Imagine standing in a crowded festival where thousands of people are packed tightly together. Suddenly, without anyone giving instructions, the crowd begins moving in a coordinated circular pattern. People sway, rotate and push against one another as if they are parts of a giant machine. Nobody is controlling the movement, yet hundreds of people begin moving together.
This is not a scene from a movie. Scientists have discovered that dense crowds can spontaneously develop these unusual collective movements, which may become dangerous within seconds.
In a study led by François Gu and his research team, scientists analysed the movement of thousands of people at Spain's San Fermín festival. Their findings reveal that crowded groups can develop large-scale circular oscillations without any external instructions or coordination. More importantly, the researchers developed a physical model that explains why these movements emerge and how monitoring them could help identify dangerous crowd conditions before a disaster occurs.
The discovery could improve our understanding of crowd safety and provide a new approach to preventing deadly incidents at festivals, concerts and other mass gatherings.
Why Dense Crowds Can Become Extremely Dangerous
Large gatherings are common around the world. Religious celebrations, music festivals, sporting events and public celebrations regularly attract thousands of people into limited spaces.
Although these events can be exciting, they can also create serious safety risks when too many people occupy a confined area.
When a crowd becomes extremely dense, individuals may lose the freedom to control their own movements. Pressure from surrounding people can push them forward or sideways, even when they do not want to move. This can lead to dangerous crowd surges, compression against walls and barriers, breathing difficulties and fatal injuries.
These incidents are often described as crowd crushes or crowd disasters. They do not necessarily begin with panic or deliberate pushing. Sometimes, the collective movement of people creates dangerous forces that no individual can control.
Despite the seriousness of these incidents, understanding how thousands of people interact and move together has remained a major scientific challenge.
Why Traditional Crowd Models Have Limitations
For decades, researchers have used mathematical models to understand pedestrian movement. These models typically examine how individuals respond to nearby people, obstacles and available space.
For example, a person walking through a busy corridor may slow down, change direction or move aside to avoid colliding with someone else. Mathematical models can reproduce such behaviour and help engineers design safer buildings and evacuation routes.
However, these approaches become more complicated when thousands of people are packed closely together.
In a dense crowd, a person's movement may depend not only on their immediate neighbours but also on pressure and motion spreading across a much larger group. Small movements can influence surrounding individuals, producing collective patterns that are difficult to predict by examining people individually.
Researchers therefore need a different way to understand these situations.
The new study takes inspiration from fluid physics. When scientists investigate the formation of a vortex in a cup of tea, they do not need to track every molecule of water. Instead, they use the general laws of fluid motion to explain the behaviour of the liquid as a whole.
Similarly, the researchers suggest that it may be possible to understand dense crowds by studying their overall physical behaviour rather than trying to predict every individual's decisions.
Thousands of People Can Begin Moving Together
To investigate this phenomenon, the researchers analysed dense crowds at Spain's San Fermín festival, where large numbers of people gather in confined spaces.
By repeatedly studying crowd movements, they discovered that dense crowds do not always behave randomly. Under certain conditions, they can develop organised, repeating patterns of motion.
Hundreds of people may begin moving in coordinated circular or orbital patterns, even though nobody is directing them to do so.
These movements are known as collective chiral oscillations.
The term may sound complicated, but its meaning is relatively simple. Collective means that many people move together. Oscillation refers to a repeating movement, while chiral describes a pattern with a preferred rotational direction, such as clockwise or anticlockwise motion.
In other words, a crowd can begin behaving like a large, coordinated rotating system.
The surprising part is that this organisation can emerge naturally from the physical interactions within the crowd. People do not need to agree on a direction or consciously coordinate their movements.
However, when a crowd is tightly packed, this collective motion can become dangerous. Individuals may be carried along by surrounding movement, making it difficult to regain control of their position.
The Science Behind These Unexpected Movements
To explain their observations, the researchers developed a mechanical model based on measurements and fundamental physical principles.
Their model identifies a key role for unusual frictional effects called odd friction.
In everyday situations, friction generally acts against relative motion. For example, when an object slides across a surface, friction usually opposes its movement.
Odd friction describes a different kind of response that can occur in certain systems that operate away from equilibrium. In these systems, forces and motion can be related in ways that are not captured by ordinary friction models.
According to the researchers, this unusual frictional behaviour helps explain how a densely packed crowd can develop coordinated rotational motion.
Confinement also plays an important role. When thousands of people occupy a limited space, their movements become strongly connected. As density increases, the interactions between individuals can create a collective pattern that would not appear in a more loosely packed group.
The researchers found that beyond a critical crowd density, the system can undergo a transition into a state of collective oscillation.
This is known as a non-reciprocal phase transition. In simple terms, the crowd changes from one type of collective behaviour to another, developing organised motion that cannot be adequately explained by conventional models of individual pedestrian interactions.
The model reproduced the main patterns observed in the researchers' measurements, offering a physical explanation for how seemingly unpredictable crowd movements can emerge.
A Connection to the 2010 Love Parade Disaster
One of the most important aspects of the study is its connection to the 2010 Love Parade disaster in Duisburg, Germany.
The electronic music festival ended in tragedy when a crowd crush in a confined access area resulted in 21 deaths and hundreds of injuries.
The researchers compared their findings with crowd movements observed at the onset of the disaster. They found similarities between the collective oscillations identified in their study and the dangerous movements recorded during the Love Parade incident.
This comparison suggests that the same underlying physical principles may help explain crowd behaviour in both controlled observations and real-world emergencies.
However, the findings do not mean that every circular crowd movement will lead to a disaster. Rather, they show that certain collective patterns may provide valuable information about changing crowd conditions.
Understanding these patterns could help safety teams recognise when a crowd is approaching a dangerous state.
Could This Discovery Help Prevent Future Crowd Disasters?
The researchers propose a strategy for monitoring crowd movement in real time.
Instead of focusing only on the number of people in an area, safety teams could also analyse how the crowd moves. Video footage and computer-based motion analysis could potentially help identify unusual, coordinated oscillations that signal increasing instability.
If such patterns can be detected reliably, authorities may be able to intervene before the situation becomes critical.
Possible responses could include controlling the entry of additional people, reducing pressure at bottlenecks, opening alternative routes where safe and directing people away from overcrowded areas.
However, further research and real-world validation would be needed before these methods could be relied upon as operational warning systems. Detecting a pattern is only one part of the challenge; translating it into timely, effective safety decisions is equally important.
A New Way to Understand Human Crowds
The study offers a fresh perspective on one of the most difficult problems in public safety. It suggests that dangerous crowd behaviour is not always random or impossible to understand. Instead, large groups can develop physical patterns that follow measurable principles.
These findings may also have broader scientific applications. Similar approaches could help researchers investigate collective movement in animal groups, living tissues and other systems made up of many interacting units.
Ultimately, the research highlights an important lesson: when large numbers of individuals interact in a confined space, the behaviour of the whole group can become very different from the behaviour of any single person.
By learning to recognise these collective patterns, scientists may help authorities better understand when a crowd is becoming dangerous—and potentially prevent future tragedies before they unfold.
Reference: Gu, F., Guiselin, B., Bain, N. et al. Emergence of collective oscillations in massive human crowds. Nature 638, 112–119 (2025). https://doi.org/10.1038/s41586-024-08514-6
Comments
Post a Comment