...
Of course the cost of our approach is that many events are discarded because the large area of vertexes is needed to provide correct illumination of the instrument. There is an approximation at the edges of simulated area - of course it is larger than maximum angles causing 3-in-a-row but still some (small) fraction of events which could cause (large off-angle) triggers are missed. I remember that I did the study what is the reasonable size of simulation area (currently 2m 40m by 2m40m) and came to this size.
To get 200K triggers I need to run about 10^8 trials.
...