Jared Carr <jared@89glass.com> writes:
> I am currently working on optimizing some fairly time consuming queries
> on a decently large dataset.
It doesn't look that large from here ;-). I'd suggest experimenting
with reducing random_page_cost, since at least for your test query
it sure looks like everything is in RAM. In theory random_page_cost = 1.0
is the correct setting for all-in-RAM cases.
regards, tom lane