I just read this excellent Viktor Nacht post which concerns beating continuous shuffle machines:
"With intimate knowledge of the model-in-question's internal machanism and the machine's current settings, you could attempt to beat the machine by taking advantage of any latency in returning the played cards into the shuffled section of the machine, or take advantage of any non-random attributes of the shuffle method. "
The constant innovation at RGE never ceases to amaze me.

