1. 100% chance to die, mauled to death by bear OR 100% chance to die, raped, humiliated and traumatized

Which one would you prefer?
I can say for myself that, in this example, I would definitely rather be mauled by a bear and I'd assume 99% of humans will agree with me on this. Let's move forward from here and see what we can learn.

2. 100% to get bear'd OR 100% to get raped, but left alive

Which one would you prefer?
Now this one is a lot more difficult to answer because in one you'd at least get to continue to live. Would that be a worthwhile life? You will continue to be shamed and victim blamed after the incident. I could see both sides here and personally I hope I'll never have to make such a decision myself.

3. 50% chance to get bear'd OR 25% to get raped, but left alive

Which one would you prefer?
We're now getting into the territory of what the “Man or Bear?” question is trying to get at. Relatively certain death is better than a comparatively unlikely, but absolutely possible rape. Women, so the argument, would rather just throw in the towel than risk the rape, and I can understand them. Taking the bear here is very reasonable. Now we get to the statistical part and the question that is being asked by the “statistical response” to Man or Bear: Where is the limit?

4. 20% chance to get bear'd OR 5% chance to get raped, but left alive

Which one would you prefer?
This is now the ultimate question and where we will move from to find some kind of conclusion. I hope these are some estimates that everyone can at least work with. Bears aren't all that interested in humans and we also know that most men wouldn't rape women even if they could get away with it without consequences. Are the odds 5%? No, probably not, but that's exactly the point of the next few comparisons:

5. 20% OR 1%
6. 20% OR 0.5%
7. 20% OR 0.1%
8. 20% OR 0.01%
9. 20% OR 0.001%

With a conservative¹⁾ estimate I assume that one third of all women do get raped. If the average woman lives 80 years, 3 women live 240 years.
You would need 0.001% some 100,000 times to fill 100%. If we think of this as days, 100,000 days are 273 years. So I think that a woman, per day, has a chance of roughly 0.001% to get raped. Give or take. The numbers here are wildly off anyway, but this is good enough²⁾.

The point here is that somewhere here, most women would definitely stop taking the bear. I know this because at 0.001% this stops being a “choose one”, and it's now more of a “you live a normal day, but you can choose to also have a 20% chance of dying to a bear”. If the rape chance was any less likely then it would actually improve (ie. reduce) one's daily chance of getting raped and it would be kinda dumb to choose anything else. So…

Where is your limit?
It is now upon you to decide where your cutoff is. Even at 10x more likely than normal daily chance you'd probably still rather take that than the bear. But when does that stop? It might be useful to really think this through and think about where on the scale you'd go, and where too much is too much for you. Of course any of the rape chances above daily normal chance are too much if repeated multiple times, but this is a one-off thing. Whatever you choose happens once and then, if you didn't die, you go back to the rest of your life with normal daily rape chances.

The other question is, how dangerous really is a man? Like, what's the actual statistic for the likelihood of a man just sexually abusing a woman if they happened to meet in a space devoid of help and witnesses, ie. devoid of accountability?

All of us are going to find very different estimates here. Personally I think that a woman could probably go through at least 1000 men before finding someone who will actually sexually assault her. Of course many more men are going to be indecent or abusive towards her before that (which is shocking in its own right), but I think the overwhelming number of men won't do anything life-changing to a woman even in a space with no accountability. You might, and probably will, come up with your own calculations here. Personally I would put the chance at 0.1% and be good with it.

What we do now from here to reach a conclusion is easy. Where is your cutoff, where is your expected chance, and is your expected chance above or below your cutoff? Now in my case, my calculation would go roughly like this: For me, 10% risk of getting raped is too much. I know that with a 10% (or less) chance of it happening, I'm comfortable enough with my odds. The previously established 0.1% are below 10%. So, if 10% is the highest amount of risk I'm willing to go with then I'm perfectly fine with 0.1% and the correct answer to the original question is indeed man.

For some extra flavor we can also use the bear chance as a coefficient. Since I'm not guaranteed to die to the bear I can be cocky and reduce the limit of what I'm willing to take. If I take a 20% bear chance (ie. 1 in 5) then I can divide my limit like this: 10ppt / 5 = 2%.
I think this is reasonable because the bear chance matters in the chance calculation. If I am willing to accept risks with the bear then I can afford to accept less risk with men. In my case, 0.1% is still well below 2%, but this might change something for you.

The cool thing is that we can use the above-established framework to explain what women's real problem with men's justification of the Man or Bear question is. Some of the properties in this model are fixed. Whether you used the same numbers as me, the bear chance and the chance of getting raped are just fact numbers. Even if we estimate them differently, that is only because we all work with different information and have different guessing skills. But, the one thing that will be different between everyone's mental model of the formula is the limit, the amount of risk a person is willing to take before they'd rather take the bear.

This limit is, to a degree, informed mathematically, but for many (especially women) it will be an expression of their fear. While there may truly be a rational, mathematical solution to the problem, people who are afraid of things will naturally try to avoid falling victim to it. Even within women, the reported limits will differ from one person to the next and they are exclusively a measure of how afraid they are of men and ultimately what the whole conversation is about, even if they are not consciously aware of: The statistical difference of the reported limit differences between men and women.

That's not irrational, in the case of cases a bear is better than a man.

The quick answer to this question: It depends on interpretation.
The general premise of this question is the tradeoff between certain death (bear) and less likely, but much more traumatizing death (if at all)(man). Women are afraid of being subjected to rape and sexual assault/abuse and risking that is not always worth it if the alternative is a simple, quick, unprejudiced death to the local neighborhood bear. The idea is that far off of civilization, a woman is completely defenseless against a stronger male and therefore much more likely to get assaulted (or worse).

From here on, this question becomes a matter of probabilities, and that set of probabilities depends on how we define some key terms of our equation.
We use the word “encounter” and for a moment we do not define civilization strictly. If you are genuinely far off of civilization, then that means you would not be on hiking trails. These trails are manmade and therefore an extension of civilization. But let's for now stay on trails. Additionally, we define “encounter” as a genuine real-life event as it could genuinely unfold, ie. two people meeting naturally. What I mean with that is that the man in question is not just randomly teleported there, the man in question is a man as you would naturally encounter them there³⁾. On a hiking trail, the most likely male person to naturally encounter is, well, another hiker. The problem with hikers is that they are unlikely to do anything serious to a woman. The majority of men aren't psychopath hawks circling and salivating over women who go out into the woods. Yes, most violent crime is committed by men, but most men do not commit violent crime, especially sex/gender-based violent crime.
So, on that hiking trail, the pool of people you could naturally encounter contains, to the overwhelmingly large extent, mostly hikers. The number of serious threats, ie. men with genuine intent to hurt a woman (which can include some hikers), is going to be dwarfed by just how much more likely you are to encounter a normal, peaceful hiker. Uncomfortable and perhaps scary encounter, still, but the odds are in your favor.⁴⁾


The much more serious problem is when we go back to the original: “Away from civilization.”
This is how I think the question was originally meant, and it means that we are off the hiking trail and in the absolute wood of the wood. Now it becomes much harder to find people you could meet that actually are not threats. Maybe the ranger? A hunter maybe? And even they are still more likely than other men to be a genuine threat, so at this point I understand women who are beginning to calculate their odds and find their risk of getting violated is too high to justify.

An interesting bit about social media trends is that they are the equivalent of word by mouth and get jumbled up in transit. The way it was first brought to my attention was with the wording “who would you rather be alone with in the woods - a random man or a bear”. “Encounter” is already kind of nebulous and leaves lots to interpretation about how the man got there, but this one doesn't make any effort to answer that question at all and lets you fill in those gaps all on your own. My initial understanding of that for many years was that they kind of just get teleported there? Like, if the wording doesn't mention this then in my brain it felt like the way the man gets there doesn't really matter, so my brain just filled it in with “something, doesn't matter, magic, they're just there - teleportation”.

Now, in retrospect I can say this was quite a serious oversight and I just didn't think about it enough, but to be fair to me, I also did not have the original wording of the question. So, if a woman is alone in the woods, away from civilization and a random man drawn randomly from all of the world's population is teleported in front of her, this is a much better bet than having a bear teleported in front of her.

The pool of people this roll can draw from is large and all but eliminates factors that would otherwise make you more likely to draw someone with hostile intent. The reason you would be less safe in the woods is because it attracts predators looking for defenseless victims, therefore inside that pool people with hostile intent will be over- or at least more represented. If you draw from a completely random pool, this overrepresentation vanishes, and you go back to the baseline, and the baseline is that most men do not have hostile intent to women ⁵⁾. In this setting where a random man is just teleported in front of the woman, it would be stochastically unwise to choose a bear instead.


I worry that this misunderstanding may be what causes a significant portion of the drama around this question. Most men responding to this question aren't psychopath hawks circling and salivating over women who go out into the woods; but these non-psychopath non-salivating non-hawks are also unlikely to “encounter” a woman in the forest in the first place. If it comes down to “encountering”, you and I and the majority of non-hostile male people are not going to be the kind of person a woman would encounter in a forest. The people a woman might encounter in a forest are probably still part of the minority of non-hostile rangers, non-hostile hunters or non-hostile hikers, but that minority is now beginning to get worryingly small in comparison to the people who are actually dangerous.


Interestingly, if we just teleport any one random man then the woman might happen to find themselves in the forest with family or a domestic partner and - incidentally - this is the group that poses the greatest threat to any one woman. You'd rather be in the forest alone with a random man than with any one of your family members. That's one to take home for consideration.

To be fair to bears - bears are gentlemen. With the exception of polar bears, bears mostly only attack humans out of defense/struggle for survival⁶⁾, not as part of their predation. When given the opportunity, bears will typically bail, including the women offering themselves to them. Statistically speaking, you're probably fine in an encounter with a bear (unless you're stupid, too close or unlucky). Depending on the pool of people you choose to randomly draw from, odds are that your bear is still more likely to hurt you (in any capacity) than a man. However, as far as I understand this, the argument being levied with this thought experiment is not the actual probabilities.

It seems to be that the argument is an expression of significant (and justified) fear more than an actual deliberation of statistics. Personally, I'm not exactly sure where this is intended to go? Like, the correct answer lies in the statistics. Either you are more likely to get harmed by a man or not, and then you can throw in a bit of Personal Preference about just how much you think you would hate getting sexually traumatized, and come to your conclusion. There is no debate here, and men are aware that women are afraid of them. Men don't need to be told that they scare women - there is nothing useful a man can say about this, so to them this discussion will invariably become about statistics.

Obviously, some misogynists do like to hijack this argument and talk about how either 1) women are irrational about men and bears or that 2) gender-based violence isn't a thing (or that men are the actual victims). This is spurious. Women are disproportionately victimized and are justified in their fear.