
Bullshit Detection by Proxy
A tool for epistemological humility and overcoming illusory superiority
The best lack all conviction, while the worst are full of passionate intensity.
— Yeats, The Second Coming
When Russia invaded Ukraine, I told my friend (half-seriously): “okay, I’m going to guess the population of Ukraine and if I don’t get within 5% of the correct answer, remember to chastise me if I ever take my own views too seriously or expect others to.” My guess? I think it was about half of the actual population. Turns out Ukraine is big. Who knew? I call this Bullshit Detection by Proxy (hereafter BDP for short).
This type of proxy is useful for quickly sifting through the sewage of social media. After you’ve read some pundit’s hot-take or a ‘Slava Ukraini’ Tweet, ask yourself whether they would have been able to name the major Ukrainian cities before the first aerial bombardments. If they still can’t name the cities today, that’s something else entirely.
This isn’t to say that one needs an in-depth knowledge of Ukrainian history in order to feel or show compassion. Instead, it’s simply a tool to spot humility (or the lack of it) when a conversation turns serious and when people demand their view counts.
We can illustrate just as easily by critiquing the very people who claim to be above such instinctive, unthinking tendencies, like some of the people who retweet the “I Support The Current Thing” meme, for example — a meme I too enjoyed. You would bet good money that many who retweet the ISTCT meme would fall prey to what psychologists call illusory superiority:
In the field of social psychology, illusory superiority is a condition of cognitive bias wherein a person overestimates their own qualities and abilities, in relation to the same qualities and abilities of other people. Illusory superiority is one of many positive illusions, relating to the self, that are evident in the study of intelligence, the effective performance of tasks and tests, and the possession of desirable personal characteristics and personality traits.
The Wiki has a plethora of examples. The most famous one being driving:
93% of the U.S. sample and 69% of the Swedish sample put themselves in the top 50%; for safety, 88% of the U.S. and 77% of the Swedish put themselves in the top 50%
To make an obvious point obvious: we enjoy the ISTCT meme because it makes us feel like critical thinkers, the type of people who don’t fall into the slipstream of mainstream media Goliaths. We aren’t non-playable characters, we think for ourselves. At least until we’re presented with our own personally triggering issues. Then we wax lyrical with uninformed (or, to be precise: under-informed) views.
Again, this isn’t an argument for having a bachelor’s degree in a field before you’re allowed to speak. Indeed, there might be something to say for a wisdom of crowds effect with many issues. For example, British concern over immigration seemed to track the number of immigrants arriving quite closely (see David Goodhart’s The Road to Somewhere). Likewise, many people enjoy a pointless pub debate about who the best athlete is in any given sport. We might also learn something from someone who doesn’t know enough to have an informed view, but who knows more than us.
As such, these proxies shouldn’t be used to shut down humble people. The type who, if you asked them, would probably admit to not knowing enough about an issue. In fact, these aren’t tools to shut anybody down (unless the conversation has reached a tipping point of hostility and rudeness and you think that leaving them with such a question before blocking them or walking away could help). Instead, Bullshit Detection by Proxy has two main uses. The first I’ve already noted: a mental device for you to filter out ignorance. In this use case, you never actually ask a BDP question aloud. The second use is trickier: using BDP in dialogue.
If you’re going to ask a BDP question, it should be framed either as self-deprecating (if this can be done earnestly) or exploratory, ideally both. It’s also worth combining with other tools for epistemological deconstruction, like asking your interlocutor to rank their certainty in the stated proposition on a 1-7 scale and steel-manning their view. Say you’re dealing with a 7 who you suspect is uneducated on the matter of abortion. Whilst you aren’t an expert, you’re familiar with the basic biological facts and ethical arguments. The dialogue might look something like this:
Mary: I just think it’s a woman’s choice. Men shouldn’t have a say in decisions over a woman’s body. They have no right.
Matthew: I think I understand your position, but I want to be certain. Are you saying that because men cannot understand the first-person experience of pregnancy that this makes them more likely, on average, to overlook important ethical considerations and to have less empathy?
Mary: Yes, absolutely. If men got pregnant, they would have no problem with abortion, I’m telling you. But men make most of the laws and that’s a massive problem when it comes to women’s health.
Note: At this point in the conversation Matthew could ask a question like “do you know how many women are against abortion?”, assuming he knows the answer, of course. However, this isn’t really a BDP question because it’s too specific. BDP questions are, by definition, questions any mildly educated layman should know the answer to. That’s why they work. What’s more, Matthew is likely to meet swift resistance because showing people to be ignorant is a delicate matter at the best of times. When they are ethically invested, it’s a tight-rope walk of death. Mary might simply respond: “so what if 50% of American women are pro-life, they’re backward conservatives and they’re wrong.” Then the conversation is in the weeds and you have to extricate yourself to make the point.
Matthew: Right. So on a scale of 1-7, with 7 being absolute certainty, where would you say you land on the question of keeping men out of such decisions and law-making processes?
Mary: Oh 7, without doubt.
Matthew: So it’s one of the topics you feel most passionate about. I don’t know if I know enough to have an opinion, but I’d like to learn more. So what’s the legal term limit for abortion in this country?
If Mary doesn’t know the answer or even flounders, I personally think it’s safe to say that we have confirmed she is likely to know little of interest or substance, or that what she does know will be highly ideological and misleading. It’s up to Matthew to decide how to proceed after that.
Of course, this is an easy example where the person (Mary) is already behaving dogmatically. The more nuanced case might be where you’re chatting with friends and you simply want to inject a layer of epistemological humility with a BDP question:
Michael: Did you guys see the government’s announcement about sending asylum seekers to Rwanda?
Steven: Yeah
Freddy: Yes, disgusting.
Michael: Why disgusting?
Freddy: Sending them to a poor African country is ridiculous.
Michael: Well the government is paying them a hefty sum by the looks of it.
Freddy: I think it’s still appalling ethically. Rwanda has it’s own problems.
Steven: We have to do something though. The situation is getting out of hand.
*Debate ensues between Steven and Freddy*
Michael: By the way, do either of you know the capital of Rwanda? I have no idea!
If Michael can say this with a bit of self-deprecating humour, it’s good BDP. But again, I cannot stress enough that BDP isn’t meant to humiliate people or to prevent a good conversation for the sake of it. As already stated, if Steven or Freddy don’t know the answer it doesn’t mean they have nothing to contribute. It just increases the probability that what they will say should be looked at sceptically if you’re interested in the truth or becoming more informed. Another good heuristic along similar lines is: would this person be speaking as confidently if they knew I was an expert on the topic? In most cases, probably not.
Below is a list of bullshit detectors you can use to check yourself and others, sorted by popular topics. Note that BDP can be non-political. If someone is trying to convince you about a particular diet you could ask, for example, “what exactly is a carbohydrate?”. However, if you just read out the question like an autistic robot, it’s obviously not going to have the desired effect. Subtlety and charm are key. Feel free to add your own BDP questions in the comments below.
Anything where a basic statistical understanding is necessary for informed debate:
What is a standard deviation?
Geopolitics:
What’s the capital of that country?
What’s the population?
How old is that country?
What are the main cities?
Immigration:
How many immigrants are there each year?
Welfare:
How much does this country spend on welfare?
Foreign policy:
Which countries does X border?
Tax:
What’s the difference between these taxes?
How much tax do the top 5% pay?
What’s the highest tax rate?
Health:
How many doctors are there in this country?
If you liked this, please do subscribe. Maybe give it a share to spread the gospel of rationality. And if you want to support my work, you can do so with a paid Substack subscription or using the following methods:
https://www.patreon.com/Ideas_Sleep
BTC wallet address: 1KHB3Mq7njTGfquABcREsiywaxmDbP2NPY
And subscribe to my YouTube Channel here.


