How a closed, all-male system breeds harm and then buries it.
Part 2 of 6, about 11 minutes
Violence in sports often gets explained as just one bad man. Find him, remove him, and the locker room is safe now. The appeal is that it costs the rest of the room nothing.
But the research doesn't tell that story. Instead, it describes the groups, norms, and institutions that make harm more likely while also making it harder to see. None of what follows is about whether any one person did anything. This is about the conditions a closed system creates, and the reflexes it uses to protect itself.
Not one bad apple. The barrel does the spoiling. Decades of work on what researchers call male peer support, including DeKeseredy and Schwartz's Male Peer Support and Violence Against Women, describe the mechanism plainly: when a tight, all-male group holds norms that excuse or encourage the mistreatment of women, and members reinforce those norms for each other, the risk of abuse goes up. The group does not just contain a problem. It manufactures and rewards one, supplying both the attitudes and the social backing that make harm feel normal and consequence feel unlikely.1 A closed athletic environment is a near-perfect incubator for this: selected young men, intense in-group loyalty, status tied to belonging, and adults whose interests run toward protecting the group's reputation. The point is not that athletes are uniquely bad people. It is that the structure does predictable things to ordinary people.
The group does not “contain” a problem.
It manufactures one.
The structure isn't hockey's. It's men's. The structure is not unique to hockey, and it is not over. The easy version says hockey has a problem. The truer version is that men do, and hockey is one of the places it collects. The same machinery shows up wherever men organize into closed groups that reward one another, and two recent cases outside sport make that plain.
In 2024, a French court convicted Dominique Pelicot, who admitted he had drugged his wife unconscious and recruited dozens of men to rape her over years.2 What the case exposed was not one monstrous marriage. It was a system: a forum where men found each other and coordinated the assault of a woman together.
That forum was shut down in 2024. In April 2026 French prosecutors opened an investigation after it reappeared.2 In 2026, a months-long CNN investigation documented men still doing it, gathered in online groups to trade methods for drugging and assaulting women and to share the footage.3 One group it found had nearly a thousand members, and the host where such material sat drew around 62 million visits in a single month. A French lawmaker who works on this abuse called such groups an “online rape academy.”3
The mechanics change.
The structure holds.
The chatroom, the livestream, the locker room. In each, a woman's degradation becomes the currency men trade for standing with other men, and the group's first move, once seen, is to protect itself. Hockey is not the only such room, and it is not the worst. But it is one of the few that ever kept a manual.
They are the shallow end of the same water. Liz Kelly's work reframed sexual violence as a continuum, a connected range running from harassment and unwanted comments through coercion to assault, rather than a set of unrelated boxes where only the most extreme counts.4 That reframing is why the jokes are not separate from the assaults; they sit on the same continuum.
The severe rests on what's normalized below
Kelly's point is that these are not separate boxes. Selecting a level highlights it and everything more normalized beneath it, because the rare extreme at the top is held up by the everyday behavior at the base.
Disparagement humor is not just tasteless, it is functional. Research on what is called prejudiced norm theory, work by Ford and Ferguson, shows that sexist humor signals to a group that prejudice is shared and acceptable here, loosening the internal brakes on worse behavior.5 The locker-room joke is not the harmless release valve it is sold as. It is the norm being set.
It is not the release valve.
It is the norm being set.
It is the code that calls honesty the betrayal. We think the phrases give it away: “bros before hoes,” “don't be a rat,” the more violent “snitches get stitches.” All from the same social contract. A player interviewed for a 2023 study of men's hockey culture, describing dressing-room talk about girls, put it this way: “Well, on good teams it stays in there.”6 The study's authors draw the conclusion: the idea that “‘good teams’ maintain a code of silence demonstrates how recognition is impeded by insularity.”6 It is the brotherhood asking you, often indirectly, to cover for a teammate, and then recasting telling the truth as the real betrayal. This is the layer where a culture protects itself.
The reflex is not invented in the moment it is needed. It is taught long before, on every team that ever decided loyalty outranks the truth: close ranks, get the story straight, treat the woman as the problem and the question as the threat.
Loyalty is taught as a virtue.
It is also the cover.
This is male peer support again, doing its other job. The same bonds that supply the attitudes also supply the silence.1
Hockey has another code, and the two are easy to confuse. The fighting code is the set of unwritten rules, in Ross Bernstein's account in The Code: The Unwritten Rules of Fighting and Retaliation in the NHL, by which players “calibrate the level of physical contact.”7 The fighting code governs physical contact and retaliation between players. The code of silence governs what doesn't get reported off the ice.
Hockey Canada has named it too. In July 2022 it titled its response Action Plan to Shatter the Code of Silence and Eliminate Toxic Behaviour In and Around Canada's Game, and wrote that it “wants to end the culture of silence that exists in parts of hockey.”8
Everyone saw. No one moved. The failure of people to step in is not mostly cowardice, it is a documented group dynamic. The classic bystander research, Darley and Latané's, showed that the more people present, the less likely any one of them acts, because responsibility diffuses across the group until it lands on no one.9 A closed team amplifies every condition that produces this freeze: clear in-group pressure, fear of breaking ranks, and a status cost for the person who speaks.
The hopeful half of the same literature: intervention can be taught. Bystander programs that give people concrete scripts and practice measurably increase the likelihood they act.10 This is the bridge to the prevention work on the Resources page. The freeze is real, and it is trainable.
There is a sharper version of this. A culture like this does not require everyone to take part. It only requires that enough people not interrupt. In a closed room, silence is a role, and it holds the rest of the structure up.
Enough people not interrupting
is all the structure needs.
Two details sharpen it. The same loyalty that binds the room raises the cost of acting against one of its own, and a teammate is the hardest possible person to speak against. And the cost of the silence is not only that it shields people afterward. It shapes the moment. When a room treats something as normal, the one person it is happening to is stripped of the confirmation that it is not. A majority of survivors do not name their experience as rape, reaching for softer words instead, and the silence of everyone present is part of what teaches them to reach for those words.11
A crowd that does not flinch becomes the victim's reason to doubt her own alarm.
The institution protects itself first. When harm surfaces, the closed system's first instinct is rarely the person who was hurt. Jennifer Freyd's work names two patterns. Institutional betrayal describes organizations failing, or actively harming, the people who depend on them, by minimizing, disbelieving, or burying what happened to protect the institution.12 DARVO names the individual-level reflex that often travels with it: Deny, Attack, and Reverse Victim and Offender, so the person who reported becomes the accused and the accused becomes the wronged party.13
Recognizing these is not a verdict on any case. It is a description of moves you can watch institutions make in real time, which is the entire point of being able to name them.
Yes. At least once, it was written down. The culture is usually described like weather: ambient, unwritten, impossible to pin down. It is not.
In the early 2000s the Junior Hockey Bible spread as a viral website, a glossary of locker-room war stories that doubled as a marketing venture for a hockey clothing brand, Gongshow Gear.14 It was not folklore and it was not ambient. It existed, it had an audience, and it circulated widely.15
It is a glossary and a code. It supplies an alternative vocabulary for women and for group sex, assigns status for degrading a woman on camera, and sorts women into the disposable and the “pure.” The same document that rewards making a record of a woman's degradation also warns, elsewhere, to keep no copies, because evidence can be harmful. It rewards the record and fears the record in the same breath. It builds its own defense into the top of itself, telling anyone who cannot take a joke to leave.
The framing is the defense.
The joke is the alibi.
A journalist who grew up around the system described the document in 2014 as reading like a guidebook for sexual assault, and argued that team culture breeds rape culture with the manual as its framework.16 That is his characterization, and it is opinion. What is not opinion is that it existed and had a following, and that it did not come down on its own: when the brand's founders were confronted in 2004, they took the site down and issued a public apology. It has since been scrubbed from the internet.14
None of this is new. A quarter-century ago, journalist Laura Robinson documented the same machinery: boys drafted at sixteen, billeted far from family at the most malleable age, made into “men” by veterans, run through hazing rituals with sexual overtones engineered to push past inhibition.17
"Young gods," insulated from ordinary consequence and taught early that the game would protect them.
Laura Robinson, Crossing the Line (1998), her characterization of junior hockey cultureThat is the point of putting a document on this page. You do not have to prove any individual ever read it. A manual only tells you the vocabulary existed and the audience was fluent. A pattern is not a single event; it is recurrence, written down at least once and documented for decades. Naming it is how you keep the room in the frame.
Naming the pattern is how you
keep the room in the frame.
A popular statistic holds that a small number of undetected serial perpetrators commit the vast majority of campus sexual assaults. It comes from one influential 2002 study and is widely repeated. We do not lead with it, because later research following college men over time challenged that "serial predator" picture, finding offending patterns are more varied over time.18
We flag this on purpose. A site that demands honest language from institutions has to apply the same standard to its own citations, including the ones that would make the argument easier.
Rarely. Most reported sexual assaults never reach a courtroom. Combining Statistics Canada's reporting and court figures, fewer than 1 percent of sexual assaults end in a conviction (our calculation). None of this is a law of nature. Sexual violence gets treated as inevitable, as fixed as death and taxes, but most of what we file under inevitable is a set of values and attitudes, and those can change. The figures below are not a record of justice done. They are a record of how rarely it is. Canadian figures, the most rigorously tracked.
Follow the cases that do get reported. Of police-reported sexual assaults from 2015 to 2019, about one in five proceeded to court, just under half of those ended in a finding of guilt, and only about 5 percent of reported assaults ended in a custodial sentence.21
Of police-reported sexual assaults, 2015 to 2019, Statistics Canada (Cotter, 2024)
Widen the lens to every assault, not only the ones reported, and it narrows further still: the widely cited older model puts it near 0.3 percent, about 3 convictions per 1,000 assaults,20 while combining Statistics Canada's reporting and court figures, fewer than 1 percent of sexual assaults end in a conviction (our calculation).21 Sexual assault is the least-reported crime measured. False reports are rare: research puts them in the range of 2 to 10 percent (NSVRC).23 Among victims of violent crime (excluding spousal violence) who did not report to police, not sexual assault alone, women cited shame or embarrassment far more often than men (34 percent vs. 6 percent), along with feeling they would not be believed (25 percent vs. 7 percent) and that reporting would shame or dishonor their family (19 percent vs. 4 percent).19
About half of cases that reach court end in a conviction. Fewer than 1 percent of all sexual assaults do.
The difference is every case the system threw away before it started counting. This is not the exception that proves the system works. It is the system working as built. Changing the outcome means changing the culture that produces it.
The culture is the why. The pages that follow carry it forward, in order: the language the room speaks, then the system that handles what it produces.
Verified as of June 2026. Statistical figures confirmed at publish.
