another guideline that i use is, "if it's too good to be true, it probably is".