A* Pathfinding Project

Why is Symmetry Breaking Bias capped so low?


We are working with a licensed version of your software and are wondering why the Symmetry Breaking Bias for the RVO Simulator is capped to 0.2.
Is there specific reason this boundary has been chosen?
Reading from the docs there is a chance that characters will start circling a node, however it seems that another occurrence where the characters will all rush and push to a single node (never reach the node position as well) is just as likely (which would still happen if Symmetry Breaking Bias was set to 0). I would think that a higher Symmetry Breaking Bias thus doesn’t have a distinct disadvantage.
So, concluding, is there another reason why Symmetry Breaking Bias is capped (from 0.0) to 0.2?



It’s the range that I have found gives the best behaviors, but if you really think you could benefit from having a higher bias then by all means try it out. The only real limit is that it should not be greater than pi/2 ≈ 1.57 because then agents may start turning around and moving backwards when they encounter other agents.


I see, that makes sense. Thanks for the explanation, keep up the good work!


You may also want to check out the 4.3 beta which has some improvements for local avoidance. https://www.arongranberg.com/astar/download