Harlan Grove wrote:
Alan Beban wrote...

Just out of curiosity, are your names grouped or ungrouped? If they

are

in fact grouped, then the simplest approach first posted should be
considered.


Granted. If names are grouped, then Alan's original formula is optimal.

However, OP showed unsorted sample data.

No; the OP showed grouped, unsorted sample data.

What's the *reasonable*
assumption to make when presented with unsorted data?


Oh please! Why are you driven to making a contest out of the dumbest
issues? The question is whether the OP showed grouped, unsorted data
because his real data is grouped and unsorted, or because he didn't
realize that his sample data didn't represent his real data. Either
assumption seems reasonable to me; and in the absence of any knowledge
about the OP, other than that he went to the trouble of grouping the
data, the first seems a bit more reasonable than the second. Might not
be the correct assumption, but certainly reasonable.

Alan Beban