At a packed concert or festival, the crowd might feel like one big, happy mass of people enjoying the same moment. But scientists at McGill University in Montreal have discovered something that could change how organizers keep those crowds safe: dangerous crushes don't happen out of nowhere. They build up slowly, as tiny pushes between people gradually spread and sync up across the crowd.
The researchers, led by Shanwei Liu, a visiting Ph.D. student in the Department of Civil Engineering, developed a computer model to simulate what happens when lots of people crowd into tight spaces. Their findings, published in the journal Chaos, Solitons & Fractals, show that when someone bumps or pushes a neighbor, that motion can spread to others nearby. Over time, people start unconsciously copying each other's pushing, creating waves of synchronized movement that can turn dangerous.
"These incidents are often not truly 'sudden'; they may begin with small pushes between individuals and gradually escalate through ongoing interactions," Liu explained.
One of the most surprising discoveries was where the highest risk occurs. Many people assume the danger is highest right at the front of an exit, where everyone is trying to leave. But the model showed the most dangerous spots are actually along the sides of exits, where physical contact builds up as people funnel toward the door. That's a hidden danger zone that event organizers might otherwise overlook.
Unlike older theories that compared falling crowds to dominoes toppling in a line, this new model treats people as wobbly bodies constantly trying to stay upright. When one person struggles to keep their balance, that instability ripples through the crowd in more complex ways.
The researchers have also created a mathematical formula that planners can use to estimate safer ways to move large groups of people. Their recommendations include spacing people out more deliberately, using physical barriers to control flow, letting people enter and leave in phases rather than all at once, and positioning security staff to catch trouble spots before they escalate.
"We hope to better understand how rescue efforts and evacuations influence one another and explore safer management strategies," Liu said. The team also plans to study how emergency responders moving against the flow of an evacuation might create new risks — and how to avoid them.
Jiangbo Yu, an assistant professor in Civil Engineering who oversaw the project, said the findings might even apply beyond crowds. The team is curious whether similar patterns of small disruptions spreading and syncing up could help explain financial crashes or political polarization. But for now, the focus is on keeping concert-goers, sports fans, and festival visitors safe — proving that understanding how people move together can help us protect them.