What can science tell us about human stampedes?

It's hard to imagine how a crowd - especially one gathered for a peaceful religious rite - can turn so deadly. But the phenomenon is so common that experts in crowd management are consulted for most highly trafficked events.

While many researchers focus on how to prevent these so-called stampedes by keeping spaces from getting overcrowded, there's been very little research on what happens once a stampede starts - or why, exactly, they begin.

Deadly stampedes have marred political rallies, music festivals, sporting events and religious gatherings all over the world.

In 1989, in one of the worst tragedies in soccer history, nearly 100 people died in Sheffield, England, as crowds pressed into Hillsborough Stadium for a match.

In 2005, a stampede on a bridge across the Tigris River in Baghdad killed more than 960 people when rumors about a coming suicide bomb attack caused panic among pilgrims heading toward a shrine. At the time, it was the greatest single-day casualty toll since the U.S.-led invasion of Iraq two years earlier.

And in 2010, a holiday celebration in Phnom Penh, Cambodia, left at least 353 people trampled to death after a suspension bridge began swaying and thousands of revelers tried to flee.

That same year, a review of all available literature on stampedes found that, despite efforts, these incidents are on the rise. But the researchers noted that little was known about the actual triggers for these events. First responders, they noted, were rightly focused on finding and treating the injured, not on taking detailed notes of their observations of the stampede.

"International health organizations have to recognize that this is an important type of disaster," Edbert Hsu, associate professor of emergency medicine at Johns Hopkins University, said in a statement at the time. "If they made it a protocol to send someone to a trampling disaster quickly to see what happened, we would have detailed reports we could use to compare and contrast. Without [those reports], we won't really understand what we're dealing with."

In an email to The Washington Post, Keith Still, a crowd safety and risk analysis specialist, explained that actual human "stampedes" are rarely observed. The incidents that members of the public refer to as stampedes are quite different from the animal equivalent, and most could be more accurately referred to as crowd crushes.

"What this appears to be is a crowd crush as a result of two-way flow in confined space," Still explained. A compression - not a stampede.

Mechanically, these crowd crushes are tragically simplistic: Once people are pushed tightly against one another (about seven people per 10 square feet of space, according to one study) it's vital that those in the front keep moving as quickly as those behind them. Otherwise, the people in the back - unable to see the front of the crowd - will move forward seeking more space, assuming that those in the front will continue to move to make way for them.

If for some reason the paces become mismatched - because something is blocking the front of the group, or a rumor is spreading in the back that people are being crushed, causing folks to speed up - the front of the group gets squeezed, sometimes producing enough force to crush people where they stand.

It's likely that the most deadly crowd crushes begin with one or a handful of deaths, caused by the sheer force of the tightly packed group, that then cause mass panic. Smaller crushes may not be fueled by "panic" at all - people can be crushed by the weight of those around them without anyone consciously surging forward.

In fact, experts have argued that blaming the behavior of "the crowd" is a mistake, since most crushes can probably be boiled down to the physical limits of their location. Even crushes or stampedes that occur at rowdier events, such as soccer matches, Black Friday sales or music festivals, are more likely due to physical strain than they are to any specific human behavior.

Still, who previously studied the specific crowd movements during the pilgrimage to Makkah in order to provide crowd-management guidance, said it was difficult to identify what exactly happened to the crowd in nearby Mina on Thursday. But he doesn't think there was necessarily a great panic.

This incident "sounds like a compression from two opposing flow rates exceeding the safe capacity of the system," Still explained. "Once that starts, it's already too late to stop the incident escalating."

But when people do panic, that certainly doesn't help. In a tightly packed crowd, we're victims of our own biology. The typical "fight or flight" response, where one feels a surge of adrenaline, is anything but helpful. If the thousands of people vying for space could be calm and collected, crowd crushes of this magnitude simply wouldn't occur. But faced with death, most will be slaves to a racing heartbeat and hyperventilation - and an urge to run for safety at any cost.

To prevent crushes like the one that happened Thursday, Still and researchers like him work on predicting the movement that a crowd will want to take. "It's all about math, management and psychology," he told The Post in an earlier interview on crowd management. When individuals are participating in a religious ceremony, crowd managers need to bear in mind the speed and direction that they are likely to move, based on the goal of the gathering.

"Unless you can facilitate that, you end up creating behaviors that are frustrated," he said. And those frustrated behaviors can send shock waves through a tightly packed crowd.