The physics of mosh pits(kottke.org)
kottke.org
The physics of mosh pits
http://kottke.org/14/11/the-physics-of-mosh-pits
7 comments
Those black dots, which are trying desperately to get OUT of the mosh pit so that they can watch the damn concert but keep being shoved back in? That's me. At every.single.show.
I've seen this many a time. The trick is to come to a stop at the edge, and then walk out. If you are moving any kind of speed, folks on the edge will just assume you are disoriented and give you that "helpful" shove back into the mix.
I hear you. They should also model the annoyance of the black dots that seem to be the barrier between the moshers and everyone else. Those poor human fences constantly get smacks at full force from the red dots on one side, and annoyed looks from the black dots as they (unwillingly) convert the high impact into a low frequency push forward.
Guess where I always end up at shows?
Guess where I always end up at shows?
What's interesting is when there are a lower number of people, say 100. With so much space, the "moshers" tend to just go in one direction until they bump into another person, but this isn't really accurate. There needs to be some level of randomness in their movement, so that when there are large spaces they are not just coasting, but still bouncing around.
That being said, this is pretty fun, and the dynamics for crowded areas looks appropriate.
That being said, this is pretty fun, and the dynamics for crowded areas looks appropriate.
With only 100 particles, it looks more like a professional pickpocketing ring simulator.
You can get something like what you want by changing the noise strength.
Yea, that's definitely an improvement. I would prefer if there was some sort of "cohesion" measure among the moshing particles. Groups, no matter how small, tend to stick together.
That's the flock strength slider. If you turn it down to zero, you get no flocking, and red particles will only interact via direct physical contact.
Be sure to klick through to the full simulator [0]. Here you can adjust values, steer the crowd (using a,w,s,d) and see the results at the same time.
[0] http://mattbierbaum.github.io/moshpits.js/
[0] http://mattbierbaum.github.io/moshpits.js/
Interesting but disturbing article on crowd quakes -
http://arxiv.org/pdf/1206.5856v1.pdf
http://arxiv.org/pdf/1206.5856v1.pdf
Once at a Grunttruck concert, for a brief time I was the only one in a small gap of the mosh pit, just in time to receive the full brunt of a stage dive. I was out for 10 minutes.
Mosh pits are perfectly safe without stage diving.
Mosh pits are perfectly safe without stage diving.
Hah, reminds an epic stage dive fail of an acquaintance of mine - http://www.youtube.com/watch?v=Z4hB562aEDk (skip to 0:45). He broke his arm then.
The simulation seems broken when you increase the box size... If the band is good, most people (black and red alike) will push towards the band.
Looks pretty convincing as a model.
Just needs some random blank spot every now and then where they're avoiding the puke on the floor.