Taken from above, this annotated image shows a school ground in Spain and the motions of teenage participants as indicated by the dots and lines. Credit: 2026 Echeverría-Huarte et al.
Put a small crowd of people in an open space and ask them to walk around, and something odd happens.
They do not move as randomly as you might think. Again and again, in experiments in Spain and Japan, groups of pedestrians began drifting counterclockwise, as if an invisible current had started to pull them around the room.
The effect was consistent time and time again. In one early set of trials, 32 out of 33 groups showed the same counterclockwise tendency.
“When analyzing the experiments, my colleagues realized by chance that in 32 out of 33 experimental trials, as people moved and turned, they noticeably preferred to turn counterclockwise,” said Claudio Feliciani, a project associate professor at the University of Tokyo.
It seems like humans may have a small but persistent turning bias in the way they walk. The researchers tested whether culture, gender, handedness, footedness, eye dominance, crowd size or walls could explain it. Most did not.
A Crowd Pattern that Starts with the Individual
Scientists who study crowds have long known that pedestrians self-organize. On a sidewalk or in a station corridor, people do not usually collide at random. They read one another’s motion, adjust their steps and form temporary patterns, often without noticing. Researchers sometimes describe this as a kind of “social force”. It’s not a real force like gravity, but a useful way to model how people steer away from one another.
Unlike particles, however, people anticipate trajectories. Studies of moving crowds have shown that pedestrians often change course not simply because someone is close, but because a collision may happen soon, and they do so over a range of walking speeds.
At low densities, this produces order. Two streams of pedestrians can spontaneously sort themselves into lanes, each moving in one direction, much as traffic finds a smoother path when drivers stop weaving. But the same physics can turn dangerous when space runs out.
×
Thank you! One more thing…
Please check your inbox and confirm your subscription.
With about two people per square meter, walking slows. At four, bodies begin to touch involuntarily. At six or seven, movement becomes difficult. When densities get higher, a crowd acts more like a deformable material, transmitting pressure through bodies. Disasters such as the 2022 Itaewon crowd crush in Seoul, resulting in the deaths of 159 people during Halloween celebrations, showed how quickly a dense crowd in a narrow space can turn into a stampede.
That is why crowd science now draws from physics, psychology, physiology and computer modeling. Researchers want to know when a crowd behaves like a stream, when it behaves like “soft matter,” and when mass emotion or distraction can change its motion.
The new counterclockwise study does not examine social forces like panic. Instead, it asks: before crowds become dangerous, before signs and walls and urgency take over, do individual walkers already carry small biases that shape the group?
The researchers first placed groups in a circular arena in Spain. Some groups included people who had shown a preference for turning right when reaching a wall. Others included left-handed and left-footed participants. If turning preference or body dominance explained the effect, the direction of the group should have changed.
It did not. The counterclockwise pattern persisted.
“This was completely unexpected as, at least instinctively, when people walk around randomly, you imagine people turn as their needs suit them, with little sign of an overall preference,” Feliciani said. “But there was a definite, measurable tendency for people to turn counterclockwise over clockwise, all things being equal.”
The team then tried to remove the walls from the equation. In a 50-by-60-meter schoolyard in Spain, more than 100 teenagers walked in a mostly unconstrained space. The bias remained.
Next came Japan, where pedestrian habits can differ from those in Europe. In some Japanese settings, people tend to pass or form lanes on the left rather than the right. If the counterclockwise motion came from learned crowd etiquette, the Japanese trials might have looked different.
They did not. Groups in Japan also tended to rotate counterclockwise.
Not Culture, Not Handedness
The researchers kept narrowing the possibilities.
They analyzed children in a Japanese nursery school during a free-running activity. The counterclockwise motion appeared even more strongly there. The children, about 5 years old, often moved in a stable vortex-like pattern, with most of the group circling in the same direction.
Young children have had less time to absorb adult conventions about walking routes, sports tracks or public-space etiquette. It does not prove the bias is innate, but it weakens the case that culture alone explains it.
“Of all these things, the only thing that stood out was that kids tend to have a stronger bias for the counterclockwise direction, so probably age plays a role in making the effect weaker or stronger,” Feliciani said in the University of Tokyo press release.
The researchers also surveyed 168 people in Spain to see whether an unwritten social rule might exist. The responses did not reveal a clear norm favoring counterclockwise movement. If anything, many respondents expected clockwise motion.
Then came a more direct test. More than 200 people walked alone, one at a time, inside an enclosure. Their paths still showed a counterclockwise bias.
This latter result is perhaps the most telling. It suggests the effect does not require a crowd. A crowd can make the pattern easier to see, but the source appears to lie in individual movement.
The team also tested whether common body asymmetries explained the effect. Right-handed and left-handed participants did not differ significantly. Similarly, no difference was seen with right-footed and left-footed participants, people with different eye dominance, men and women, or participants who walked with one eye patched.
Previously, researchers studying crowds at the San Fermín festival in Spain found that they spontaneously synchronize into collective oscillations, with hundreds of people moving in coordinated, swirling patterns — like a human vortex.
Bartolo lab, ENS de Lyon
A Small Bias with Practical Consequences
The cause remains unknown.
“It likely does not come from the eyes, because we tried to patch people’s left or right eyes and the bias was still there,” Feliciani said. “And some people asked us if it might be large-scale phenomena like the Coriolis force or Earth’s magnetic field, but this seems unlikely given what we have managed to point to so far.”
The authors suggest that subtle biomechanical or neurological asymmetries may be involved. Previous work has shown that blindfolded people can drift in circles when trying to walk straight, possibly because of small errors in balance or body perception. But the new experiments involved people who could see normally, so the answer is probably more complicated.
The finding could eventually matter for how people design public spaces. Museums, airports, train stations, stadium forecourts and shopping centers all depend on smooth pedestrian flow. If people naturally prefer certain circulation patterns, designers might one day use that tendency to reduce friction and improve comfort. Perhaps the ideal public space is less a long corridor or square and more like a spiral.
“Our results may appear to be a minor, insignificant discovery, but in nature, most phenomena related to locomotion show that animals mostly walk without directional preference,” Feliciani said. “The strong bias found in people hints at some asymmetry at the biomechanical level.”
The team now plans more detailed experiments with individuals to hunt for the source of this asymmetry. The answer may sit somewhere in the body’s machinery for balance, stride, attention and motion.
Until then, the next time a crowd begins to drift left, just go with the flow. It’s just human, after all.
The findings appeared in Nature Communications.