How to handle break shot computation?

[PubohH]

Hi Evan and pooltool community!

I want to say thank you first Evan for starting and maintaining this awesome project and the carefully written blog on the ins and outs of the physics engine. I enjoyed reading them as a pool enthusiast myself.

My question is about simulating the breakshot or when multiple balls are touching. I don't seem to find anything related to this in your blog post. If ball-ball collision happens one at a time, I understand that we can just use the equations for motion to update each ball's motion, but when balls are touching, do we still update their governing equations pairwise? If so, does the order in which we enumerate the ball-ball collision matter? Or is there a different approach to this situation? I apologize for not formulating my question clearly, but I hope you can get the idea.

An interesting phenomena is when two balls (say 1-ball and 2-ball) are frozen and the cue ball collides the 1-ball, in-line with the line of centers of the frozen balls (so it looks like: CB-1B-2B). In reality the 1B moves slightly to the right, the 2B gets almost all of the speed of the CB, and the CB bounces slighty back. I wonder if pooltool takes this into account as this requires treating the collision with deformation. Related article https://drdavepoolinfo.com/technical_proofs/new/TP_B-29.pdf and video https://youtu.be/dYKRAW2zBvA

So I'm wondering if the breakshot simulation would be much less accurate due to this phenomena.

I appreciate any answer and help!

[ekiefl]

Hey @PubohH, thanks for your interest. Currently, all pooltool events are pairwise and instantaneous.

For a demonstration on where this falls apart, I think you should checkout the #physics channel of the Discord. Someone just posted an excellent demo of the shortcomings of this approach, and also how we can account for multi-ball collisions to improve accuracy.