[zljones]

In a earlier thread I started we talked about what determines rarity, I myself developed a whole new insight and will stop using the term "rare" so much.
Now the next similar topic I would like to discuss is population of a card. Although there is no way to get an exact population of a card maybe there can be a formula used to estimate how many ungraded and graded are out there.
What we all know is, is that SGC, BVG, and PSA do not show all their graded cards on their sites. They make mistakes and leave some out. For example I am the proud owner of a N172 portrait Comiskey Browns Champs. Between PSA and SGC they only show 3 copies ever graded. That is incorrect because I own a PSA AUTH Comiskey, and PSA shows zero for Auth grades. Also I discovered a PSA 6 move at one of the auction houses; however, PSA claims zero 6s.
Another thing we know is that there are cards out there in private collections, ungraded, that have never been made public. My question is how many of those secret stashes could possibly be out there? You would think by now most of them would have been sold or somewhat made public, by grading or forum posts. Since the 1970s people have been tracking down cards and many many elderly owners have sold off or given away their collections. My theory is, is that there are not that many secret stashes out there compared to those that are both graded and also changing hands. So that can narrow it down, right? I don't know
Ungraded populations can be tough to pin down. Someone would have to pull Terapeak at least annually and track every card that sold as raw. Then other auctions and online card shops would have to be tracked as well. The best part about those is there are sales records some where. Wouldn't it be great if there was a service or research group that would ask every dealer on the net to share sales records? If so alot of dealers probably would consider that too much work to dig those up.
What I am getting at is, could there be a formula used to estimate the population of certain cards using PSA, SGC, BVG, and GAI population numbers. For example maybe say population is 30 between all four grading companies, then multiply it by 10 to get a rough estimate of the whole population. Then as far as those cards that are more popular that have a higher graded population, maybe divide it by 2 then multiply it by 10 or whatever. I am just making these formulas up as examples. There can also be different formulas for individual sets like T206 is multiplied by 6 and R319 is multiplied by 15.
I hope I do not sound ridiculous, but it kills me wondering how many of my cards are really out there. I do not need an exact number just a closer estimate. I have a hard time digesting the fact that there are less than 10 N172 Comiskey's portrait. Or that there are only like 50,000 89 UD Griffey rookies.