How an ex-FBI profiler helped put an innocent man behind bars
latimes.com
latimes.com
Edit: it's an ironic coincidence that this follow-up feature was published on the same day of the news of O.J. Simpson's parole: http://www.latimes.com/nation/la-na-oj-simpson-parole-board-...
Simpson is being paroled for a robbery conviction 9 years ago but of course his most famous run-in with the law was when he was acquitted of double-murder despite a huge amount of forensic evidence linking him to the homicides (he later lost a civil suit, which has a lower standard of evidence).
In the case described by the OP, the jury voted to convict in the total absence of physical, forensic evidence. It's astonishing that this profiler's beliefs were enough to sway the jury beyond a reasonable doubt but even more grotesque is how a clusterfuck of dumb, unverified assumptions led the profiler to his conclusions. He blames the detectives for not doing "Investigation 101", and says he would not have testified if only he had known the detectives were so negligent.
As much blame as the profiler deserves (he basically comes off as reliable as a psychic), seems like a large portion of blame should go to the district attorney, who presumably knew he was so short of evidence that he had to base his case on the profiler's testimony. Can't imagine what was going through the jury's heads; prosecutors often complain how modern juries demand the presence of DNA because of how TV dramas like CSI portray and hype DNA and other evidence as ubiquitous and unimpeachable, but this jury seems to have had the completely opposite mindset.
https://blog.simplejustice.us/2017/05/08/short-take-magic-co...
tl;dr: cop goes for a 160h course, gets a Drug Recognition Expert certificate, then his opinion "based on his training" is enough to establish probable cause and make an arrest and then you sit in jail for however long it takes for tests to get back (if that even helps because "tests come back wrong all the time")
There is plenty of blame to go around for the cascade of errors that led to Ray Jennings spending 11 years behind bars for a murder he didn't commit. But most of the institutions responsible for the tragedy — from the D.A.'s office to the judiciary — have owned up to their mistakes.
Not so for Mark Safarik. Despite retracting his opinion in court, he is now publicly insinuating that Ray may actually have been guilty, repeating the very falsities that my father and I debunked in order to overturn the false conviction.
Everything that Safarik says in the article — from the supposed evidence of sexual assault, to the fact that Ray should have been able to see the killer — is addressed and refuted in our initial 34-page letter to the Conviction Review Unit. The truth is available for anyone who cares to read it: http://www.ehrlichfirm.com/jennings/Ray-Jennings-Letter-to-C...
I note that, before we even hired a profiling expert, I was able to point out gross errors in Safarik's work merely by contrasting his testimony to the standards listed in the behavioral-profiling "bible" that he contributed to at the FBI. (See Letter, pp. 20-22.) This is supposed to be the profiling equivalent of the DSM, but Safarik blew off the numerous criteria that contravened his thesis.
Safarik's attempt to shift all the blame to the local detectives is unpersuasive. Their failure to interview everyone at the crime scene was obvious from the investigative materials that Safarik was provided by the Sheriff's Department's. Indeed, Safarik was specifically cross-examined about that failure when he took the stand, yet he clung to his opinion, and thereby allowed an innocent man to be convicted of murder.
Only Safarik himself is responsible for basing his analysis on the false premise that Ray was the only person present who could have committed the crime. When he presented his "expert" opinion, Safarik acknowledged the presence of other individuals in the parking lot around the time of the murder, but inexplicably brushed that fact aside because it did not fit his desired conclusion.
If Safarik is the best that the "science" of profiling can offer, it has no place in our courts.
I can imagine that profiling might have a use in generating leads for investigators to follow, perhaps leading them to actual evidence. But it can't be considered evidence itself.
>The driver’s-side window of the victim’s car had been lowered several inches, suggesting to the profiler that the teen had rolled it down when someone who looked trustworthy approached. And her tube top was askew — a sign, the profiler said, of a botched sexual assault.
In Sherlock Holmes stories, that is a fairly reasonable inference. Here, for example, is an exert from the story "The Hound of the Baskervilles":
>The address, you observe is printed in rough characters. But the Times is a paper which is seldom found in any hands but those of the highly educated. We may take it, therefore, that the letter was composed by an educated man who wished to pose as an uneducated one, and his effort to conceal his own writing suggests that that writing might be known, or come to be known, by you.
The issue is in real life, this sort of inference is absolutely absurd. There are a dozen other possible explanations that fit the evidence, and to make such wild speculations gets you nowhere except chasing dead ends and accusing innocent men. It works in Holmes' stories because, and this is very important, they are fictitious. Doyle can have Sherlock spout any kind of insane gibberish, and still have him be right in the end.
Though I can't fault lay persons for not understanding why this technique is faulty, both because pop culture treats it as valid and because it's a classic instance of sampling bias: Say 100 profilers failed spectacularly, and 10 succeed and produce perfect profiles. The 100 failures aren't interesting or worth talking about, some stranger not being correct one time is not a good story. The 10 successes get talked about endlessly at conferences, at the local bar, when people give advice to each other. Old veteran cops 20 years later still talk about that one time a profiler solved a cold case in a snap. So, over years, a mythology is built up and the objective abysmal quality of profiling is hidden by the fact everyone everywhere has heard of the 10 successes while the 100 failures are lost to the dustbin of history. Rather, appropriately, like Doyle's own real life attempts to play detective - the dozens of times he was humiliated are (mostly) forgotten, while the handful of successes are the stuff of (nerdy) legend.
-- Terry Pratchett, Feet of Clay
The books would have caused a thousands kinds of outrage had they been written today.
In the new system, professional jurists would be paid for by the courts, but not directly employed by them or the government. The prosecution and defense could have their pick that they must agree on just like now, but they would be basing their pick on the professional reputation of the jurist, not exploitable characteristics or ignorance.
Jury selection today specifically selects against "peers" of the accused. (and as you also mentioned, against "peers" of anything in general :-P )
Peers in context have been interpreted to mean your fellow citizens, nothing more.
Although I'd like to note that letting a jury of bankers sit on a financial crime might mean no banker ever got convicted. So the whole more exacting definition of peers argument has some definite problems.
No, it absolutely does not, because neither the phrase nor any alternative phrase on the subject incorporating a reference to “peers” appears in the Constitution, which instead provides (in the Sixth Amendment) for an “impartial jury of the State and district wherein the crime shall have been committed, which district shall have been previously ascertained by law”. (It is a common rephrasing of a guarantee—“the lawful judgment of his peers”—in Magna Carta which unmistakably was an product of the barons’ insistence that members of their formal social station should render verdicts against them.)
You have heard wrong. The term is a rephrasing of a guarantee in Magna Carta [0], and refers to the right to be judged by others of the same formal social station.
[0] in one English translation, “No free man shall be seized or imprisoned, or stripped of his rights or possessions, or outlawed or exiled, or deprived of his standing in any way, nor will we proceed with force against him, or send others to do so, except by the lawful judgment of his equals or by the law of the land.”
Jurist are as impartial as you can get. Making someone a paid professional jurist doesn't magically strip them of all their biases. Courts are intimately familiar with this, precisely when they ask jurists "Do you think Police are less likely to lie under oath than a normal person" in voir dire because judges know that police are just as corruptible as the rest of us.
Your proposal is basically creating a system by which a lawyer cannot ensure that your are judged by a jury of your peers. The problems with the judicial system are people based /not/ jury based.
Jury duty is the one chance that citizens can make a direct action in the power of the state and you are proposing changing that and some how having the government to pay that person but also simultaneously not being paid by the government.
I am proposing that the expense of the jury be covered by the court (as it is now for citizen juries, and it is also is with public defenders), but that they not be employees of the court. Again, like public defenders.
It certainly wouldn't be a perfect system, but I believe it plugs some of the holes that have appeared in our current system of jury selection.
There are reliable tells that imply "pacifying behaviours" for stress, and those can be useful to pay attention to in order to get people to trip up in interrogations, but there are many possible reasons for that kind of stress, including an interrogation itself being a stressor. But people who try to identify lies based on their theories of such tells without intimately knowing that person in advance generally do no better than chance.
(A good book on the subject, which gives a very sober assessment of what body language can actually reliably tell us vs. where you need to thread carefully is Joe Navarro's "What Every Body is Saying" - Navarro is a former FBI agent and trainer)
We like to think of courts and juries in the idealistic view given to us in school. Its just like "12 Angry Men". Everyone gets their fair shake in court. Even if 11 people want to convict and go home the 12th one will stand up right? After all, you're a reasonable person. You can't imagine just convicting a guy without a reasonable discussion. Surely enough of your fellow citizens are the same?
The truth is a shocking amount of people in society just have deference to authority figures. They'll trust whatever expert the prosecution produces. You would think that courts have standards for what counts as experts, but the truth is the standards are horribly lacking. Shows like CSI give the average citizen a cops and robbers view of reality. When is the last time an episode of CSI didn't end with the guy getting caught and convicted? Have they ever made a mistake in CSI? Made an ambiguous decision?
And why wouldn't the DA choose to prosecute as much as possible? You don't want to appear weak on crime. After all, you have an election to win if you want to keep your job.
Trial by jury is only as good as the society within which it exists.
Does it have a beyond a reasonable doubt standard, because if it does this implies a validation of the 28% crazification of the population theory.
Actually there was one episode where they knowingly sent down an innocent man in CSI:Miami. A mentally challenged man had been unwittingly framed by his brother for a murder he didn't commit. The piece of evidence that sends him to jail if I recall correctly was a tie used in the crime that the brother asked his mentally challenged brother to retrieve from the back of the truck putting his DNA or fingerprint on it. The team know the kid they're sending down is innocent, but the evidence is the evidence and their hands are tied.
I can't for the life of me find a link to the episode. Perhaps someone with more Google-fu than me could assist.
Of 28 examiners with the FBI Laboratory’s microscopic hair comparison unit, 26 overstated forensic matches in ways that favored prosecutors ... The cases include those of 32 defendants sentenced to death. Of those, 14 have been executed
In fact, in the linked article it says that DNA exonerated people tried based on hair comparison – 25% of those exonerated to boot. Sounds like DNA is great.
What you seem to have a problem with, which is valid, is procedures, how evidence is introduced, and how the authority/validity of that evidence is communicated to the court, correct me if I'm wrong.
forensic toxicology and chemistry also work fine (they use the same tools and techniques that analytical chemists use for assaying unknown substances, or your doctor's office uses for measuring how much of your pain killer is in your system).
the problems from those labs come when the government-run ones decide they're on team blue, instead of team truth, and start making shit up.
> “Like other killers I’ve known, he’s also arrogant and narcissistic — fatal traits that led to his demise,” Safarik says. “This was all his doing…. Ultimately, he was responsible for it.”
I only see arrogance and narcissism coming from one person in the story told, and it wasn't from Jennings.
If I had made a mistake - even an honest mistake, the kind where one could not have done any better - that caused somebody to be falsely convicted - not even of murder: any kind of crime that is serious enough to end up in court! - I strongly think that I would feel absolutely terrible about it and at the very least apologize to that person. Beg for forgiveness, actually.
[1] http://www.sfgate.com/news/article/Mixed-Success-for-FBI-Age...
But I am curious: apart from TV shows, how important is this kind of evidence in most trials? Is it actually uncommon?
Proceedings of the National Academy of Sciences (PNAS)
Bradford T. Ulery, 7733–7738, doi: 10.1073/pnas.1018707108
Accuracy and reliability of forensic latent fingerprint decisions
Bradford T. Ulerya, R. Austin Hicklina, JoAnn Buscagliab,1, and Maria Antonia Robertsc
Edited by Stephen E. Fienberg, Carnegie Mellon University, Pittsburgh, PA, and approved March 31, 2011 (received for review December 16, 2010)
ABSTRACT
The interpretation of forensic fingerprint evidence relies on the expertise of latent print examiners. The National Research Council of the National Academies and the legal and forensic sciences communities have called for research to measure the accuracy and reliability of latent print examiners’ decisions, a challenging and complex problem in need of systematic analysis. Our research is focused on the development of empirical approaches to studying this problem. Here, we report on the first large-scale study of the accuracy and reliability of latent print examiners’ decisions, in which 169 latent print examiners each compared approximately 100 pairs of latent and exemplar fingerprints from a pool of 744 pairs. The fingerprints were selected to include a range of attributes and quality encountered in forensic casework, and to be comparable to searches of an automated fingerprint identification system containing more than 58 million subjects. This study evaluated examiners on key decision points in the fingerprint examination process; procedures used operationally include additional safeguards designed to minimize errors. Five examiners made false positive errors for an overall false positive rate of 0.1%. Eighty-five percent of examiners made at least one false negative error for an overall false negative rate of 7.5%. Independent examination of the same comparisons by different participants (analogous to blind verification) was found to detect all false positive errors and the majority of false negative errors in this study. Examiners frequently differed on whether fingerprints were suitable for reaching a conclusion.
http://www.pnas.org/content/108/19/7733.full
Authors
Bradford T. Ulery
aNoblis, 3150 Fairview Park Drive, Falls Church, VA 22042;
R. Austin Hicklin aNoblis, 3150 Fairview Park Drive, Falls Church, VA 22042;
JoAnn Buscaglia bCounterterrorism and Forensic Science Research Unit, Federal Bureau of Investigation Laboratory Division, 2501 Investigation Parkway, Quantico, VA 22135; and
Maria Antonia Roberts cLatent Print Support Unit, Federal Bureau of Investigation Laboratory Division, 2501 Investigation Parkway, Quantico, VA 22135
One might wonder why such an assessment was not done a long time ago.
This is a report on the Brandon Mayfield case:
https://oig.justice.gov/special/s0601/exec.pdf
Bottom line is fingerprints are much more accurate than chicken entrails but hardly infallible as used to be widely believed.
In addition, the 0.1 percent false positive rate is based on a small sample of around 1000 cases. The Federal fingerprint databases such as the ones used in the Brandon Mayfield case have millions of people in them and may eventually have all US citizens (over 300 million people) in them. How does this "small" rate extrapolate when a fingerprint is compared to every fingerprint in the US or the world?
> I could understand things like trying to guess based on similarities with a partial print
That's really the rub, pun intended I guess, there are no clean neatly pressed fingerprints left on object collected by police, generally speaking.
[1]http://www.sciencedirect.com/science/article/pii/S2211368113...
[2]http://www.cbsnews.com/news/fingerprints-infallible-evidence...
>The big crime is how little epistemological support there is for some of the big forensic tools (fingerprints, DNA, bite marks, arson spread, etc), how little interest there is in researching these areas, and how trusted they are. But I am curious: apart from TV shows, how important is this kind of evidence in most trials? Is it actually uncommon?
Second, have we programmed a computer to do it? At this point, something like "fingerprint analysis" should be quite open to machine learning. If we can't program a machine to do it, then probably humans aren't actually accurate either.
I think juries should be clearly educated on the limitations of DNA testing, and other kinds of forensic evidence should be banned. Then again, I recognize that for every person unjustly convicted based on lab error, there are a dozen who just want to throw that possibility out there hoping to confuse the jury enough to get an acquittal.
[1] Courtrooms are open to the public, and I'd encourage everyone to go to your local federal courthouse and sit in on a dozen sentencing proceedings.
I think it's fairly widely used as part of the evidence, but I don't think statistics are kept, and even if they were it's hard to know how important it is in convincing juries. When the issue with fiber analysis and the admission of it's nesr complete lack of validity by the FBI broke, there were hundreds of criminal convictions, dozens of death penalty convictions, and several who had already received the death penalty identified directly tied to FBI testimony on fiber analysis, and concerns of tens of thousands of convictions based on work by locsl, FBI-trained, fiber analysts.
So, its at least not insignificant in potential impact.
An introduction with links (caveat Frontline): http://www.pbs.org/wgbh/frontline/article/forensic-tools-wha...
Strengthening Forensic Science in the United States: A Path Forward (2009), Summary: https://www.nap.edu/read/12589/chapter/2 This https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4581003/ puts it in more recent context.
> arson spread
For example: "Progress" In Fire Investigation: Moving from Witchcraft and Folklore to the Misuse of Models and the Abuse of Science[2](PDF; not as good as the title; depressing) from [3].
> how little interest there is
One challenge is the forensic expert "system" is extremely heterogeneous. In funding, honesty and integrity, professionalism, research, oversight and regulation.
> how important is this kind of evidence in most trials
And since most "trials" are plea bargains, how important in pleas?
And the US plea bargain system is its own pit of systemic dysfunction.
[2] http://www.firescientist.com/Documents/PROGRESSIN%20FIRE%20I... [3] http://www.firescientist.com/publications.php
I feel like some sort punishment is due towards ruthless opportunistic people like these. (I would say the same about the people who pursued Swartz even after MIT & JSTOR dropped the case.)
How could you possibly prosecute people for the Swartz case? They were simply seeking the punishment that the law specifies. If you don't like the law, then lobby to have it changed. If you want it to be less vigorously enforced, then vote for a President/Governor/DA that will enforce it less vigorously. You can't put people in jail for following the law.
When wrongly convicted inmates are freed from prison, I think it reasonable to remunerate them per year of jail time, with accelerating penalties (as more years are lost to jail, opportunity is taken away at an accelerating rate; 20 years lost to prison is far more than 10 times worse than 2 years lost to prison, for example).
The base rate of $550,000 tax-free per year for lost work/family/life opportunity and pain/suffering seems eminently fair to me, with escalating amounts per year, as stated before. Probably could get behind a permanent waiver of personal income tax for life on any income under $200,000/year as well.
no history of gang crime, "didn't have a boyfriend or a criminal record", "wallet left", "Mustang wasn’t taken", "parking lot was lighted and patrolled", Her tube top was pulled down exposing her breasts." -> "sexual assault”. "wipes his hand over his brow" -> guilty.
Imagine Safarik was a Machine Learning algorithm. More than a decade's worth of training data led to a model that predicted Jennings to be guilty.
The counter-evidence (phone missing, no scratch marks) wasn't part of the model.
In his defence, Safarik denies "assigning too much weight to [evidence]. It was the totality of things that shaped his finding."
Likewise, it is difficult to fix errors in a Machine Learning model once it is trained. The "totality of things" is the history of other crimes the model has investigated.
Can Deep Learning do the same job as criminal profilers? Are they as accurate? Should their predictions be trusted, when the consequence of failure is 11 years of prison for an innocent man?