We Are Weak and They Know It
Manipulation without the feeling of being manipulated is the most effective kind.
28 February 2026
997
997
Hi, I'm Richard. On this blog, I share thoughts, personal stories — and what I'm working on. I hope this article brings you some value.
I’m Surprised by the Confident Use of Words Like Certainty and Causality
Data, statistics and knowledge limits
Listen to this article
0:00 / 0:00
Today I came across a post on LinkedIn by a digital specialist. He confidently claimed that with an A/B test, we can determine not just correlation, but true causality. He used words like “certainty” as if statistics were part of Newtonian physics — clear, absolute, unquestionable. I’m surprised by that level of confidence. I don’t have it.
We See Causes Where There Are None
Our brain craves order. When something happens after something else, we instinctively think: “the first thing caused the second.” Got a headache? Must’ve been the coffee. We’re built to look for causes — even when they’re not there.
From an evolutionary perspective, this makes perfect sense. If you hear a rustle in the bushes, it’s safer to assume there’s a tiger and run, even if it’s just the wind. Evolution has taught us it’s better to be wrong than dead. Maybe that’s why we tend to see patterns in randomness, connections in the unconnected.
In the Middle Ages, people believed comets brought disaster. Halley’s Comet appeared in 1066 — followed by the Battle of Hastings. Case closed.
Continue reading
Enter your email to unlock this article and join the newsletter. You can unsubscribe anytime.
Common questions about this topic?
Why is the confident use of the word certainty problematic?
Because certainty is far rarer than commonly assumed. In the article, a LinkedIn post by a digital specialist who claimed A/B tests can determine true causality prompts the reflection: statistics operates in probabilities, not absolutes. Even well-designed experiments produce higher probability, not certainty. Confusing statistical significance with proof leads to overconfident decisions based on incomplete understanding.
What is the difference between correlation and causation?
Correlation means two things occur together; causation means one actually produces the other. David Hume argued in the 18th century that we never directly observe causality — we only observe that B follows A repeatedly and assume a causal link. In the article, this philosophical insight is applied to modern marketing and data analysis: statistics can show that two variables are related, but not which one causes the other.
Why does the human brain see causes where there are none?
From an evolutionary perspective, assuming causation was safer than ignoring potential threats. If a rustle in the bushes might be a predator, it is better to run and be wrong than to stay and be dead. In the article, this survival mechanism is identified as the root of a persistent cognitive bias: we instinctively look for causes in random events, see patterns in noise, and construct explanations where none exist.
Can A/B tests prove causation?
A/B tests provide stronger evidence than observational studies because they use randomisation to control for confounding variables. However, they still operate within probability — they increase confidence that a difference is real, but they do not deliver absolute certainty. In the article, the claim that A/B tests determine true causality is challenged: even experiments produce higher probability, not proof in the Newtonian sense.
What does David Hume say about causality?
Hume argued in A Treatise of Human Nature (1739) and An Enquiry Concerning Human Understanding (1748) that we never perceive causation directly. We observe constant conjunction — that B regularly follows A — and our minds create the expectation of a necessary connection. But this connection is a habit of thought, not an observed fact. In the article, this insight is applied to challenge the casual use of the word causality in data-driven fields.
Why does this matter for professionals working with data?
Because overconfident causal claims lead to wrong decisions. In the article, the concern is that professionals in marketing and data analysis use words like certainty and causality as if they were dealing with Newtonian physics — clear, absolute, unquestionable. This false confidence can result in strategies built on correlations mistaken for causes, optimisations based on incomplete understanding, and a culture where questioning assumptions is discouraged.
If you have any thoughts, questions, or feedback, feel free to drop me a message at mail@richardgolian.com.