|
These are interesting questions to ponder, Mark. From a rigorous statistical perspective, I think there is one more parameter that should be considered, the fraction of surviving examples that can be sampled using public internet data. Most likely, this is actually a relatively small fraction, but I don't have a good idea of how to estimate it. In other words, my anecdotal impression is that the majority of these uncommon and valuable references trade hands privately (or have not traded hands in the internet era), and may never appear in auctions or public listings. There are many one-owner watches, family heirlooms, watches that have been sold privately, watches sold only in the 20th century, etc.
I realize that I am describing the "unknown surviving fraction", which is technically not your topic, but my intuition tells me that this fraction may actually be significantly larger than the numbers you are estimating. And the total surviving fraction is really the quantity of interest.
|