evidence against you. The worst it can do is make you look guilty, whether you are or not. That's a pretty bad deal for you.

Injustice against the ignorant or less intelligent is still injustice.

No argument here. My point is that bad ideas are still bad ideas (although perhaps my point could have been better-phrased).

Victim-blaming is cool, sometimes? Just not when its a PC issue.

What the actual fuck?

If by "victim blaming", you mean "criticizing someone for taking a stupid action that a reasonable person could assume would end badly".

Don't forget that people who appear on those shows get paid. At least scale wages if not more, plus accommodations, all of which is considerable for people living within ordinary means. Combined with the thrill of being on tv it can b.s. An overwhelming experience, which explains why so many people do it. But much of it cones down to financial and emotional manipulation.

> "criticizing someone for taking a stupid action that a reasonable person could assume would end badly".

Its her fault she got raped, she wore a SHORT SKIRT.

Edit: fuck you.

People like you never seem to distinguish between (a) pointing out what is unreasonable/imprudent, and (b) claiming that a given misfortune was deserved.

It would be comical if you weren't so unpleasant and rude.


The subjectivity of this word amuses me to no end. I can distinguish whatever I want whenever I want, as long as facebook and twitter agree with me.

So let me make sure I have this straight...

You don't believe there's a distinction to be made between "that was a bad idea with predictable consequences" and "he/she deserved [insert misfortune here]"?

I'm saying that the line that separates <"that was a bad idea with predictable consequences"> and <"he/she deserved [insert misfortune here]"> moves whenever the internet twitter/facebook mob decides it does. The distinction is mob mentality. Hence my amusement.

Or just plain stupid.

It seems to me that the intelligence communicate uses lie detectors less as tests for nefarious activities and more as personality tests.

Let's assume that lie detectors actually work. To "pass" such a test you have to be reasonably certain about your answers. Have you ever passed information to the enemy: Yes/No? Is everything in your application truthful? Only some people can be completely secure about there answers. The more knowledge and experience a person has in an area, the more the likelihood that they cannot give that yes/no answer. So what they get is a bunch of young, mostly male, kids who are not the type to self examine. They are sure that they are correct, that they have done no wrong. That infeed might explain much of the direction that the community has taken in recent years.

Or perhaps they're just selecting people who are good at lying.

Since fMRIs are likely the future for this type of thing, this is probably the most disturbing job posting I've ever seen:




Date Posted: 05/03/2012 [...]

[...] seeking contract physiologists to provide technical expertise in central nervous system (CNS) studies related to credibility assessment. [...]

Candidates will provide expertise in the use of CNS technologies such as functional magnetic resonance imaging (fMRI), [...]

A Top Secret security clearance is highly desirable. Candidates without a security clearance who possess superior qualifications may be processed for the required USG security clearance before commencing work.

There are already commercial operations using (or claiming to use) fMRI for lie detection [0]. I can understand the potential appeal of similar positions for applicants — there's a scarcity of academic jobs for even highly skilled cognitive neuroscience PhDs.

As I've posted elsewhere in these comments, the best available evidence suggests that fMRI is no more useful than polygraph for lie detection. Unfortunately, the power of brain images to induce credulity in otherwise intelligent people is well-known [e.g., 1].

[0] http://noliemri.com/

[1] http://link.springer.com/article/10.1007/s12152-007-9003-3

I wouldn't be so quick to discount the technology's potential. While the commercial operation you cited is probably dubious at best, and while it may be true that the current state of the technology is as unreliable as you say for the purposes of credibility assessment, I still believe the future of fMRI research holds incredible potential.

Keep in mind that as far back as 2011, researchers succeeded in reconstructing images from the visual cortex[0].

Moreover, the intelligence community has far more resources and motivation to perfect such a technology. It's possible they may already have advanced well beyond what's currently known today. If not, then they certainly will in the future, provided the technology continues to hold its promise.

[0] http://news.berkeley.edu/2011/09/22/brain-movies/

I wouldn't be in cognitive neuroscience if I thought our methods held no promise, but there's a vast gap between reconstructing images from visual cortex responses and accurately determining whether a person is lying. Farah et al. [0] recently summarized the best available evidence and concluded as follows (emphasis mine):

"... different policies should be considered for different applications of fMRI-based lie detection. We do not join calls to ban fMRI-based lie detection across the board. Despite the enormous shortcomings of the current evidence ... we suggest that restrictions should be proportional to the outcomes and principles at stake. Risk reduction in dating calls for different standards of certainty and different protections of individual rights than the interrogation of terrorist suspects."

[0] http://www.nature.com/nrn/journal/v15/n2/abs/nrn3665.html

My apologies, I didn't check your profile prior to replying and thus was unaware this was actually your field. Good to know. :)

On that note, aside from your assessment almost certainly being far more accurate than my own, I would say yours is largely more desirable from a societal point of view. Even if the technology advances at a most glacial pace, society may still not be properly prepared for the implications upon its arrival. The slower the better, perhaps.

That study is likely closer to BS than you might think. http://www.wired.com/2009/09/fmrisalmon/

True, although I cannot help but feel that your supporting evidence for this claim is somewhat fishy.

If it is true that it's no better than a coin flip, shame on everyone for using it.

And, the idea that a good citizen with a conscience will give a "bad" response easier than a hardened criminal certainly rings true.

Its not used for diagnostics, its used as an interrogation tool. The idea is that you grill someone much easier if they're connected to a strange machine that can seemingly read their thoughts. Decent article about this from 2000 here:


Polygraphs don’t have to work to be a deterrent. People just have to believe that they work and can reveal whether they have committed crimes. The DOE doesn’t have to believe they work, either.

More important, polygraphs are an immensely effective interrogation tool; they need not detect lies. Lykken tells an anecdote of two cops interrogating a suspect at a time when copy machines were not familiar objects. Lacking a lie detector, the cops put a piece of paper in the copier that said “He’s lying!” They made the suspect place his hand on the strange machine while they asked him questions. When they didn’t like his answers, they’d hit a button on the machine. It would groan, whir, stink and shoot out a piece of paper that read “He’s lying!” Realizing that denial was useless, he confessed.

“If I was in the police business I would use [the] polygraph,” says Lykken. “It’s a powerful inducer of confessions, and you don’t have to hit ’em with any clubs. I can’t blame the police for using it; I only blame them for believing it.”

A 1983 report from the Office of Technology Assessment says, “It appears that the NSA [National Security Agency] (and possibly CIA) use the polygraph not to determine deception or truthfulness per se, but as a technique of interrogation to encourage admissions.”

edit: I am not advocating for this. Personally, I think they should be illegal to use.

We've been led to believe that NSA/CIA hire smart people. Smart people probably know things that were common knowledge in 1983, so what is the use of a polygraph administered to NSA/CIA staff thirty years later?

As the article mentions, the polygraph results are still occasionally used in court. Regardless of its utility as an interrogation tool, there are serious concerns about the scientific validity of its output. The use of polygraph output to bar people from jobs and as evidence in court deserves scrutiny.

Is that a good excuse to maintain a pseudo-science complete with "professional organizations" and "journals" and other publications? It pollutes a space that should be for the pursuit of truth.

> Realizing that denial was useless, he confessed.

Is this a win or a loss for society?

I really dislike entrapment based sting operations. The way they trapped him with those two undercover agents and use that to show that he would have aided a criminal stinks... It's not the actions of a just judicial system.

Entrapment is when you're coerced into committing a crime you otherwise wouldn't have committed. For example, if the police show up and say that they're going to arrest your relative unless you rob a certain store, then you rob the store and they arrest you for robbery, that would be entrapment. If they merely pose as a person interested in your criminal activity and get you to commit a crime, which you were happy to commit anyway, that's not entrapment.

The problem here isn't the sting operation, it's that it's a crime to help someone defeat a polygraph in the first place.

According to what I understood defeating the polygraph is not necessarily a crime. What is a crime is to "knowingly help someone lie to a federal agent". If the agents didn't give him any stories about a crime they committed, then he could have claimed that he didn't knowingly help them to lie since there was no lie that he knew about.

By having one of the agent talk about the cocaine smuggling or the other agent talk about sexual abuse, it made the case much stronger because it proved clearly that he was knowingly helping someone to lie to a federal agent since the lie had been made clear by the agents.

The reason I say that it's entrapment is that they've manipulated him into committing a crime by doing this.

One thing I'm not clear about American law is if, in this kind of cases, all records are given of the interactions between the agent and the defendant or not. If not and if the defendant's lawyer doesn't have access to the records, then it's very easy to use certain parts of the recordings to make a much stronger case than would be possible otherwise. Out of context quotes can incriminate people very easily. And in that case, depending on what would be hidden, it's entirely possible that he wouldn't have actually helped someone in a similar situation.

Now this is my interpretation based on what I understand, I'm not a lawyer and all that :-)...

>Entrapment is when you're coerced into committing a crime you otherwise wouldn't have commit.

Coerced is the wrong word, induced is the correct one, and that's what happened here.

I'm not sure I see the difference. I also don't see how it applies here. "Otherwise wouldn't have committed" is key. The guy did this stuff routinely, even advertised it. This isn't a crime he otherwise wouldn't have committed. It just shouldn't be illegal at all.

"Coercion" is specifically the use of threats. "Inducement" is a broader category that includes other methods like begging, badgering, and wheedling.

So offering someone a million dollars to commit a crime would qualify as inducement, but not coercion? Makes sense.

What makes this fascinating to me is , since the undercover agents hadn't actually committed the crimes they told him, he wasn't, in fact, helping them lie to federal agents.

As the purported 'lie' was actually the truth, where is the crime?

Conspiracy. Intent goes a long way, especially in cases like this. Such as cases where FBI makes a fake bomb, ask suspect to flip this switch to arm it, ask suspect to place fake bomb at target, arrested for terrorism.

The whole concept that aiding someone to beat a lie detector is a crime is ludicrous. How many sociopaths beat it without any preparation? How many people with nothing to hide exhibit a stress response at an inopportune moment? I've read accounts of the grillings that people get at some of these interrogations and it's a really shameful practice.

The way this guy was busted though, I don't think I have a problem with it. He knew the stakes he was playing for and clearly knew that a sting was possible yet he went ahead and helped someone who claimed to be a drug trafficker and another, a child rapist. Yeah, technically he was not caught aiding a criminal, but I don't think this was entrapment, where someone got him to commit a crime he wouldn't have otherwise.

If cleverly designed almost anyone could be entrapped for something.

Wilson reminds me somewhat of right-to-die advocate Dr. Kevorkian, in that the passion of his advocacy eventually led him to cross some bright red lines. Kevorkian assisted several chronically-ill patients with their suicides, and juries repeatedly let him off the hook because of the patients' videotaped testimonials expressing their suffering and desire to control how they died. Eventually, though, Jack took it to far and injected a patient himself (instead of waiting for them to push the button), and got sent to prison at age 70.

Maybe the lines weren't quite so bright red with Williams, but that pushing and pushing to the point where you lose sight of the bigger picture seems characteristic of many "solo" advocates (expanding the word "solo" to include not just loners but maybe also founders of non-profits who don't give up the leadership reigns, like Stallman at FSF).

By contrast, really great advocates can look beyond their cause and keep the wider world in perspective. Nelson Mandela is a good example of this (since he focused on uniting post-apartheid South Africa instead of seeking vengeance on the White gentry like his wife and many others wanted to do).

Penn&Teller Bullshit episode on Polygraphs https://www.youtube.com/watch?v=bScv6kfxRyE

Why don't agencies use fMRI instead of something as unreliable and questionable and discredited as polygraphs?

I'm not sure you'd use fmri as widely as polygraphs are today, but maybe for targeted investigation they'd be a useful tool?

There are claims of up to 90% accuracy[1] while not perfect, are better than polys.


Because they are no more or less useful. Here's a link to one of the best headlines Wired ever ran:


Essentially, even a stinking dead fish will show positives in an fMRI. And yes, these studies have really rocked the fMRI field and called into doubt many experiments.

fMRI as a method is not bankrupt or ill-founded. However, as the study you linked shows, it is susceptible to statistical artifacts that must be controlled to support rigorous conclusions.

fMRI is not bankrupt or ill founded, but it shows what it shows, that is more or less the geography of brain oxygen consumption.

The ill-founded and bankrupt part comes from all of weak and tenuous interpretation that comes after that. fMRI can teach us much about brain physiology, but when blood flow is linked to psychology we should exhibit heightened skepticism.

Don't believe the hype. Applying fMRI methods to lie detection is at least as problematic as the polygraph. Here's a good review:


If you’re looking for something that only occurs one-tenth of 1 percent of the time, running a test that’s 90 percent accurate doesn’t help you.

Also known as the false positive paradox. https://en.wikipedia.org/wiki/False_positive_paradox

"a startup called Converus has developed an exam based on eye movement and pupil dilation"

Philip K. Dick has the prior art

They have no evidence the guy actually aided a criminal yet he's facing jail.

Is America the only first world country where such an extraordinary thing is possible?

What do you mean? He knowingly helped two undercover agents lie who disclosed crimes to him. One that supposedly smuggled drugs and another who touched a 14 year old. If he immediately severed contact when they confessed to these crimes he would be free. If you are doing something in a gray area make sure you dont make mistakes.

Here is the witness tampering law: "(c) Whoever corruptly—... (2) otherwise obstructs, influences, or impedes any official proceeding, or attempts to do so,"[1]


I'm guessing a lie detector test is an Official Proceeding.

"He knowingly helped two undercover agents lie who disclosed crimes to him."

What crimes?

Read the article. If you dont understand why then talk to a Lawyer or a Judge.

Or you could just explain what you are talking about.

First Crime: "On Oct. 15, 2012, Williams got a call from a man named Javier Domingo Castillo. He told Williams he was a Department of Homeland Security inspector who’d helped a friend smuggle cocaine into the country."

He helped him beat a lie detector test knowing that he committed a crime. Now I would assume being administering a lie detector test is a "preceding before an executive department."

From wikipedia on witness tampering (this is what he was charged with): "Witness tampering is the act of attempting to alter or prevent the testimony of witnesses within criminal or civil proceedings. Laws regarding witness tampering also apply to proceedings before Congress, executive departments, and administrative agencies."

Police in the US are allowed to lie or use deception to get someone to get a confession.


No. You will be charged with "possession of a controlled substance" in the UK even if the substance was not a controlled substance just so long as they can demonstrate you thought it was. E.g. the cops can set up a "drug deal" to sell you cocaine, sell you talcum powder and then arrest you for buying it, so long as they can demonstrate you think it is cocaine - such as saying "here's that cocaine you wanted".

Same is US, if you sell fake drugs you will be prosecuted as if you sold those drugs.

This seems fair, especially if we want the opposite to be true: if you buy what you legitimately think is talcum powder, and it ends up actually being cocaine, that shouldn't be a crime!

Probably not.

No. A prosecutor must present evidence to a judge that a trial is warranted. If a judge OKs it, then the attorney(s) will have to present that evidence, and possibly more, to a jury. Any prosecutor trying to do this without evidence of any kind wouldn't get anywhere.

Now, I haven't read the article, but you have to careful that such information is present and stated and not ignore. News services all too often, nowadays, leave facts out of stories and, to be honest, I don't view Bloomberg as a source for these types of crime stories.

It seems they busted Williams based on the bogus stories provided by the DHS guy and the deputy sheriff used to set him up. Both "admitted" they broke the law and Williams, knowing this, continued to help them. Had he turned them away could this have turned out differently I wonder?

Personally, I think polygraphs whether used as lie detectors or "personality tests" should be illegal and have no place in a modern society except perhaps in a museum.

I read that the data in the stolen OPM records also contains polygraph data.

I wonder that if, say the US would decide to retaliate and steal similar data from China, they would find records with I-Ching readings? :-)

And soon it appears he'll be a political prisoner. :-(

Applications are open for YC Winter 2020

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact