The Polis Inventory catalogs 1035 poleis in a consistent and uniform, therefore comparable, construction of various physical and historical parameters.
What might come as a surprise, however, is that many of these ancient cities are not definitevely located or are not located at all. The Inventory reports that 166 poleis (>15% of the catalog) are "unlocated". We know something of these poleis from historical writing and record keeping, but we have no definitive archeological evidence for them. Improvements in locations are constantly being made and in this dataset 123 of 1035 poleis remain unlocated.
Given the inspiration of the Polis Inventory App is to plot all these "cities" geographically on an explorable map, what are we to do with these "location" orphans?
My decision was to group them by region and pick a group location that is near to their assigned region, typically, but not always, located over water to avoid any overlap with polies that have verified locations.
An example is given below, when you find a gray bounding circle (as opposed to blue or white that relates to spatial influence (size)) you will know that group of polies have unknown locations but are related to the nearby region.
Just for fun they are plotted as a Fermat spiral so they do not overlap each other and to visually enhance the fact that their locations are contrived.