science physics mosh_pits music heavy_metal computer_simulation computer_modeling building computing behavior
Science of Mosh Pits May Mean Safer Crowds, Better Game and Movie Graphics

image

by Michael Keller

The casual observer of talks being given at this month’s American Physical Society meeting, happening this week in Baltimore, can be excused if the first association he or she makes isn’t with a freewheeling heavy metal concert.

Yesterday, March 20, saw a session on spin multiplets and plasmon satellites in photoemission spectras, a talk on block versus stripy antiferromagnetism in Fe-based spin-ladder materials, and new crystal structures in hexagonal CuInS2 nanocrystals, among other indecipherable—but probably very interesting—subjects.

Today, though, the physicists meeting at the Hilton Baltimore Holiday Ballroom 1 will be getting all riled up.  From exactly 5:06 p.m. to 5:18 p.m., the sky will split open, the heavy metal gods will descend, attendees will thrash and gnash like tomorrow will never come, and all will be right with the world.

That is, if the Georgia Tech team scheduled from 4:54 to 5:06 keep their talk on ants engineering habitats that reduce locomotion control requirements on schedule.

Cornell University physics graduate students will be presenting “Mosh pits and circle pits: Collective motion at heavy metal concerts,” their work to understand how people pressed into close proximity during these shows bounce around in the chaos.

“Humans are complex sentient beings,” says graduate student Jesse Silverberg, who has been going to metal concerts for 13 years and is a coauthor of the study. “When large groups get together, a lot of that complexity gets washed away by basic rules of human collective behavior.”

Though their study is founded on the entertainment enjoyed by a small segment of society, the model they’ve created from it might help planners channel the chaotic flow of protests and other public events to prevent injury. It might also help builders and concert organizers design structures that can better handle panicked crowds trying to escape buildings.

“You can’t reproduce riots to study just for the sake of science,” Silverberg says. “But we now have a consistent lens to explore what happens in those extreme conditions.”

Birth of an idea

Matt Bierbaum, another physics graduate student and coauthor of the research, guesses he’s been to 20 metal shows in the last six years.

His inspiration to study mosh pits came, he says, when he took his girlfriend to a show some five years ago. He would normally throw himself into the pit, sending his body slamming into other anonymous moshers in a violent and ecstatic ritual that would appear strange and perhaps sinister to the uninitiated.  This time, though, he stayed back to make sure his girlfriend stayed safe inside the pit.

“Being on the outside for the first time, I was absolutely amazed at what I saw — there were all sorts of collective behaviors emerging that I never would have noticed from the inside,” he says.

Silverberg and Bierbaum went looking to study motion in mosh pits at live shows around Ithaca, N.Y., where Cornell is located. But crowds were too small, so they turned to Youtube. They downloaded videos of mosh and circle pits, where people follow those in front of them to create large swirling vortexes, and wrote programs to quantify the motion frame by frame. “By recording the changes through all the frames, we could build up a map that showed motion as a function of time,” Silverberg says.

“People bounce around like the molecules in a gas”

What they found is that people in these pits exhibit two different types of collective motion. In mosh pits, they behave like molecules moving around in a gas. This motion can be described and simulated using statistical models. In circle pits, people following others create ordered movement like that seen in flocking animals such as schooling fish, which use that motion as a defense against predators.

“With computer simulations, we reproduced both behaviors using a few very simple rules, and moreover, could transition from one to the other simply by changing the tendency of simulated moshers to move in the direction of their neighbors,” says Bierbaum.

(The mosh pit simulation.)

(The circle pit simulation.)

Their findings were published in arXiv in early February.

Though their work is being done purely in the pursuit of knowledge, there are already some clues that it might have legs beyond their lab.

A band manager who handles a handful of groups has reached out to them, offering to collaborate so they can record some good, clean data from upcoming shows. It could be a sweet deal for the two metal fans, who have so far funded this research out of their own pockets—meaning they have been paying for their own concert tickets. “The idea of science and metal as a collaboration appeals to these bands,” Bierbaum says.

Besides the public safety aspect, there are also obvious applications for their simulation in the entertainment business—it could create more realistic panic and crowd movement in CGI movie and video game scenes.

“There are smart people studying these social conditions—sociologists, anthropologists and others—but what we’re looking at are the general rules that govern these extreme conditions,” Silverberg says. “Matt and I think movement is simpler than it lets on in a crowd.”

Top Image: Patrícia Marcelino Sacco via Compfight cc

15
15 notes
https://www.tumblr.com/reblog/45916998137/BufRn5FI
Permalink
  1. n0stalgicserendipity reblogged this from txchnologist
  2. semantic-sugar reblogged this from txchnologist and added:
    #science #ftw
  3. behumblr reblogged this from txchnologist
  4. jonnpilgrim reblogged this from txchnologist and added:
    Wtf…
  5. slimdusty09 reblogged this from txchnologist
  6. bobbfett reblogged this from txchnologist
  7. djlegeauxblox reblogged this from txchnologist
  8. wolflambda reblogged this from txchnologist and added:
    Oh the many ways Metal can save the day c:
  9. txchnologist posted this
text

LATEST

blog comments powered by Disqus