The Magic Octagonal Coin: Why Only 9% of People Can Predict Its Flip?

An ordinary-looking octagonal coin has an arrow pointing up on its front face and an arrow pointing left on its back face—the two at a right angle to each other. When you rotate it 45 degrees clockwise and flip it around its axis, where will the arrow on the back point? Most people confidently give the wrong answer—and that is precisely the magic of the "Magic Octagon." According to Raymond Hall, founder of the YouTube channel Physics Fun, only about 9% of people can correctly predict its behavior, and scientists still do not fully understand the cognitive mechanism behind it.

Rules of the Magic Octagon and Common Misconceptions

The rules of the experiment are simple: the front of the coin has an arrow pointing up, and the back has an arrow pointing left. The operation has two steps—first flip around the Y axis (front-to-back flip), then rotate 45 degrees around the Z axis, and finally flip around the Y axis once more. In the end, the arrow on the back points in a direction opposite to intuition.

According to the video demonstration, after rotating 45 degrees clockwise and flipping, the arrows on the two faces point in opposite directions; but after rotating 45 degrees counterclockwise and flipping, the arrows on the two faces point in the same direction. This asymmetry catches most people off guard. Raymond Hall, in collaboration with Lauren Lacks of Fresno State University, found that the most common errors occur at the first or second vertical flip. Participants often overlook that the coin has thickness—it is actually an octagonal prism, not a two-dimensional planar figure.

This thickness causes the change in arrow direction during flipping to exceed the intuitive range of everyday spatial reasoning. Preliminary research suggests that humans have clear limitations when tracking multiple spatial transformations—here, three transformations must be tracked simultaneously: the initial Y-axis flip, the 45-degree Z-axis rotation, and another Y-axis flip. The brain's "spatial visualization" ability is prone to overload in such multi-step operations.

Octagonal prism is the geometric essence of the coin in this article: it is not a two-dimensional octagonal plane, but a three-dimensional prism with thickness, whose cross-section is a regular octagon. This means the coin has multiple faces (front, back, and eight side faces), and each face has a clear orientation.

In the flipping operation, the thickness of the prism is crucial. When the coin flips around the Y axis (front-to-back flip), the plane on which the arrow lies moves from the front to the back, but because of the thickness, the actual orientation of the arrow in space changes rather than simply being mirrored. More importantly, the superposition effect of multiple rotations causes the direction changes not to obey the commutative law—for example, flipping first and then rotating yields a different result from rotating first and then flipping. This explains why clockwise and counterclockwise rotation followed by flipping produce different results: the direction of rotation changes the relative relationship between the prism's faces and the flip axis.

Understanding the thickness of the prism is the first step to correcting intuitive errors. In daily life, we often regard coins as thin slices, but in this experiment, thickness is precisely the key that breaks two-dimensional intuition.

A Cognitive Science Perspective: A New Explanation from Relevance Theory

In addition to the limitations of spatial reasoning, the video's author proposes a more ambitious hypothesis: Dan Sperber and Deirdre Wilson's "Relevance Theory" may explain this phenomenon. The theory holds that human communication and cognition prioritize information most relevant to the current situation. When the coin is described as "up," "left," and "45 degrees," we tend to focus only on its appearance from the front view and ignore the changes brought about by rotation.

To test this hypothesis, the author designed a controlled experiment: instead of directly saying "up," "left," and "45 degrees," describe which corner the arrow points to. For example, the front arrow points to one corner, and the back arrow points to another corner. If people think about the coin in terms of "parts" rather than "views," can they better track its motion?

This approach shifts the question from "the brain's insufficient spatial ability" to "how we decide what information is worth paying attention to." If Relevance Theory holds, then changing the way the coin is described may significantly improve prediction accuracy. Currently, this experiment has no public results, but this direction offers a new entry point for cognitive science—the illusion may arise not because the brain is not strong enough, but because it is too good at filtering information.

Relevance Theory was proposed in 1986 by French anthropologist Dan Sperber and British linguist Deirdre Wilson, originally to explain pragmatic reasoning in human communication. Its core idea is that the human cognitive system instinctively seeks "maximum relevance"—that is, obtaining the greatest cognitive effect with the least cognitive effort. In information processing, we automatically filter out the cues most relevant to the current situation and ignore other information that may be equally important but appears secondary.

In the Magic Octagon scenario, when the coin is described as "arrow pointing up" and "arrow pointing left," these directional words activate our two-dimensional representation based on the front view. Relevance Theory predicts that the brain will prioritize this explicit directional information and ignore the implicit spatial transformations brought about by the coin's thickness and multiple rotations. This is not because the brain lacks ability, but because it efficiently follows the principle of "maximizing relevance"—except that in this particular task, that principle leads to systematic bias.

This theory provides a new perspective for understanding visual illusions and spatial reasoning errors: illusions may stem from the cognitive system's misjudgment of information relevance rather than a mere deficit in spatial ability.

Multi-Source Cross-Verification and Unsolved Mysteries

According to analyses by multiple science channels on Bilibili and YouTube, the core mechanism of the Magic Octagon has been consistently confirmed: the coin's thickness and the superposition effect of multiple rotations are the main causes of intuitive bias. The sources largely agree on the figure that "only 9% of people can correctly predict it," but they do not disclose the specific experimental sample size and statistical methods in detail.

The point of divergence lies in the explanatory framework. Some cognitive psychologists favor individual differences in spatial reasoning ability, while the Relevance Theory proposed by the video's author remains a single-source hypothesis lacking peer-reviewed support. Lauren Lacks's team research is still ongoing, and a complete paper has not yet been published. Therefore, Relevance Theory can currently only serve as an exploratory viewpoint and cannot be regarded as conclusive.

Regardless, the Magic Octagon reminds us that everyday intuition is unreliable when it comes to three-dimensional rotations and multiple transformations. It is both an interesting desktop toy and an ingenious experimental tool in cognitive science. As the trivia at the end of the video notes—it is actually not an octagon, but an octagonal prism—our understanding of the world often differs by that little bit of "thickness."

Peer review is the core mechanism in scientific research for ensuring the reliability of conclusions. Before a study is formally published, it must undergo anonymous review by experts in the same field, who evaluate its experimental design, data analysis, and the reasonableness of its conclusions. Hypotheses that have not undergone peer review (such as Relevance Theory in this article) are usually regarded as exploratory viewpoints with limited evidentiary strength and cannot be cited as scientific conclusions.

The Lauren Lacks team research mentioned in this article has not yet published a complete paper, which means the relevant experimental data and methodological details are not public, and other researchers cannot reproduce or verify them. Therefore, Relevance Theory can currently only serve as a heuristic hypothesis and must await subsequent research (such as preregistered experiments and multi-laboratory replication) for testing. In science communication, it is crucial to distinguish "confirmed conclusions" from "hypotheses awaiting verification" and avoid conflating the former with the latter.

One-line Summary

The Magic Octagon coin flip fools 91% of people; thickness and multiple rotations cause intuitive bias, and Relevance Theory offers a new explanation.

Paragraph Summary

The Magic Octagon is a cognitive illusion experiment: an octagonal prism coin with a front arrow pointing up and a back arrow pointing left, after flipping and rotation, only about 9% of people can correctly predict the direction of the back arrow. Common errors stem from ignoring the coin's thickness (it is actually an octagonal prism) and the superposition effect of multi-step spatial transformations. Researchers propose the Relevance Theory hypothesis, suggesting that the brain prioritizes explicit information such as directional words while ignoring rotational details, and that changing the way the coin is described may improve accuracy. Currently, this theory remains a single-source hypothesis lacking peer-reviewed support. The phenomenon reveals the limitations of human spatial reasoning and also suggests that illusions may stem from the cognitive system's efficient filtering rather than a lack of ability.