Brendan Dassey appears at the Manitowoc County Courthouse in Manitowoc, Wis., in April 2007. (Dan Powers/The Post-Crescent/Pool)

Over at the New York Times, Adam Liptak writes about the possibility of the U.S. Supreme Court taking up the case of Brendan Dassey, the 16-year-old whose interrogation and confession were included in the Netflix series “Making a Murderer.” Dassey’s attorneys argue that his confession should have been suppressed, citing the way both Dassey and his mother were manipulated and basically bullied by interrogators, as well as emerging evidence that juveniles are especially prone to false confessions, especially under harsh questioning.

The Wisconsin Supreme Court disagreed and upheld Dassey’s conviction. But a federal district court judge and a panel from the U.S. Court of Appeals for the 7th Circuit found that the confession should have been suppressed. The full 7th Circuit then reversed again. That’s where the case stands today.

The U.S. Supreme Court could simply decide not to hear the case, in which case Dassey’s conviction would stand. They could also uphold the conviction on the grounds that the 7th Circuit did, finding that his claims aren’t compelling enough to get him past the 1996 law (commonly known as AEDPA) that limits when federal courts can review state court rulings. In other words, the court may not delve much into the issue of false confessions among juveniles, and find instead that this is an issue that, under that 1996 law, should be left to state courts to decide.

In our new book, my co-author and I look at a case that I think provides a pretty vivid illustration of why this shouldn’t be an issue left to the state courts. Tyler Edmonds was just 13 when he was accused of killing his sister’s husband. According to prosecutors, Edmonds and his sister waited until the victim was asleep, then the two both held a gun and simultaneously pulled the trigger, killing the man. Edmonds would later say that his sister confessed to him, but urged him to take the blame, and claim that he accidentally fired the gun while playing with it. Since he was a juvenile, she said, they’d go easy on him. If she was convicted, she told him they’d send her to the electric chair.

During his initial questioning, Edmonds denied any role in the murder. But he was later separated from his mother. At one point, he was put in a room with his sister, who slipped him a note again urging him to confess, or she’d be given the death penalty. After that, Edmonds confessed. But his confession wasn’t consistent with the crime scene, or with other facts known to police. And he confessed in a way that didn’t completely take full responsibility for the crime — it was Edmonds who first laid out the two-people-holding-the-gun theory. Three days later, after talking to his mother and other relatives, Edmonds retracted the entire confession, explaining that he thought if he took partial responsibility, he’d save his sister from execution.

As it turns out, Edmonds’s confession raised a number of red flags. He wasn’t just a minor; he was just 13. He was questioned outside the presence of a parent. He was emotionally and psychologically immature for his age. He also adored his sister, making him more susceptible to her manipulation. Edmonds also confessed in a way that minimized his role in the crime. His confession wasn’t consistent with the facts of the case. And he quickly and completely retracted. All of these are common indicators in a false confession. The defense was prepared to call an expert witness to testify to all of this. But prosecutors objected, and asked for a hearing (known as a Daubert hearing) on the scientific validity of expert testimony about false confessions. The trial judge granted a daylong hearing on the question before ultimately deciding that the research in support of the idea that juveniles can be pressured into false confessions wasn’t scientifically reliable. Edmonds’s expert would not be permitted to testify.

The judge first did what judges often do in these situations. He looked at what other judges had done. He found that about half the courts that ruled on the matter had ruled against allowing such testimony in. Thus, he concluded, the notion of false confessions had yet to be accepted in the scientific community. But how judges have ruled on a challenge to expert testimony is quite a bit different than what science says about the matter. And in fact, more than a decade before Edmonds’s trial, a review of peer-reviewed research on the matter found more than 800 published reports in support of the idea that false confessions do indeed occur. And as former Mississippi Supreme Court justice Oliver Diaz would later point out in a concurring opinion, while the trial judge claimed that only six people were researching the issue at the time, in fact more than 60 authors had published on the topic. In support of his opinion to exclude the testimony, the judge cited just three experts. One of them had never published a peer-reviewed study on the topic. Another was former federal judge Paul Cassell, who is not a psychiatrist or psychiatric researcher, but a partisan legal pundit and longtime victims’ rights advocate.

If that all wasn’t bad enough, the state also put on its own scientifically dubious evidence. As the prosecutor told jurors in her opening statement, “You’re going to hear how Kristi stood behind him and held him and you’re going to hear how they both put their finger on the trigger and you’re going to hear how they both shot and killed Joey Fulgham.”

That’s an extraordinary claim that would be nearly impossible to prove. Yet the state tried, with the controversial, longtime Mississippi medical examiner Steven Hayne. When asked whether the wounds in Fulgham’s body were consistent with Edmonds’s confession (as opposed to his retraction), Hayne replied, “Within a reasonable medical certainty, it’s consistent with the scenario provided to me.” That, of course, would be the scenario in which two people fired the gun. Asked later to clarify — was he really claiming that the bullet holes suggested there were two people holding the gun that fired the bullets? — Hayne replied, “I would favor that a second party be involved in that positioning of the weapon.”

This, of course, is nonsense. There’s simply no way a medical examiner can look at bullet holes and know how many people were holding the gun that fired those bullets. Edmonds’s attorneys rightly objected and requested a Daubert hearing of their own. Incredibly, the trial judge — the same judge that granted a daylong hearing after which he ultimately refused to allow in expert testimony about false confessions — wouldn’t even grant a hearing on the scientific validity of the two-people-holding-the-gun theory. So that’s what the jury heard. And it convicted Edmonds.

Edmonds appealed and lost at the Mississippi Court of Appeals. He then appealed to the Mississippi Supreme Court. He lost there too, at least at first. The justices initially voted to uphold his conviction. But after Diaz wrote a blistering dissent excoriating his fellow justices, they voted again and came down 6-3 in favor of overturning the conviction because of Hayne’s testimony. That in itself was noteworthy — it was the first time in two decades of cases that they had tossed out the testimony of the now-discredited medical examiner. But the majority also ruled that the trial judge was right to exclude the false confession testimony, finding that it “couldn’t be empirically tested.” (Never mind that courts routinely allow all sorts of expert testimony that can’t be “empirically tested,” including most fields of forensics.)

Edmonds was tried again. Without the dubious forensic testimony, he was quickly acquitted. Remarkably, the state later accepted his false confession argument in an attempt to prevent Edmonds from being compensated for his wrongful conviction. The state argued that because his confession was “manufactured evidence,” he had aided in his own conviction. Under state law, he was therefore ineligible for compensation.

According to the defense attorneys I’ve talked to in Mississippi, the portion of the Mississippi Supreme Court’s Edmonds decision dealing with false confessions is the most recent such decision from that court. This means that any defendant who wants to introduce expert testimony on false confessions would face an uphill battle, given the court’s line about about empirical testing.

AEDPA was a bad law for a lot of reasons. But its limitations on federal review of state court decisions might be the biggest one.