At approximately 24:15 he says that the new marker value to be .00001, four zeroes. No idea what that does for the entire multivariate score but it's what I heard him say.