Had another look at the forums - plenty to read. Interesting, especially for what is left out - nothing about Tversky and Kahneman, anchoring, biases, choice theory, JND etc. Plenty on maths especially SVD, but all (or at least all the items I read) ignore the fact that there are humans actually making the ratings. Surely, this piece of information should be informing the way people go about trying to build their models?