The physics of crowds can kill

Almost two years ago, 21 people died when they were crushed to death in the crowd at the Love Parade music festival in Germany. Now, scientists have been able to pinpoint exactly what lead to those deaths. Here's a hint: It wasn't a stampede, there's no evidence of intentional pushing, and it doesn't look like mass hysteria had anything to do with the deaths. So how did those 21 people die? Physics. (Via Jennifer Ouellette)


    1. – 1883, and everyone involved were children. Slightly different, in that they were moving down a staircase and hit a door that only opened inwards – but still. Incidentally, that’s the disaster that led to panic bars and doors having to open outwards.

  1. This is really interesting. My tendency is to just blame stupid people when something like this happens. Now I can blame stupid physics.

  2. Helbing and Mukerjee say most deaths occurred elsewhere in the crowd because of a phenomenon known as crowd quakes. These occur when the density of a crowd becomes so great that individuals are forced into bodily contact with each other. When this happens, the forces are transmitted through the crowd in chains  from one body to the next. Crucially, when this happens, the transmitted forces add up.

    ok so the victims’ bodies had broken ribs?  or at least the cause of death was asphyxiation?  that is, what did the bodies themselves tell us about the forces caused by these theorized “crowd quakes”?

    1. I think the very next section addressed that point. And they didn’t gather this from autopsies or examining injuries. They came to this conclusion by examining crowd movements and reactions by watching video.

      The point is it wasn’t stampeding that crushed people as they were trodden underfoot. It was a cascade of kinetic energy from one body to another in sections of the crowd where there were many people forced into contact by the dynamics of the crowd. It built to a point that it started to knock people off their feet. This caused a domino effect of people landing on people. The one’s on the bottom get crushed/asphyxiated. I would imagine that any injuries caused by the energy transfer would be incidental next to the gross damage caused by getting crushed in a scrum pile.

      At least that’s how I understood it.

      1. that article seemed to indicate there was a blend of insufficiently-deployed crowd control and line-cutters.

        the Who disaster was a case of an easily-identifiable rush of people in stadium seating– if there had been assigned ticketed/numbered seats, then no rush; so the line cutting in this latest case was the primary rush, then the bottleneck, then an unreasonable press of herded humans.

        this is one of the reasons i don’t even like shopping malls– you can smell the cattle stampede potential in the air.

  3. This may lead to some interesting research in crowdspace architecture. I forsee quake-breaking formations that prevent force chains reaching dangerous levels, installed or designed-in around likely risk points.

    1.  I saw a video on getting people out of an enclosed space in case of fire and preventing fatal pile-ups around doors. The researchers concluded that positioning a pole about four feet in front of the door allowed bodies to flow out the door safely instead of crushing each other there.

  4. They have a link in the article to bystander videos demonstrating their findings, but I don’t dare click on it.  Too triggering.  I was in such a crowd once.  It’s possibly the scariest 45 minutes of my life, and I have NOT lead a sheltered life.  I could feel the force they’re talking about.  For many minutes at a time, my feet were not on the ground.  And yet, there was no anger in the crowd, or overt panic: people were scared, no question, but they weren’t pushing….if anything, everyone was trying to do everything they could to calm themselves and those around them in an effort to quell the force we were feeling all around us.  We knew, instinctively, that we were in a life-threatening situation despite no active triggers except for the sheer numbers of people and lack of exit points.

    *whew*  I wasn’t expecting to flashback to that day just now.  Yeah, you don’t ever want to be in that situation.

    1. I experienced something like this during Halloween in the Castro in San Francisco several years ago. The closer to the intersection of 18th and Castro you got, the more frequently the crowd completely pressed in all around you and you lost control of where you could move.

      I got out as quickly as I could. At least there it was caused by massive amounts of people converging on the intersection of 18th and Castro and not being boxed in with no exits.

  5. I suppose WAVES start moving through such a crowd, and the characteristic of standing waves is nodes & antinodes .. regions of high and low pressure. So yes, I guess high pressure regions exist in a crowd undergoing standing wave patterns. Very scary! The trick will be to work out the resonance condition … when does a crowd start behaving like a wave?

  6. It was the dangerous overcrowding caused by the extra letter in the word “led.” 


  7. Is this not basically the same thing that happened at one of the “Stoning the Devil” disasters? I seem to recall a fairly in-depth study of that one, that came up with some flow density and flow dynamics pictures very similar to the Love Parade one in the linked article.

    As I recall it, the ingress route had more capacity than the egress route, which was furtherr hampered by a sharp corner. Crowd density built up around the hole in the bridge in fairly predictable ways, if you’ve ever watched water flowing down a stony stream.

  8. This happened to me in Melbourne once, on the Swanston Street bridge on new years eve at about 11 PM. There were so many people on that bridge that none of us could move a single step. I felt that if a few people started to fall then we would all go like dominoes and be crushed.

  9. I don’t say this research is without merit, but even without these new insights that disaster should not have happened. This is about politics and incompetence as much as it is about physics. There are rules how to set up large events, both technical and legal, and they were broken deliberately by the organizers and the authorities, with criminal disregard for human life. The city chief of police recognized the setup as unsafe in advance but was sacked by the mayor. After the disaster, the mayor refused to step down until he was voted out of office two years later. The rules were probably sufficient, but they were broken.

    1. He couldn’t step down because he would have been utterly wiped out. 

      And while I think that he was an ass to spearhead the love parade, he was a spearhead, carried by the momentum of both politics and the press.

      When the catastrophe happened, I’ve started scouring the press for the alleged “we told you you” articles they supposedly published well in advance.  

      Didn’t find a whole lot but for *readers’ comments*, who warned about the area and especially the tunnel. 

      Also, IIRC there were quite a few problems with the setup police and fire services provided. How to implement crowd control and security was their job and no, politics in Germany aren’t so corrupt that they couldn’t have put their foot down. Witness, for example, the  fire inspector at the new Berlin airport. 

Comments are closed.