Round-up: A Lot of CSI-Style Forensics Have Turned Out to be Bogus

rubyvroom:

rubyvroom:

cartoonsandcommunism:

rubyvroom:

rubyvroom:

Basically a lot of it is pseudoscience that was never rigorously tested in controlled situations to see if it actually worked.

This is because it was not developed by scientists, but by police, and mainly with an interest in putting people in prison rather than uncovering the truth.

• At least two dozen people have been falsely convicted due to “Bite Mark Analysis”. 
• “Burn pattern analysis” put an innocent man to death in Texas
• “Blood Spatter analysis” such as that shown on the TV show Dexter is actually completely unreliable even according to the US Department of Justice
• Forensic hair comparison is also widely believed to be junk science and the FBI is currently reviewing convictions based on hair analysis due to the unreliability of their results
• Handwriting analysis has an unreasonably high error rate, by some accounts as high as 43%
• Lie detector tests, or polygraphs, are notoriously unreliable and based on bad science. Even though everybody knows this, they are still constantly being used in criminal investigations among other places.
• Toxicology labs can be poorly supervised and badly run, producing false and even fraudulent results
• Due to sloppy procedure at many labs and lack of regulation even DNA testing is often unreliable
• Even when correct results are produced, genetic profiles may be less useful than we have been lead to believe
• Fingerprinting analysis is not foolproof and actually has not been thoroughly tested, as this Frontline special discusses

Here are a few more articles on how unreliable modern forensics are.

Unfortunately due to TV shows that stress forensic investigation, juries are demanding this kind of evidence at trial, and have little idea of how untested and unreliable it really is.

HEY REMEMBER WHEN I WROTE ABOUT THIS TWO YEARS AGO? SPECIFICALLY THE PART ABOUT FBI REVIEWING ITS FORENSICS HAIR ANALYSIS CASES? WELL THE RESULTS ARE IN AND WHOOPS: EVEN THE FBI ADMITS THAT IT’S BOGUS NOW

In case you are stopped by the paywall here’s a Slate article on the same thing and here’s another one. 

Hair analysis alone has been used in thousands of trials. The FBI is reviewing 2500 cases out of “21000 federal and state requests to the FBI’s hair-comparison unit between 1972 and 1999″. Even if this review exonerates some of those convictions, that doesn’t even begin to cover the hundreds of state and local “experts” trained by the FBI in this bogus “hair analysis” technique to do things like this:

Santae Tribble served 28 years for a murder based on FBI testimony about a single strand of hair. He was exonerated in 2012. It was later revealed that one of the hairs presented at trial came from a dog.

So anyway remember anytime you hear about “forensic evidence” that a lot of it is bullcrap and not scientifically validated and a lot of so-called experts are just pulling conclusions out of their ass.

the forensic hair analysis thing is terrible, the FBI literally invented a branch of forensic psuedoscience with no evidence behind it in order to boost conviction rates, then taught the bogus technique to thousands of forensic investigators in the us and around the world. we have no idea how many people have been wrongfully convicted, and this is just one in a very long list of forensic techniques that lack rigorous scientific evaluation

It’s been another year or two so here’s an extremely recent article about how “Criminal Profiling” is totally bogus and TV shows like Mindhunters continue to focus on it because it looks cool and makes good stories, but it really only works in the movies. 

Profiling was trendy in the 70s-90s but has been falling into disrepute ever since. This 2007 analysis showed that Criminal Profilers do not outperform regular detective work. Here’s another analysis finding Profiling unreliable in its current form and suggests ways to make it more scientifically rigorous. Here’s another. 

Hi everybody it’s eleven years later and just for fun I went and looked up the latest news on forensic science and guess what? It’s still bullshit.

They’ve made up entirely new fake techniques since the last time I posted about this. Like “9-11 Call Analysis” which is supposed to identify from a person’s voice on a phone call whether they are lying. Surprise surprise: it doesn’t work.

Despite the seeming pervasiveness of the technique, researchers who have studied 911 calls have not been able to corroborate [the original] claims. A 2020 study from the FBI warned against using 911 call analysis to bring actual cases. A separate FBI study in 2022 said applying 911 analysis may actually increase bias. And academic studies from researchers at Villanova and James Madison universities came to similar conclusions. Ultimately, five studies have not been able to find scientific evidence that 911 call analysis works.

There’s also something called SCAN – Scientific Content Analysis – that examines the statements given to the police for falsehood. This one is actually kind of promising because it uses a real scientific analysis, a grammatical formula, to map out a written personal account kind of like using an AI algorithm to diagram a sentence which can reveal that actually no I’m lying it’s total garbage

Why Are Cops Around the World Using This Outlandish Mind-Reading Tool?

ProPublica also describes a kind of “photo analysis” which is basically whatever a Hollywood movie writer expects to happen when the detective says ENHANCE and of course it doesn’t work either.

So why are there even more forensics techniques that don’t work every few years? If forensics is fake science, why do we keep inventing more of it?

Well, because like I said way too many years ago why have I been here so long, forensics are not created by scientists, they’re created by police. And the justice system is interested in convictions, not uncovering the truth, and these techniques get you convictions. Many of them are wrongful convictions, but oh well.

Anyway in conclusion, John Oliver agrees with me and ACAB