frrol, 1. What is your education level? Did you ever worked with computer models? You do not have to understand all the details of model all you ned to know its implications. It is all on the slides 23 & 24. It is not garbage in, so relent with discounting it. this is how far we got with analysis over of 10 years writing those. 2. Yes 50% is coin flip but if you are in those 50% you are improving about +3 MMSE/year. See the data with the eyes of a scientist.
As I heard Missling, in Part B patients had continued on the dose they where on in Part A. The strong correlation coherently across all measures were that those on the higher doses continued to improved and those on too low doses did not.