Matrix,
I did not try any of Rubro's hints, in fact I even didn't look into it! The main problem is numenclature; you do not know if what he calls "confident interval" or "confident gap" is what you or anyone else understands from it!
The other problem is that the guy just popped here for some minutes wrote this, and took off without saying anything else after... in fact I don't even recall seeing him again, but LT can check this if someone requests it!
Take for fourier analisys, you can use it on the history of a given position from a given lottery (assuming that history could be described as a finite function) and by doing so, theoretically you would obtain the frequency spectrum of that history! The first problem is that fourier analisys is a base change into a complex number plane; your result are a series of complex numbers how do you interpret them! The second problem is that in lotto all numbers are positive, this means that for example the fourier transform of any series taken from it will have a strong DC component, that will make the resolution on the AC so low that he might make it unreadable!
Does this invalidate what Rubro, presented... No, it only points to the fact that if he managed to get it working properly, he is not sharing the detail that makes it work (and why should he?!), which means that you or anyone else will have to study the problem carefully to try find it also!