While you are correct, you are also incorrect. Indeed, each week there is a (1/6) chance of Xur bringing the class item you're hoping for at face value. However, it's become blatantly obvious that some items are weighted more heavily than others which effectively eliminates the (1/6) chance of any particular class item. BUT, the probability of Xur bringing the same item 4 weeks in a row (Calculated at purely face value) is indeed calculated with [(1/6)^4]. That's simple statistics. Simply because OP is taking events that have already transpired and calculating against their relative probabilities. The same could not be said about extrapolating this numbers into future occurrences.