If we can make the chances of a successive sequence that will completely beat us rare enough to be a once in a lifetime event, we have a practical HG. By completely beat us, I mean beat all of our recovery attempts successively and faster than we can recover.

I find calculating the actual probabilities here to be extremely difficult by simulation and to my mind math as well IF one is willing to properly apply cumulative probability to the calculations (which is the only way to be accurate); no matter how you try and calculate it, the iterations are far too many to deal with.