Joe: I know this - that's why I chose a slightly different way.
about a month ago (one and a half by now) I copied all the items in my fridge into an own table. I erased the items I don't need - coeff. 0 items of course, and also "bad" items because I always avoid them (e.g.

, but also items I avoid only sometimes like

). Unfortunately, I also had to erase items that *can* sometimes appear as coeff. 0 like

.
And

is not countable too^^
Then, about a month later (one week ago), I did the same thing and compared the numbers. I have not finished this part yet, because I think a month is not enough - for example I am totally unable to distinguish between coeff. 5 and 6 items; and I have not found some coeff. 4s too (pencil case, Armor of the Night, ...). So I will have to wait longer. But in the end I will certainly find clear numbers - it is already well possible to decide if an item is coeff. 1, 2 or 3; coeff. 4 only needs a bit more time. Coeff. 5 and 6 will be more difficult.
I started the whole thing at a point where I was (and still am) able to find all items (except for those that end appearing:

), so I think it is a good and statistically agreeable method.
Of course I will have to think about the "bad" items and their impace... But at the end, when it's all finished.