Hacker News Reader

Comment by jakelazaroff | Hacker News Reader

Relevant excerpt:

> “The undersigned finds that a biometric feature is analogous to the 20 nonverbal, physiological responses elicited during a polygraph test, which are used to determine guilt or innocence, and are considered testimonial.”

So it's analogous to a passcode because you're divulging something stored by your body, as opposed to e.g. a metal key which is artificial.

Tangentially related, it's astounding to me that polygraphs are still considered valid evidence given how widely they've been discredited.

View on HN

I thought that polygraphs were not accepted as evidence in court, but are still used by some government and private agencies.

Also, just curious, have they been discredited as in that double-blind trials show no more than a 50/50 chance of being correct (i.e., flipping a coin)? Or is it that a high enough percentage of people will fail them even if innocent, or pass them even if guilty?

The thing I don't like about them is the method of reading them, where it is up to the subjective judgment of the examiner. That is, there isn't an algorithm that can be written down so that, for example, an untrained person or computer program can read it (and therefore be subjected to large enough random tests).

What does it mean for a polygraph to be "correct?" They're not lie detectors, they are stress detectors. You can't "fail" a polygraph in any reliable sense.

> I thought that polygraphs were not accepted as evidence in court...

Only if you "pass". Fail one and admissible or not, it'll be brought up.

I thought this smelled like totally useless Reddit-style cynical speculation, and 30 seconds of Googling confirmed my suspicion. No state allows admitting polygraph results without both parties' consent. Federal courts may or may not allow them depending on the judge, but if the judge allows them then the defendant is free to submit their passed polygraph (for whatever that's worth) just as equally as the prosecution is free to submit a failed one.

That's my understanding as well.

https://www.hg.org/legal-articles/is-a-polygraph-test-admiss...

> admissible or not, it'll be brought up

That's not how it works, and even a public defender working 20 hour days who doesn't know their defendant's name would be able to get a mistrial.

In the stairwell outside the court room:

"Look, I know your guy went for a polygraph, but you've not brought it up, so I assume he didn't do well. Here's what I can offer..."

The polygraph test is an interrogation technique. The examiner uses the reading to guide the questioning. I think taking a human out of that loop would be as difficult as a Turing test.

> have they been discredited as in that double-blind trials show no more than a 50/50 chance of being correct?

Yes. The reading of polygraphs is very subjective and it's been found that they reflect only whether the person administering the test thinks the subject is guilty.

They also reflect whether or not the subject of the test believes the test works. In practice polygraphs are used as a form of psychological torture, used to frighten people into giving confessions.

Of course whether a polygraph can frighten an innocent person enough to falsely confess is another matter, but if the subject is guilty and understands how polygraphs (don't) work, they can deceive the person administering the test very easily. There is no shortage of spies that have passed polygraphs regularly and reliably with flying colors.

Obligatory scene from The Wire: https://youtu.be/AJ5aIvjNgao?t=115

You can read the National Academy of Sciences' conclusions on the polygraph and lie detection in their paper here - https://www.nap.edu/read/10420/chapter/10

I know this will sound stupid - but its a serious showerthought of a question:

Are "clearers" from Scientology any 'better' at reading a polygraph than non-"clearers"

Auditors in scientology are in fact using a primitive polygraph device to get gullible people to divulge their deepest and darkest secrets. This is something the "church" has used to blackmail and discredit their enemies.

They're notoriously inaccurate, but I don't think that courts have given up on them yet. And if you think that's bad, just wait until you're given an inaccurate brain scan to tell whether you're lying.

https://www.sciencedirect.com/science/article/pii/S016926070...

>The rates of correct detection in guilty and innocent subjects were 86%, which was better than other previously used methods.

Eh, that sounds close enough to 'beyond a reasonable doubt', right?

Eyewitness testimony is also extraordinarily inaccurate, but courts overwhelmingly consider it as strong evidence of guilt.

https://www.scientificamerican.com/article/do-the-eyes-have-...

People win the lottery every day and those odds are very much against thrm, but they play the odds in hopes of a positive outcome. When talking about life and liberty and negative consequences there should be as few chances taken as possible. I don't think 86% is the best we can do.

Individuals can also 'train' for polygraphs, basically learning to spoof the readings. So really you just can't rely on polygraphs for anything at all.

People can also simply lie, but statements are still considered testimony.

Yea the polygraph adds nothing useful here

Clench your anus, a la ‘The Americans’ tv series.

Look up studies on the reliability of witness testimony if you want to get really scared on how courts run.

Even forensic science can be used questionably. Really, once you're in the court room it's entirely rhetoric and fudge-factor.

Agreed.

Pretty much everything in forensic science turns out to be junk science over a long enough timeline. Hair, bite marks, fire progression, all junk. There's even a few cracks in the fingerprint wall. Don't get me started on all the field test kits of various types that LEOs use.

The only things that seem to be reliable are things that don't have their roots in forensic science (e.g. using DNA to identify people) and even then you still have to depend on a crime lab (run by the people doing the prosecuting) to not be sloppy.

And along those lines, I've heard that DNA tests aren't reliable either.

DNA tests prove a very specific narrow fact, and it’s up to lawyers to string together a hypothesis in order for it to have relevance.

Unfortunately all too often the DNA test result is used as proof of a hypothesis, not of the mere presence of surprisingly similar DNA in a swab sample.

The problem with the DNA tests we use today, assuming even perfect cases with zero contamination or mistakes (whole other argument there), is that (in my non-expert understanding):

    * Taking the samples and using several solutions to snip them at given patterns.
    * Taking another solution and promoting replication of the fragments.
    * Using some dye and a weight sorting channel to bin the fragments by weight.

This results in a kind of 'bar code' that graphs the distribution of weights as a VERY crude hash of samples of DNA.

It's pretty useful for determining things like IF it is LIKELY that individuals are related.

It's also pretty good at confirming negatives (We're sure someone is NOT related to X).

It's not that great at confirming positives. That is, the results are both subjective and ambiguous given the hashing. A "positive" result here is really more of an "OK, it's likely we should run the real and expensive check, evaluate if someone might be a suspect by other merits, etc."

I would, offhand, consider a "positive" above to be enough evidence to produce /suspicion/ and /warrants/ to locate other specific evidence to ascertain an actual guilt or innocence based on harder evidence. If an actual "sequence the whole set of samples" option enters the realm of feasible tests then it would also warrant actually doing that.

Come to think of it, I'm not sure how the Ancestry/etc novelty DNA tests work. I'd assume they've isolated a few specific markers they're looking for and the processes are optimized towards identifying those and comparing combinations for those specific traits.

You’ve described electrophoresis DNA fingerprinting correctly, then jumped to the unjustified conclusion that it’s a VERY crude hash.

Not so, it’s a fairly excellent hash, one with few collisions. Better than a fingerprint. If it looks like your DNA, it most likely is.

Source: was an electrophoresis tech (not forensic) in a previous life.

And yes, the usual genetics sites do SNP tests, which are cheaper than full sequencing but mean that they only find what they’re already looking for.

The price of full sequencing is dropping rapidly, I look forward to it being a standard part of medical practice, it will save a lot of lives and improve quality of care.

Better than a fingerprint is still not great!

Well, several decimal orders of magnitude better than a fingerprint, to be accurate.

Apropos of this, but about human incompetency rather than DNA tests, the Phantom of Heilbronn. A serial killer so well concealed that her very existence was only given away by DNA traces.

https://en.wikipedia.org/wiki/Phantom_of_Heilbronn

Your Wikipedia link says that the Phantom is a myth and the DNA was contamination from the kit factory.

Yes it does, good reading skills. That's the point. Human incompetency, related to DNA forensics.

I didn't outright say that, though; the idea was that I would prepare you for a story of a serial killer so well concealed that her existence could only be inferred from trace DNA, but THEN there's the surprise twist - there is no serial killer, it's a trick of human incompetency. BOOM!

The twist has a much greater effect because I didn't give it away at the start, but the twist also ties back in to my original statement about human incompetency - but NOT where you may have been expecting, and subverted expectations are a key part of surprise twists - and the story is also about DNA forensics. That's the previous subject again, the reader having been led around in an interesting and surprising circle.

> and even then you still have to depend on a crime lab (run by the people doing the prosecuting) to not be sloppy.

Which CSI has done a wonderful job of turning into super-cops in the eyes of jurys.

This has had the opposite effect of what you seem to be implying. It's something that's come up quite frequently in discussions of jury nullification [1]. Jury nullification being when a jury ends up voting not guilty, even when they believe the defendant guilty, generally because they do not find the law (or possibly the punishment) just.

For decades a rate of about 5% for hung juries was typical. In more recent years some jurisdictions have seen more than 20% of cases end in hung juries. [2] Some have seen this as evidence of a rise in jury nullification, but others have argued that it's due to the 'CSI Effect' [3] with these individuals believing that, because of shows like CSI, the standard of proof required for a conviction seems to be rising - along with a reduced weighting given to things like circumstantial evidence.

[1] - https://en.wikipedia.org/wiki/Jury_nullification

[2] - https://www.washingtonpost.com/wp-srv/national/jury080299.ht...

[3] - https://en.wikipedia.org/wiki/CSI_effect

That sounds like a good thing to me to be honest

The article about the CSI effect says it's believed to work both ways. Jurors expect more forensic evidence but put more faith in it.

High quality video can be quite useful for getting closer to the truth in many cases. Fits under your tech not developed for forensic science category though.

>crime lab (run by the people doing the prosecuting) to not be sloppy

Oh yeah. Not just incompetence, but actual malice. Recently in Massachusetts there were two high profile state forensic drug lab scandals: One where a chemist was functionally incapacitated by taking all the drugs she should have been testing [0], and one where the chemist was falsifying positive tests (!!!!!!) [1]. More recently, the state police are known to be falsifying overtime records [2]. I can't even begin to imagine what it's like in "stereotyped as corrupt" states.

[0] https://www.cbsnews.com/news/6000-drug-cases-linked-to-rogue...

[1] https://www.cbsnews.com/news/more-than-21000-drug-conviction...

[2] https://www.masslive.com/expo/news/erry-2018/12/1fc00a248688...

While we're at it, most of "ballistic fingerprinting", as well.

My favorite case was where someone was accused of arson because the wood floor was found to contain traces of gasoline, even though none was found on the carpet or padding on top. Turns out that further testing by the defense showed lead also. Which means that the gasoline came from a cheap contractor several decades ago that used (at the time leaded) gasoline as a lacquer thinner for the hardwood floor. And the burn patterns on the floor which indicated an accelerant, were due to wear patterns on the carpet above which allowed more heat to penetrate.

> Really, once you're in the court room it's entirely rhetoric and fudge-factor

I agree with all the skepticism of particular types of evidence in this thread including witness testimony and forensic evidence (I'd also add even confessions to this list). But this rather extreme conclusion doesn't follow. Evidence can be uncertain and hard to interpret, but still lead us to truth. You just need more than one data point.

In a typical criminal trial (assuming the defendant is guilty) the prosecution generally will not hang on a single dubious piece of evidence. They will often assemble a veritable mountain, if they can, because they know that a competent defense attorney will make sure that the jury knows better than anyone the various ways in which evidence can mislead.

This is not to say that everything is rosy in the criminal justice system, and that juries always get it right. There are lots of big problems: Juries are often mysterious and unreliable. Defense attorneys often do a bad job (meaning, among other things, that the prosecution can get sometimes get away with presenting a weak case based on shoddy evidence). And, lest we forget, huge numbers of criminal prosecutions end in a guilty plea, and never go to trial in the first place (sometimes for good reasons, sometimes for bad).

But I want to push back on the extreme notion that the outcome of every trial is all about rhetoric. Sure, it plays a role and some cases are better than others. But I've never seen one that was "entirely rhetoric and fudge-factor." (I'm sure it happens, but not often enough that I've ever seen it.)

"Forensic science" has a long history of being flagrant psuedo-science, very often being a morally bankrupt abuse of statistics.

My favorite example:

>In one case, West claimed to have matched the bite marks in a half-eaten bologna sandwich to the defendant. The jury convicted. (The conviction was overturned on appeal when defense lawyers discovered that the autopsy report recorded a partial bologna sandwich in the stomach of the victim.)

https://www.washingtonpost.com/news/the-watch/wp/2015/02/17/...

There is a lot of sketchy evidentiary techniques that get in, for sure. That said, the vast majority of prosecutions are based on extremely strong evidence. Reading the news gives a misimpression because you only see the edge cases. But for every “making of a murderer” there are ten guys who are convicted based in being caught on security cameras, and credit card reciepts tying them to the scene of the crime, and witness testimony.

And jurors.

This one's my favourite:

The jurors who used a Ouija board to find a murderer guilty - https://www.abc.net.au/news/2018-05-08/ouija-board-juror-mis...

The Drunkards Walk talks about the misuse of statistics in court as well. Not only do they get the math wrong easily, or make common mistakes ("the odds that it's raining given the ground is wet, is the same as the odds that the ground is wet given it's raining"), but if someone accurately states there is only a 1 in 100,000 chance of something happening by chance, the jury will think that means the "something" was not by chance, instead of "oh, so it happens by chance to thousands and thousands of people" like they should.

I once had to go to court as an individual had attempted to stab myself and another (we had tried to stop them from escaping the police).

During the suspect identification process I failed completely; the suspect was bald and had a spider tattoo on their face. I realised that I couldn’t even remember how they looked, it shocked me quite a lot. I assume it was the adrenaline.

> considered valid evidence

They're considered a valid tool for coercion.

Stupid criminals who watch too much TV think they work and some cop comes in and says he failed a lie detector test and they confess.

In that way they actually do work.

Assuming all you want is a confession but if the person actually provides evidence of their own guilt they so be it.

"Tangentially related, it's astounding to me that polygraphs are still considered valid evidence given how widely they've been discredited."

They generally are not considered valid evidence.

(to expand, they are not admissible without consent of both parties, and even then, the court would likely find it not relevant)

I'm not sure how the courts will differentiate between biometric data related to authentication and biometric data from blood tests, breathalyzers or DNA tests. The latter are, in a sense, examples of forcing you to divulge something stored in your body.

The only difference I can see is that in the case of alcohol, biometric data is the thing you are looking for.

In the case of telephones, biometric data is being used to get access to something else.

Implied consent (the authority the police use to try to force you to divulge your blood alcohol content) is a well-established area of law.

The courts can differentiate it by saying that the legislature has written implied consent laws that cover what you are deemed to have consented to when operating a motor vehicle, and that the legislature has not written implied consent laws for cell phones/etc.

Right now it's a pretty easy case for the courts, though higher courts have justified much worse with much less.

It's possible a legislature could write implied consent laws for cell phones, but I don't think they have yet. Not sure if they'd withstand court review, but I am afraid they might.

> Implied consent (the authority the police use to try to force you to divulge your blood alcohol content) is a well-established area of law.

Indeed - warrantless blood tests have been found unconstitutional by SCOTUS three years ago (Birchfield v. North Dakota).

The UK fixed this for drink driving by making it an offence to refuse to submit to the alcohol tests. So they don't force you to implicate yourself for drink driving, you can choose instead to be prosecuted for refusing to provide a sample and have everybody assume (not at all unreasonably) that you refused because you knew you'd fail badly.

Sometimes people are _so_ drunk they refuse to provide a sample without really comprehending that this will itself directly lead to getting arrested and prosecuted but minus any opportunity to get the results thrown out for any reason since it'll just come down to "Officer, did the suspect refuse consent to the required sample?".

You couldn't do that exact workaround in the US because of the Constitution, but I'm sure a creative work around can be invented.

> You couldn't do that exact workaround in the US because of the Constitution, but I'm sure a creative work around can be invented.

The "workaround" in the United States is similar. When you apply for a driver's license, you agree to submit blood/breath/urine upon demand or lose said license automatically.

This is basically how it works in the US too. You are punished for not submitting to the test by revocation of your license and also, the refusal to submit can be used against you to help establish you were DWI.

The article is quite clear about that differentiation:

"the judge wrote that fingerprints and face scans were not the same as “physical evidence” when considered in a context where those body features would be used to unlock a phone."

(Emphasis mine.)

It's the context of using those body features to unlock a phone, that makes them testimonial rather than physical evidence. Outside that context, they can still be physical evidence.

But does that mean the police can't use physical evidence to unlock your phone? I don't know if it's possible to get a fingerprint from a glass and use that to unlock a phone, but if it was, would that still be legal? What about if someone wrote their password on a note?

My understanding is that the police use polygraphs simply as a 'primed' question-asking session for people who agree to take them. It is no different from sitting down and asking someone questions, except maybe they will believe you can tell that they are lying, and so might be more honest. The polygraph doesn't have to be evidence because your answers to questions are evidence by themselves. Whether you are hooked up to a polygraph is irrelevant.

> Whether you are hooked up to a polygraph is irrelevant.

I'm going to go with strongly disagree here.

So you're willing to argue that the output of a polygraph is evidence of something in absence of any responses you give during questioning?

I'm arguing that the idea that anyone is allowed to connect anything to me while they ask me questions is perverse, reprehensible, and absurd.

The comment you replied to pointed out that they are voluntary.

Voluntary is the sense of someone(s) in a position of authority or power asking if you're willing to submit to a voluntary procedure.

I've be interviewed by police twice after being charged with crimes that carry potential prison sentences. The interviews are voluntary in the sense you're not required to answer any of the questions, and I could have chosen to wait in a cell for my lawyer to turn up days later ... but I volunteered to respond to each question by making a mockery of their questions and processes.

These people in authority, they are not there to help you, with their badges and their guns and their attitude problems, and I don't believe it's possible for someone to "volunteer" when such a power imbalance exists.

It's an absurd use of the word. You can't "volunteer" when the other person is carrying a gun.

Absolutely you could have waited for a lawyer. You voluntarily decided not to.

"voluntary" - done, given, or acting of one's own free will. You said yourself that you could have waited, but decided not to.

Alternatively you could have just done a "no comment" interview, which - let's be honest - is what the lawyer would have advised you to do, no matter if you're guilty or not.

That has nothing to do with what I'm talking about. You can (and should) simply refuse to take any polygraphs. It will never work in your favor.

Also "no comment, lawyer please" is a good answer for any question.

In this respect they're not really any different from an officer administering a field sobriety test or a breathalyzer to someone they've stopped for a suspected DUI: these prime the suspect to give more information than legally required.

And worse, these will provide opportunities for the state to build probable cause for a stop that otherwise might not exist.

> So it's analogous to a passcode because you're divulging something stored by your body, as opposed to e.g. a metal key which is artificial.

Also, shouldn't the act of entering someone's phone be analogous to entering someone's house? I.e., wouldn't a search warrant/court order be required?

The article is about the warrant. The warrant was for forcing unspecified non-suspects to be forced to unlock their phones so they could be searched for evidence. The judge judges to be "overbroad".

Reply on news.ycombinator.com