Morning Coffee Feminism: Risk Assessment and Dying in the Park

Content Note: assault, sexual assault, rape, trauma, abuse, violence, rage

I had to calm down before I could write this post. It’s taken about a week and a half since I cracked, crying, unable to stop, unable to believe what so many people seem to think.

I mention this not to demonstrate my emotional fragility (I’m actually in pretty good shape), but to emphasise a point before we go any further: this is personal. I don’t just mean for me – I mean for women in general, or for people socialised as women. It’s personal, and it has a deep, heavy weight to it, one that I didn’t even realise until I finally caved under the onslaught.

But for people not affected by it, it seems to be a thought experiment. A mild sense of discomfort, a desire to play “Devil’s Advocate”, to run the numbers, to make a wry face and look offended. They dive eagerly into the discussion, tossing aside heart-wrenching personal experiences, ripping apart horrifying statistics of abuse and suffering, insisting that they and they alone are the logical ones, the sensible ones, and they demand attention right now.

This makes the discussion inherently unbalanced, right from the start. It makes it exhausting for one side, because we are cracking ourselves open to try and get through to people, and there is only so much feeling and horror you can share before you are wiped down to nothing; but the other side is tireless, relentless, and apparently lacking in the ability to shut up and listen.

Be kind. If you don’t know, if you don’t understand, if you need to question and discuss – be kind.

I don’t do well in those kinds of exchanges. I simply don’t have the patience.

But this? This I can do.

The objection appears to be thus: women should not be angry or suspicious of male behaviour just because a tiny minority of men attack women in parks and murder and rape them.

If women are suspicious and scared, if they do things like cross the road when a man is walking behind them, if they carefully assess their exit strategy before being alone with a man, they are – apparently – being awful discriminating bigots because they are tarring all men with the same brush. If they point out that, among men, there is a very common attitude of entitlement to women’s time, attention and bodies, an entitlement that we have always dealt with, and usually tolerated with a smile because the alternative is personally dangerous? How dare you generalise. I don’t have that attitude at all.

If they dare to suggest that maybe raising men the way we do is a problem – and here, I mean limiting their accepted emotional expression to superiority, lust, anger, and other feelings on that axis, and then tightly tying those emotions to masculine identity, and then heavily emphasising the overwhelming importance of gender to their very sense of self, not to mention making it almost psychologically impossible to ask for help when they’re at breaking point?

Well, you can’t say that. That means you hate men.

(I mean, I point out that we’ve tied men in psychological knots socially, because I would like less men to commit suicide, actually, which kind of suggests that I like men, and I would like them to be less miserable, not to mention less homicidal)

Apparently, the only way to treat male violence towards women is as an inevitability, a natural disaster, an unstoppable force. It just happens.

Except it’s not. It’s not an accidental, unavoidable tragedy. It’s an action that someone chose to take. There is agency here.

And we, as women, as apparent potential victims, don’t have any control over that choice, made by people bigger than us, stronger than us, who aren’t limited by assumptions of civilised behaviour.

We can only make choices about our own actions, and it’s a complex mess of risk assessment.

Let’s talk about risk assessment, because if you’re wondering why you’re going to be tarred with the brush of potential threat, you need to understand this.

We’ll start with Type I vs. Type II errors.


It took me a long time to be diagnosed for Crohn’s Disease, because the sneaky bastard was hiding deep within my small intestine, a location that is largely inaccessible to most diagnostic procedures. If the disease is advanced enough (ultimately, mine wasn’t), it can be seen in a particular kind of abdominal MRI.

Now, an MRI isn’t a statistical test, but it will illustrate my concept nicely, so bear with me.

When we test for things, statistically, we’re generally looking for the effect of one thing on another thing. We’re trying to see if two variables have a relationship.

Our basic assumption, to start with, is that there is no relationship. We refer to this as our null hypothesis.

Then we have to test this null hypothesis. We need to find out if we should accept it (no relationship between those two things, back to the drawing board) or reject it (aha! Something is afoot! MORE TESTING!).

When I go in for an abdominal MRI, my null hypothesis is that I don’t have Crohn’s Disease. That’s situation normal. Nothing will show up. (again, I am stretching this a little. Hush)

The MRI will either give a negative result (null hypothesis true) or a positive result (null hypothesis false – I have Crohn’s).

Of course, these tests have to have options for failure. No test is perfect. If we mistakenly reject our null hypothesis, thinking that something is going on when it isn’t? That’s called a Type I error. If we mistakenly accept our null hypothesis, thinking that nothing is going on, when something in fact is going on? That’s called a Type II error.

In medical tests, the likelihood that the effect really is there when the test says it’s there is referred to as specificity. Conversely, the likelihood that the effect isn’t there when the test result is negative is known as its sensitivity.

Now, an abdominal MRI looking for Crohn’s in the small intestine has a specificity of – effectively – 100%. That means it’s almost never going to make a Type I error. If that test says I have Crohn’s, well, we can be pretty confident that I have Crohn’s and no further testing is necessary.

The test, however, has a sensitivity of about 86%. That sounds high, right? But that means that if the MRI says I don’t have Crohn’s disease, there’s about a 1 in 6 chance that it’s wrong.

The MRI did in fact say that I didn’t have Crohn’s. It came back negative. And it was wrong.

Type II error in action.

So now that I’ve given us a solid personal example of a Type I vs Type II error (in the absence of any actual statistics and tests of p-values, which isn’t really relevant here), we can move on to Basic Risk Assessment.


As a marine biologist, I’ve embarked on fieldwork where I’m required to SCUBA dive. This involved submitting risk assessment to multiple institutions involved in the project, and I won’t lie, I was genuinely intimidated by the paperwork I had to fill out for this.

But a basic risk assessment rubric requires two primary components:

  • Likelihood
  • Consequence

The issue of likelihood vs. consequence came up in one of my field trips in particular. We were in Tasmania, at Binalong Bay in the north-eastern corner of the island, and we were gearing up for a dive. We’d finally found a place where we could easily access the water, and the local environment looked like it would support the species that we were looking for.

I was halfway into my dry suit when two women walked past and eyed the three half-dressed SCUBA divers. One of them said, “You do realise there’s a shark out there, right?”

We laughed, because we get this all the time. Not a day goes by on a field trip that some winking fisherman doesn’t say “Aw, mate, you better watch out – I saw a shark out there this morning, and he was THIS BIG” while yanking his hands out as wide as they can go. People like to mess with divers. We don’t worry about it.

“Yeah, I’m sure there is,” one of us said (I don’t remember who) with a grin.

“No – really. There was an attack yesterday. A thirteen year old girl was attacked and a bunch of the beaches are closed. You didn’t notice the helicopters?”

We collectively blinked, and looked up. Sure enough, there was a helicopter, apparently patrolling the area in search of a shark.

We thanked the women for their warning, and then had what we later referred to as an emergency meeting of the OH&S committee in the carpark, half-dressed.

We knew – as marine biologists working in Australia, we knew – that the shark in question was almost certainly miles away by now. They don’t hang about. Significant encounters are uncommon to start with (I’ve been diving on the south coast of Australia for 12 years and never encountered a large shark), but repeat encounters are just about unheard of.

We knew – based on experience, statistics, migration studies and the advice of genuine experts – that the likelihood of us encountering a large, dangerous shark in Binalong Bay on the planned dive was very small.

But the consequence, if we did?

There are usually about five levels of consequence in risk assessment rubrics, ranging from “minor” (bruises and grazes, minor damage to property) to “major” (hospitalisation required) to “catastrophic” (death, permanent disability, multiple deaths, etc.).

It’s fair to say that a shark attack on a group of research divers would fall under the “catastrophic” consequence level. Level 5.

Now, when you calculate the risk value for paperwork, you multiply the likelihood by the consequence, and in this way, extremely unlikely events become “high risk” by way of their catastrophic consequence.

We decided not to dive. If nothing else, we’d feel pretty damn stupid in those last awkward moments, and no-one wants that.

The potential consequence was too high. We leaned towards a Type I error over a Type II error.


Let’s put this together, and we’ll bring in the third factor here in our risk assessment: controls. Controls are necessary in order to bring down the level of risk. We want to reduce the likelihood of the event, or even the consequence. If you’re wearing the appropriate protective gear, perhaps the worst consequence of a particular laboratory disaster is a few bruises instead of death.

We enact controls all the time.

When women assess their personal safety in regard to the risk of male violence, we’re deeply aware that our Type I and Type II errors are unbalanced.

Let’s say that our null hypothesis is that Fred is a nice, normal guy, who honestly has no intention of attacking or raping anyone. Let’s expand that: he wouldn’t even continue to kiss a woman if she pulls away, even if she’s already said yes. He wouldn’t even have sex with a woman who’s passed out drunk. Guy’s a real prince (yeah, I have a bugbear about consent. Fight me).

Our null hypothesis is basically: Fred is not a threat to my safety.

Let’s look at our Type I and Type II errors in terms of consequence.

Our Type I error is that we assume Fred is a threat when he isn’t. This, by the way, is what so many men seem to be up in arms about. They’re furious and angry that women might treat them like potential rapists (or, hey, murderers). That’s the Type I error.

I’ve decided Fred could be a threat.

I enact controls.

Some example controls might be:

-crossing the road if Fred is walking behind me; getting off the train if Fred is making me feel scared of uncomfortable; calling a friend, either to make it clear I have help on the line, or even to come and pick me up; hiding in the toilets; taking out my earphones so I can hear if Fred’s footsteps are coming closer; perhaps making sure that I’m never alone in a room or an elevator with Fred, if I’ve encountered him at a social event; not accepting a drink from Fred-

It depends on the circumstances.

All these are designed to reduce the risk. They involve living in fear, or at least suspicion, and a heightened sense of stress and awareness, any time you’re in a vulnerable situation. The point so many women are trying to make is that we are constantly making these judgements and assessments, and constantly deciding whether or not we have to enact these controls, whether we have to sacrifice yet another piece of our autonomy, yet another moment of a life we’d like to live without fear that someone will decide they have a right to our bodies, to our lives.

It’s tiring living in this kind of interrogative framework.

Alright. Let’s talk about the consequences of our Type I error. I’ve enacted my controls, my risk assessment has said Fred could be a problem, but I’m mistaken. Fred’s apparently a mensch, and not really a threat.

Oh no. I’m wrong. I’ve made a mistake. It happens.

What’s the consequence?

Fred’s feelings are hurt. Fred feels affronted. Fred feels taken aback. He feels frustrated, maybe even angry, that someone has jumped to conclusions about him. Does he really seem like that kind of guy? Really? It’s not fair. He’s never done anything to deserve this.

And I’m sorry, Fred, because in our scenario, that’s absolutely true. You haven’t done anything to deserve this. And it isn’t fair.


Let’s talk about our Type II errors.

Let’s say we’ve decided that we can’t live in fear. It’s all overblown hype anyway. Fred seems decent enough. Who cares that he’s following a little closely? Who cares that he seems way too interested in where you live? And maybe there’s not even anything like that, and he really just does seem okay. You’ve got mutual friends. Maybe he is a friend of yours, or a family member. For whatever reason, you’ve decided he’s fine.

We’ve decided Fred isn’t a threat. He’s not going to do anything. Great! We don’t need to enact any controls around him.


Now we have to talk about the consequences of a Type II error, because if we’re wrong about Fred, it could get ugly. Really ugly. I’m not talking about him being an inconsiderate douchebag who’s disappointing in the sack, or any number of relationship crimes. He could rape you. He could maim you. He could kill you. He’s strong enough – stronger than you. If he decides to ignore your refusal, if he loses his temper violently, the outcomes for you are really, really bad.

They’re level five: catastrophic.

And you didn’t enact any controls, because this is a Type II error. Or maybe you did, just out of habit, and it didn’t reduce the risk enough, or maybe it did and you just got unlucky.

Stepping back and summing up: when a woman is assessing her safety in regard to a particular guy, she has to make assumptions. She runs the risk of being wrong, no matter what those assumptions are, because while some guys give off dead fucking creepy vibes, as a rule, they don’t wear signs or pointy hats indicating what side they’re on. She can be wrong in a Type I sort of way – assuming he’s a threat when he isn’t – and then she’ll enact controls, and he’ll be offended and hurt, because it’s not fair that he should be assessed as a danger when he isn’t one.

Or she can be wrong in a Type II sort of way – and be killed.

That’s what we’re talking about. That’s why this hurts so much. When men start getting angry about women jumping to conclusions about potential rapists and murderers, when they start talking about how it’s unfair and about how we can’t talk about male violence because it’s rude and mean and generalising (in spite of the fact that male violence is a fucking problem), it hurts. Because what they seem to be saying is that it’s okay if we die in parks, if we’re beaten bloody and violated and living in fear, as long as we don’t talk about male violence and offend them.

I mean, those things just happen, right? That’s just how it is.

It’s not as though someone made a choice to do it. It’s not as though it matters why that choice was made or how we can prevent it.

Let me tell you: no one ever made the world a better place by saying that’s just how it is. No one ever fixed a problem by refusing to talk about it, or ignoring the reality.

And if we’re emotional about this, it’s because we’ve been living with this weight our whole damn lives. We’re raised with it. We’re saturated with it, constantly told what we’re supposed to do to avoid it. The statistics on domestic violence are horrific – men we trust, men we live with do this to us (and I am aware that women commit domestic violence, given that I am an actual survivor of one, so spare me that – and it’s a conversation that needs to happen, but it’s not this conversation), but we’re also warned about strangers – men we don’t know, men we can’t see, men we can’t defend ourselves from because there are monsters out there in the dark.

It is impossible to assess the other component of our risk assessment rubric, likelihood, because the cultural saturation, the stories we hear, and our own personal experience simply reinforce the fact that we can so easily be hurt, or killed, and it happens so often. Do we calculate the likelihood of being attacked by a man in a park after dark? Do we calculate the likelihood of it in this park? What about dusk? What if it’s a main street, but it’s late at night? Or maybe we should calculate the risk of a date rape – we’re alone with a man we’ve been our with a few times, and things seem fine – but they’re not, because he thinks a couple of drinks mean yeah, of course. Should we calculate the risk of domestic violence, of the man we married losing his temper and punching us so hard that our brain ceases to function? It is profoundly more likely that we will be hurt by someone we know than by a stranger.

But these specific likelihoods are not accessible to us in the moment.

All we’ve got is a general idea of threat. And it’s skewed. Things that we hear about seem more likely (see “availability heuristic”), and no amount of people shouting not all men is going to make the dark seem safe, or going to make us want to accept a drink from a guy we don’t know.

And if we do decide that the risk is low, that it’ll be okay, that we can walk home at night?

We’re told we don’t enact the right controls. We don’t work hard enough to be safe. We aren’t – fucking hell – situationally aware.

And the reason, dear god, the reason women are so angry about this is that yes, we are situationally aware. Of course we are. Don’t teach your grandma to suck eggs, you condescending muppet. We are trained subtly, from childhood onwards, to fear. To assess. To look for exits. And even when we do everything perfectly, we die, and are told we just didn’t do it right. If we dare to put down the weight, for a night? If we decide we can’t live this way, and it’s a kinder world to live in if we just don’t make those generalisations that are apparently so awful and offensive and unkind? Not only do we die, not only are we attacked and hurt, but then we’re judged for not being more suspicious.

And then quite often we blame ourselves, because we can’t help it, even if we know better, even if we’re enraged and furious and we know that it wasn’t our fault – we blame ourselves. And the trauma deepens. And the world perhaps blames us, and we can’t get help. We can’t get support. We can’t be heard.

Because in that case, we weren’t suspicious enough.

This is why it hurts. You’d rather we suffer and die, you’d rather view those deaths and those attacks as an unfortunate tragedy that just happens in the world rather than as something that is deliberately done to us. You’d rather that this keep happening than talk about the problem, because talking about the problem makes you feel bad. It makes you feel as though maybe you should do something about a “women’s problem”. Because thinking about it as “male violence” makes it your problem.

You didn’t do anything wrong, so you don’t feel you deserve the weight of this problem.

Guess what: I didn’t do anything either. I don’t deserve the weight of this problem either. And at this point, I don’t really care that you feel uncomfortable or generalised about. I just care about getting out of this lose-lose cycle. I care that the little girls growing up now don’t have to perform these kinds of bullshit, imperfect risk assessments. I care that they don’t grow up carrying this weight. I don’t want them to have to learn these controls or solutions that don’t even fucking work. I don’t want them to try to figure out whether it’s their fault somehow when they get hurt.

I… don’t want them to get hurt in the first place. Not like that.

The fact is, there’s no perfect set of controls. There is no winning move in this game. At some point, we need to be able to live in the world.

We need to be able to walk home.

And that’s why we’re furious.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s