But we don't have ranges. It's not like a physical measurement where we know that the actual value falls in a certain range with a Gaussian distribution. In his

But we don't have ranges. It's not like a physical measurement where we know that the actual value falls in a certain range with a Gaussian distribution. In history we have disparate sources, often contradicting each other when you have more than one and historians try to figure out which is the most reliable. So all you can say is that you have approximations. It's not "frivolous", they are figures that are interesting for themselves because you can still see trends (like which Empire was dominant, or how depopulation starts to creep down in the late Roman Empire), with the caveat that the figures are "to the best of our knowledge".