Two days after Judge Steven Jones allowed a Georgia Tech sexual assault judgment to stand—even as the judge admitted that conduct from the university investigator “very far from an ideal representation of due process”—the university will be back in court, facing another lawsuit calling into question its fairness.

Just as in the first case, an internal appeals process overturned a decision by Georgia Tech’s “single investigator,” Peter Paquette. Just as in the first case, the student wound up being expelled anyway. Just as in the first case, the accused student is noting the curricular impact of the expulsion—requesting an injunction since Georgia Tech is in the process of changing its required courses, so if he can’t enroll this spring, he’d be permanently harmed as classes vanish from the curriculum. If anything, the facts of this case are even more compelling. (You can read the complaint here.) Fortunately for the accused student, the case was assigned to Judge Richard Story, rather than Judge Jones, the jurist indifferent to Georgia Tech’s unusual conception of due process in Wednesday’s ruling.

The case involves two students, one gay and one bisexual (John Doe, or JD, the plaintiff), who aren’t identified in the complaint. After sexual activity in April 2014, the accuser waited just under a year to file charges with Georgia Tech, even though he had regular social contact with JD during most of that time. As occurs in the overwhelming majority of campus cases, there’s no evidence the accuser went to the police.

The two students hooked up twice in April 2014; but the accused student soon made it clear he wasn’t interested in a relationship. Nonetheless, shortly after their second hook-up, the eventual accuser rented a nearby duplex for the following academic year. He thus, as the complaint notes, “arranged to rent a house that was literally fifteen feet away from the person who he would accuse almost a year later of sexual misconduct.” Photos from fall 2014 show the two socializing together, even though they no longer had any type of sexual relationship. In early February 2015, the would-be accuser sent a message asking the plaintiff to come “cuddle with him”; the plaintiff refused and said he had no interest in a sexual relationship.

That message appears to have marked a turning point in the two students’ relationship. The accuser shortly thereafter asked JD to delete their private Facebook correspondence; in what turns out to have been a major mistake, JD complied with the request. By the end of the month, the accuser unfriended him on Facebook and ceased all contact with him. He filed a complaint in April 2015. JD, who hadn’t been open about his bisexuality, was forced to reveal his sexual orientation to his parents as a result.

Georgia Tech investigated. The school—whose policies, as was seen in the previous case, designate a sexual assault accuser as the “victim,” and witnesses as “informants,” thereby belying the presumption of innocence—uses the “single investigator” model, with Paquette running the show. As shown in the previous case, Paquette’s investigative technique is unusual. He interviews students, but doesn’t record the interviews, and shares with the accused student only summaries of his private interviews with his “informants.” (An accused student has no right to sit in, or have a legal representative sit in, on any of Paquette’s interviews.)

In this case, according to the complaint, the accuser presented Paquette with edited version of Facebook conversations he’d had with the student he accused; Paquette (like all college investigators) lacked subpoena power to compel production of the entire exchange. Nonetheless, after speaking to JD, the investigator didn’t sound too sure that anything untoward had occurred. He produced a report conceding that “it is reasonable to believe that based on the nonverbal action of [the accuser] that John Doe believed he had consent.” Case over, right?

Guess again. In a bizarre interpretation of the presumption of innocence, the investigator argued, “However, the charge of this investigator, however [sic], is to determine if one of the stories is more likely than not.” In other words: Paquette seems to view his job not as determining whether it was more likely than not the accused student reasonably believed he had consent. Instead, he utilizes some vague standard of attempting to determine which person’s story the investigator found most likely. So if Paquette was 99 percent certain an accused student believed he had consent, that student theoretically could still be deemed a rapist by Georgia Tech if Paquette was 99.5 percent convinced the accuser had too much to drink.

How did Paquette reach his conclusion in this case? He determined—based on 12-month-after-the-fact conclusions—that the accuser was “very intoxicated.” (The second hookup occurred at a party; both students had been drinking.) According to Paquette, being very intoxicated made it “more likely than not” that the accuser was “incapacitated and incapable of giving consent.” In other words: it was more likely than not the accused was very intoxicated, and it’s more likely than not (according to Paquette’s wholly subjective standards) that a very intoxicated person will be incapacitated.

No wonder Judge Jones concluded that Paquette’s vision fell “very far from an ideal representation of due process.”

JD appealed. As occurred in the previous Georgia Tech case, the university appeals committee looked skeptically at Paquette’s work (especially his odd determinations of credibility), and overturned the finding. The accuser had five business days to file an appeal to the president; he didn’t do so. Case over, right?

Guess again. On the sixth business day, an appeal was filed—but not by the accuser. Instead, his parents filed the appeal. Rather than dismiss this as procedurally improper, the president heard the appeal, and restored Paquette’s finding. JD then exercised his final right of appeal, to the Board of Regents. Remarkably, they overturned the president’s actions (it seems very likely that the board was also aware of the dubious conduct in the other Georgia Tech case as well). Case over, right?

Guess again. The Georgia Tech president remanded the case to the appeals committee, but gave them a new charge, reiterating an extreme affirmative consent policy (“the initiator must obtain consent at every stage of sexual interaction”) and holding that “intoxication,” in and of itself, is enough to have “incapacitation,” and thus sexual assault. Under this extraordinary definition—which would make a good chunk of the sex between Georgia Tech students to be rape—the appeals committee reversed its original findings, and upheld Paquette’s judgment.

The juxtaposition between the two Georgia Tech cases is remarkable. In the first case, Paquette proclaimed that once a key witness lied to him, that decision alone undermined the witness’ credibility. So because the accused student in the first case lied to Paquette about sexual contact with another female student (regarding which Georgia Tech said he did nothing wrong), the investigator no longer deemed him credible regarding the allegation that led to his expulsion. But here, according to the complaint, the accuser made several false or misleading statement to Paquette—with no loss of credibility.

Will Judge Story show the courage that Judge Jones did not?

While those enraptured by the narrative seem not to care (CNN essentially turned over its editorial standards to the producers), the last few weeks haven’t been good for The Hunting Ground. Nineteen Harvard Law professors criticized the film’s accuracy. E-mails from one of the film’s producers, Amy Herdy, revealing that “we don’t operate the same way as journalists—this is a film project very much in the corner of advocacy for victims, so there would be no insensitive questions or the need to get the perpetrator’s side.” Stuart Taylor sharply criticized the film in National Review. An NPR piece by David Folkenflik highlighted a study in JAMA Pediatrics that contradicts the film’s reliance on the controversial researcher David Lisak. And lawyers for one of the film’s targets, Brandon Winston, put together a document-filled website containing trial transcripts that the film’s producers either ignored or distorted to make their case.

It’s a striking commentary on the two sides’ credibility that the accused student has done everything he could to have people read the trial transcript, and the “documentary filmmakers” seem utterly disinterested in the factual record. That transcript, by the way, vindicated the first major criticism of the “documentary,” from Emily Yoffe.

Throughout this process, director Kirby Dick and producer Amy Ziering haven’t taken criticism particularly well, and they’ve held true to that pattern. Last week brought a veiled threat, as the duo informed the Crimson, “The Harvard Law professors’ letter is irresponsible and raises an important question about whether the very public bias these professors have shown in favor of an assailant contributes to a hostile climate at Harvard Law.” In other words: criticize us and you might face a Title IX complaint with OCR. As HLS professor Janet Halley recognized, “What it really means is that they don’t want the debate.”

At the very least, it might be presumed that “documentary” filmmakers making veiled threat against professors’ speech would outrage those professors’ colleagues. But one (non-tenure track) member of the Harvard Law faculty has consistently sided with Dick and Ziering in their contempt for campus due process. In an earlier piece on Inside Higher Ed, Diane Rosenfeld was quoted as saying, “I fully support the Hunting Ground film, which is all about ending the silencing of survivors. I am bound by the principles of confidentiality under which the hearing was conducted, so I cannot say anything about the substance of the case. I can however say that the signatories of the press release represent only a minority of the [Harvard Law School] faculty.” It’s true that 19 HLS professors signed the open letter. It’s also true that that the only HLS instructor who appears to have defended The Hunting Ground is Rosenfeld. (One is, obviously, a much smaller minority than nineteen.) Dick and Zeiring nonetheless raced to put the Rosenfeld statement on their site.

Rosenfeld’s argument seems to be that the initial stage of the Harvard process—which produced a guilty finding that an appeal at Harvard overturned—needs to be trusted as the sole determinant on what happened in the case. She reaches this conclusion even though the initial panel (whose findings, again, were found unsubstantiated) didn’t have access to accuser Kamilah Willingham’s sworn testimony, didn’t have access to the physical evidence that came out at trial, and in particular didn’t have access to the cross-examination of Willingham. Neither Willingham nor The Hunting Ground team has chosen to release the transcript of that proceedings, which doesn’t speak well for their credibility.

Rosenfeld, meanwhile, asks us to trust her word that the initial panel got it right. She hasn’t said whether she read the trial transcript, or even that she’s read the transcript of the initial disciplinary hearing. She hasn’t even said whether such a transcript exists. She has strongly implied, in a Boston Globe editorial, that all students found guilty of sexual assault (however broadly the college happens to define it) in due process-unfriendly college tribunals should be expelled.

Hers is, to put it mildly, an odd approach to the use of evidence for a lecturer at Harvard Law School—even one whose research and teaching interests reflect a very specific type of gender-based ideology. Rosenfeld’s most recent publication, “Who Are You Calling a ‘Ho’?: Challenging the Porn Culture on Campus,” gives a taste of her scholarly interests. And her personal website labels her an “advocate” on the issue of campus sexual assault, with a background in representing sexual assault accusers who filed Title IX complaints against their universities. That someone with such a preexisting bias on the issue of campus sexual assault had any connection with Winston’s case—as Rosenfeld’s Inside Higher Ed comment implies—raises significant questions about bias.

Rosenfeld has offered a broader description of her attitude toward how colleges should handle campus sexual assault in an eye-opening essay for the Harvard Law Review. Here, she spoke positively about the “affirmative consent” standard—which shifts the burden of proof to the accused student and effectively requires him to prove his innocence before a campus tribunal. She favorably cites the research of David Lisak—which Reason has subsequently discredited. And she conceded that the college disciplinary system winds up falsely branding some students rapists. But the fear of a false accusation is “overblown,” Rosenfeld assures readers, and in any case isn’t all that relevant. “If the private [emphasis added] and public costs of false negatives are at least as high as the private and public costs of false positives,” then the due process-unfriendly college system needs no change—except, perhaps, to make it even more procedurally tilted in favor of the accuser. Rosenfeld’s standard is one that never would justify a college respecting due process, since advocates like the Harvard Law lecturer could always claim “private” (and thus unmeasurable) costs for the “survivors” that she represents.

Rosenfeld’s essay also features a series of vignettes about purported cases in which she was involved. Here’s one of them:

Another case involved a young female student who was out with friends drinking. When her friends wanted to go elsewhere for the night, they dropped her off at a male friend’s room, who said he would take care of her until the effects of the alcohol wore off. Instead, he and another male friend raped her. One of the accused men then bragged to a third friend about what the two men had done, who in turn then circulated the story in an email to his living group. The email corroborated the young woman’s account, including particular sounds she remembered hearing. She brought a complaint against the men through her school. They were held responsible and suspended, but then hired an attorney who prevailed upon the school to overturn the decision.

How did the attorney “prevail[] upon the school”? What was the school’s appeal process? Did the attorney file a lawsuit, or threaten to do so? Did this “case” on which Rosenfeld said she “worked” produce any documentation, so a reader could evaluate her credibility? Here’s Rosenfeld’s answer: “As I was researching this case for any public record to supplement this case description, I instead found a recent case from the same university . . .” In other words: there might (nor might not) be written documentation to test the veracity of Rosenfeld’s recollection. But she’s not going to provide any of it, and she’s not going to do anything to identify the parties (or even the date) of the case so someone can check.

That sounds like an approach that The Hunting Ground team would have preferred reporters use in the Winston case. No wonder they’re so fond of Rosenfeld.