TL;DR: The human brain is wired to detect patterns, so it often detects patterns that do not exist. This phenomenon is called apophenia, and a familiar form of it is pareidolia, seeing faces in inanimate objects. This is not a pathology but a side effect of a survival mechanism, becoming a problem only when accompanied by an unusual sense of meaning or leading to poor decisions. Research shows that people who believe in the mystical tend to report seeing faces in ambiguous stimuli more often, but most of the difference lies in a more lenient response threshold rather than in perceptual sensitivity.
Why humans see patterns where there are none: from faces in clouds to repeating numbers in life
What is apophenia?
Apophenia is the tendency to see meaningful connections between things that are not actually related. The term was coined by psychiatrist Klaus Conrad in 1958 in his study of the onset phase of schizophrenia (Conrad 1958; see also Mishara 2010), where he defined it as "the seeing of unfounded connections, accompanied by an abnormal feeling that everything has meaning".
One point worth distinguishing: apophenia is not the same as hallucination. In hallucination, the brain produces a perception with no external stimulus. In apophenia, the stimulus is real, only the interpretation is wrong. You really do see the streak of cloud, you really do see 11:11 appear on the clock, only the meaning you assign to them is the part your brain adds on its own.
Why does the brain go looking for patterns everywhere?
The brain cannot process each individual event independently, it is forced to compress information into patterns to save energy. This process is called pattern recognition, and it operates through three main mechanisms.
- Template matching: the brain compares a new stimulus with templates stored in long-term memory. The letters D, d, uppercase D, handwritten d are all recognised as the same letter. But when this mechanism is applied to complex data such as clouds or stock charts, the brain can match the wrong template and produce a false positive.
- Prototype matching: the brain does not need an exact match, only a match with the average image in the mind. Because it does not demand a perfect match, this mechanism easily leads to concluding on the basis of just a few similar features, and that is fertile ground for apophenia.
- Feature analysis: the stimulus is broken down into separate features, then compared against memory through four steps: detection, separation, comparison, identification. An error at the comparison step is enough to generate a phantom pattern.
Pareidolia: faces in clouds and on slices of bread
Pareidolia is the most familiar form of apophenia, in which people see images or sounds in random stimuli. People around the world see the "man in the moon", see a face in the headlights and grille of a car that is "smiling", see a religious face in the burn marks on a slice of bread.
The mechanism behind this phenomenon is fairly specific: the fusiform face area, the part of the brain responsible for face recognition, sometimes mistakenly processes an object with a few face-like features as an actual face (Liu et al. 2014). Liu's group recorded face pareidolia as the illusory perception of faces that do not exist, measurable even on purely noisy images. There is relatively strong evidence that hallucinogenic substances increase pareidolia. Another famous example is the "face on Mars", which NASA announced to be a natural landform turned into a face by light and the camera angle (NASA, Astronomy Picture of the Day, 6/4/1998, A Face On Mars).
Do believers in the mystical see more?
Yes, but most of the difference lies in the response threshold rather than in actual detection ability. According to research by Tapani Riekki and colleagues in 2013 published in Applied Cognitive Psychology, when 47 participants took two tests, both those with mystical beliefs and those with religious faith detected face-like areas in landscape images at a higher rate than sceptics, yet at the same time they also made more mistaken reports. When analysed using signal detection theory, the authors found that believers had a more lenient response threshold, while actual sensitivity did not differ from sceptics. This is a small sample and the results have not been widely replicated, so it should be treated as a suggestive direction rather than a firm conclusion.
Michiel van Elk (2015) in PLOS ONE conducted two studies, one at a mystical fair and one online, and also recorded that believers in the mystical tend to report seeing faces in ambiguous stimuli more often. He also examined the link between this tendency and conspiracy beliefs, and found the link to be weak and inconsistent. In the first experiment, sceptics were more cautious when having to classify an ambiguous stimulus as a face, suggesting the two groups use different internal criteria when making perceptual decisions.
Jan-Willem van Prooijen, Karen M. Douglas and Clara De Inocencio (2018) in the European Journal of Social Psychology ran five studies and found that people who see patterns in random coin-toss sequences or in abstract paintings are more likely to believe in conspiracy theories and in the supernatural. Notably, when the authors primed participants with an awareness of patterns, those beliefs also increased, meaning the link appears to run in a causal direction rather than being mere correlation.
When does seeing patterns become a problem?
At an ordinary level, apophenia is just a side effect of an efficiently working brain. At an extreme level, it is a symptom of mental disorder, for example in paranoid schizophrenia, when patients see hostile patterns in ordinary behaviour and believe there is a conspiracy against them.
Apophenia is also the typical raw material of conspiracy theories, where random coincidences are woven into a deliberate script. Psychologist Michael Shermer calls this tendency patternicity, "the tendency to find meaningful patterns in meaningless noise" (Shermer 2008).
The key difference lies here: an epiphany brings new understanding of reality, while an apophany only creates the feeling that everything revolves around oneself, repeating over and over and leading nowhere.
Why is gambling an ideal environment for apophenia?
Gamblers often think they see patterns in lottery number sequences, in card hands or on the roulette wheel, where in reality no pattern exists. The most common form is the gambler's fallacy, the belief that a result that has not appeared for a long time is about to appear.
In statistics, apophenia is an example of a type I error: falsely detecting a pattern in data. This is precisely the false positive in hypothesis testing, and it is the reason researchers must strictly control the number of hypotheses tested.
The clustering illusion and a lesson from medical data
The clustering illusion is a form of cognitive bias that makes people see patterns in a sequence of random numbers or events. Many theories have been rejected after this bias was pointed out.
One notable case is the cluster of breast cancer cases among female staff at the ABC broadcasting site in Toowong, Brisbane: 10 cases among 550 people from 1994 to 2006, a risk six times higher than that of Queensland women in general (Stewart 2007). A thorough investigation found no correlation between this unusually high rate and factors at the workplace, nor with the genetics or lifestyle of the staff. A disease cluster that looked very "patterned" turned out to be a random fluctuation in a small dataset.
Why does the brain prefer a false alarm to a miss?
Evolutionary psychologists explain apophenia through error management theory, proposed by Haselton and Buss in 2000. According to this, detecting patterns is not a flaw of the brain but the result of survival pressure across many generations.
A related experiment is Skinner's 1948 "superstition" experiment: hungry pigeons were given food regularly every 15 seconds, regardless of what they were doing (Skinner 1948). A bird happened to be performing some action when the food dropped, and instead of attributing it to randomness, it repeated that action. Six out of eight birds developed this kind of "superstitious" action. This is a form of "superstition" in animals, and it shows that apophenia has roots far deeper than human culture.
The survival logic is very simple: if you hear a noise in the bushes and ignore it, you may lose your life because it is a predator. If you run away because the noise was only wind, you lose only a little energy. The brain was selected to lean toward the false alarm, and the price paid is that it also sees patterns where there are none.
How is apophenia different from related biases?
Many cognitive biases are related to apophenia and are often confused with it. The table below distinguishes the commonly encountered concepts.
| Concept | Short definition | Example |
|---|---|---|
| Apophenia | Seeing meaningful connections between unrelated things | Believing that coincidences during the day are signs of fate |
| Pareidolia | Seeing images or sounds in random stimuli | Seeing a face in a cloud |
| Clustering illusion | Seeing patterns in a sequence of random numbers or events | The cancer cluster in Toowong, Brisbane |
| Gambler's fallacy | Believing a result that has not appeared for a long time is about to appear | Betting on red because black has come up many times |
| Confirmation bias | Seeking, interpreting and remembering information in a way that supports existing beliefs | Reading only news that reinforces one's own views |
| Synchronicity | A subjective correlation between two events, without any claim of a causal relationship | Thinking of someone and then that person calls |
Confirmation bias: apophenia's companion
Confirmation bias is the tendency to seek, interpret, favour and remember information in a way that reinforces existing beliefs, values or decisions. It is the missing piece for understanding why a phantom pattern lives so long: apophenia creates the pattern, and confirmation bias keeps that pattern from being broken by contrary data.
This bias operates through three channels: biased information seeking, biased interpretation and biased memory. These three channels explain four specific effects:
- Attitude polarisation: disagreement becomes sharper even when the sides are exposed to the same evidence.
- Belief perseverance: a belief survives even after the evidence for it has been proven wrong.
- Primacy effect: relying more heavily on information encountered early in a sequence.
- Illusory correlation: seeing a connection between two events that have no connection.
An experiment at Stanford University clearly showed this mechanism. Supporters and opponents of the death penalty read two fake studies on the deterrent effect of capital punishment. Initially both groups shifted slightly in the direction of the study they read first. But after reading the detailed descriptions, each side found the study agreeing with them more convincing, and the overall result was that the two sides moved further apart (Lord, Ross & Lepper 1979). Both supporters and opponents rated the study supporting them as having better methodology, with very specific arguments.
How does the digital environment strengthen apophenia?
On social media, confirmation bias is amplified by filter bubbles and "algorithmic editing", so that each person is shown only information they are likely to agree with, while opposing views are filtered out. When the information source has already been filtered to match existing beliefs, apophenia has more data with which to weave ever tighter patterns that are less and less challenged.
What to do about this tendency?
Apophenia cannot be switched off, and it does not need to be. Schools of psychology do not fully agree on where to draw the line between "healthy intuition" and "phantom pattern", but there are a few fairly universal principles.
- Ask yourself whether the data is large enough. A cluster of three events in one week says nothing. Random fluctuation in a small sample looks very much like a pattern.
- Look for disconfirming evidence, not only supporting evidence. Humans tend to test in one direction only, seeking out cases that match the hypothesis. Asking the reverse question is the cheapest way to detect a phantom pattern.
- Distinguish the feeling of meaning from evidence. The feeling that an event is "too coincidental to be random" is a psychological state, not an argument.
- Pay attention to emotion. Confirmation bias is strongest for outcomes we desire, emotionally charged issues and deeply ingrained beliefs. That is when the most caution is needed.
In short, apophenia is the price paid for a brain designed to survive in a dangerous environment. It helped our ancestors react quickly to threats, and it also makes us see faces in clouds, repeating numbers on clocks, and patterns in random data. Knowing the mechanism does not make it disappear, but it helps us pause for one beat before turning a coincidence into a destiny.
