Domestic violence research has yet to pinpoint a definite predictor of why one person and not another becomes a victim. They don’t look or act a certain way. Some live paycheck to paycheck; others never worry about the rent. Some have Ph.D.s; others never made it through high school. Some are outspoken; others are wallflowers. Their one commonality: They gravitate toward men who want to control them. This is why falling in love can turn out to be the most dangerous thing a woman ever does.

It is nearly as difficult to anticipate who will become an abuser as it is to predict who will be abused. Still, there are signals. A partner may seem too good, too perfect. What begins as flattery turns commanding. You’re too beautiful. You’re flirting. You’re slutty. You’re a whore. And so on. A disconnect emerges between actions and words. Inevitably, it gets worse.

There is one semireliable way to tell if someone might become a victim: Children who grow up around domestic violence are 15 times more likely to be abused themselves. Even worse, boys who witness abuse at home are twice as likely to become abusers because they’ve been rewired to think it’s normal. Living in a violent domestic environment also can slow sensory growth in the brain and increase depression, suicide risk, and problems such as withdrawal, criminality, and sleep difficulties. Even if these children aren’t actually hit, they rarely escape the abuse at the time—or later in life.

For those unable to flee, the abuse and the make-ups develop a cyclical quality. After a crisis, an abuser may start to resemble who they at first seemed to be: caring, apologetic, changed. Then another violent episode happens. Then another vow to change. This is why many domestic violence victims aren’t identified until they land in the hospital. They come in complaining of bruises on their neck, broken ribs, or black eyes. They usually have an excuse, claiming to be klutzy or accident-prone, which can fool the untrained eye. Unfortunately, less than one-fifth of domestic violence victims get hospital care.

Moreover, the psychological impact of domestic violence in the United States accounts for $18.5 million in mental health care visits each year—and that’s just the victims who do seek help. Those who don’t may slip into something akin to Stockholm syndrome, in which a hostage feels sympathy for her captor and starts doing things simply to make her abuser happy and fend off the inevitable next attack for just a little longer.

It’s easy to wonder: Why didn’t she leave? Why didn’t she pack her kids up in the middle of the night and run? He wasn’t holding a gun to her head. She had a choice.

Most abusers don’t raise their violence or abuse beyond a certain level. They are shouters or slappers, not killers. Legal methods often work; most restraining orders are effective. Once they’re confronted with the legal and personal risks, abusers usually back down. But absent any outside pressure to stop, abuse can—and will—escalate.

So as many victims do, Krystal stayed. Because she loved him. Because she thought having this man as a dad was better than her kids growing up without one. Because she was afraid. Because she thought he’d stop. Because no one ever asked the key question: Why did he hit her?