Agree that this should probably be in the water cooler, since no one is talking about the OP any more. I think the basic assumption that bridge players use a statistical definition of range is flawed. I think of the usage here just as "the set of allowed values" - I'm not a statistician(?), and this definition is more intuitive to me. The bridge usage is not really a measure of dispersion, it is more a measure of cardinality. "A 3-HCP range", to me, is just shorthand for "A range with 3 HCP (value)s" - just as "a pumpkin pie" is "a pie made from pumpkins".