"I served for 8 years on a university sexual misconduct board and at the end of that distressing tour of duty concluded the following..."

"(1) the combination of alcohol abuse by both parties (which is the case in the vast majority of charges), absence of witnesses, and absence of any forensic investigation in the student led process makes the charges almost impossible to prove by any standard of evidence; (2) a very small number of sexual predators can create a lot of misery on a campus (3) peer pressure and buddy systems by both male and female students are probably the best form of prevention; and (4) cases of sexual assault should go straight to the police and courts. Universities aren't equipped to handle these cases and need to stop trying to serve as a parallel justice system. This is not a place for amateur hour."

That's the second-highest-ranked comment on the NYT article: "A Majority Agreed She Was Raped by a Stanford Football Player. That Wasn’t Enough."

"A Majority Agreed" means that 3 members of a panel of 5 found that the male student committed a sexual assault. It "Wasn't Enough" because Stanford required at least 4 out of 5 on the panel to find a preponderance of the evidence against the accused. Preponderance is the lowest standard of evidence, and the panel can include students (as well as faculty and administrators), but the NYT still called this "an uncommonly high bar."

Stanford has changed the procedure to a panel of 3, and now the panel must be unanimous:

“In deciding we wanted well-trained, long-term panelists, it made sense to go to a three-person panel,” said Pamela S. Karlan, a Stanford Law professor who is now chairwoman of a sexual assault advisory committee, “and having three people decide something by a preponderance of the evidence seemed to us the appropriate way of deciding whether a life-altering sanction should be imposed on somebody for his or her behavior.”...

Michele Dauber, a Stanford law professor... said she doubted that the university’s proceedings complied with Title IX.... “You have to look at the process holistically, and when you see a series of hurdles and roadblocks, this becomes a very unfriendly place, if not one of the most unfriendly in the nation,” said Ms. Dauber, one of five Stanford professors (including Mr. Palumbo-Liu) who wrote an open letter in December 2015 to the provost complaining about the new policy. “The victim should not need to garner three votes to win while the respondent needs to garner only one. That is basic inequality.”

Those last 2 sentences of Dauber's are — as her use the word "garner" suggests — disingenuous. The accused and the accuser are not in the same position, and the panel isn't voting to pick one of 2 individuals who are vying for the same thing. It's not a question of equality. It's a question fairness — giving enough process to the one who faces a very serious deprivation. 

ADDED: I don't like Dauber's reference to the "victim" and the "respondent." The absence of parallelism shows the strain. Before the outcome is determined, it is emotive and biased to call one person the "victim." Dauber is decorous enough to avoid saying the word that goes with "victim":  "perpetrator." She says "respondent." The word that goes with "respondent" is "complainant."