On A Mallows-type Model For (Ranked) Choices