Originally Posted by
Darvamos
What you trying to say is they have a static HQ rate independent of the success rate and only the NQ rate is dependent on the success rate. This would be really silly imo and super hard to test! Bigly amounts of samples needed. It also would of taken more code/work on SE's part.
Code:
Normal: |HQ|-----NQ-----|----Break|
Success+: |HQ|-------NQ-------|Break|
Doesn't seem complicated at all to code it this way. The easiest way to verify would likely be to do a large enough Rusk/HQ+ sample w/o success+ and see if overall break rate drops at all. If not, then it likely does follow my idea.