Simply looking for alternatives and better performance.

I have gotten myself into a bad habit of only using UNION to avoid duplicates and had quickly brushed that thought away. As I think it over now I can see how the additional over head of filtering out duplicates could be costly, but now that you bring that to my attention I've run a little simple testing. When I run both UNION and UNION ALL at the same time and review the execution plans, they are both at 50%. If this is because the size of the selection is not enough to matter, how to I determine the size (aka volume of data) that it will take to make a difference between the 2?

But that is really another question and I am still open to ideas of different queries to produce the same results.