This is quite interesting. At first I couldnt understand why the probabilities didnt sum to a value of one, but I assume its due to the setting of the movement range, so at an extreme, if you set a target and stop of 100 pips, but only specified a 20 pip movement range, you'd end up with zero probability of either being hit. That's pretty neat

I also like the mean offset bias feature

Now I'm going to have to spend the rest of the day brushing up on stats and reverse engineering this thing