BEagle - if a technique works then it should work on RANT, a system so simple that the variability even of a simulator has been removed, and results should be most consistent. I agree that fudgy must be doing something wrong, as although I have neither taught nor used the technique I am not aware of large errors, but that is not an excuse for being patronising. It in fact goes against your insistence that SCA is the only technique to use and infalibly simple! Perhaps fudge should simply be using a different method, that might suit him?