anyhow for the dewey decimal system one i totally lazied out and didn't even write a quiz, it just changes your name to some numbers and works it to three digits. the color quiz, on the other hand, has all these complicated algorithms that take into account your answers, other people's answers, common colors, the works. Both quizzes are successful because there are tons of answers... you're unlikely to get the same one as someone else on your friends page. They're also both complimentary and the answers are a little bit generic. I learned a while ago when I made a "how dumb are you" quiz that people only post results that are flattering.
But colors go in a nice order, and the dewey decimal system doesn't, so I didn't bother writing an algorithm. The numbers don't correspond to anything terribly logical, it's just a kind of random index.
The revelation is that I think each quiz has about the same odds of getting posted alongside a comment about how accurate it is. that's half of what i look for when i read blog posts with my quizzes... are people saying "this is so me" or not? and that doesn't seem to have much of anything to do with how random I make the quiz. Some people are a little disappointed that the dewey quiz doesn't ask any questions... it's made me think i should just post some fake ones. But no one is disappointed with its accuracy.
but I can't make fake questions to make people feel better, because part of me honestly thinks that if you have to answer questions about yourself to get a quiz answer, you'll have higher expectations of its accuracy. so if it only halfway makes sense you won't post it. if it's based on random nothing, and halfway makes sense, that's pretty good!
I have no grand conclusions from this, I'm still studying this whole internet thing, but I'll let you know if I discover anything brilliant. either way the quiz is doing well and it's good to maybe have a successful one again, because that color quiz was 2003, I was starting to feel like a has-been.