I suspect the guys who poured silver into individual coin flans got very good at estimating rather like your grandmother could cook without a fancy set of 1/4 teaspoons. I also have suspected that some issues were either done better than others or held to a more strict standards simply because I have seen more scoop adjusted coins for some issues than others. I recall seeing more than one other coin of M. Volteus as shown above that had a scoop. That could mean that the guy who poured the flans started with more metal and did more adjusting or that other issues cheated a bit more and started with exactly the right amount of metal and were able to pour exactly the right number of coins with greater skill making scooping unnecessary. Answers to these questions would require examination of a thousand coins of each type and I am lucky to have one. Perhaps we will eventually have huge databases starting with data we have from CNG, acsearch etc. and people with free time can find patterns previously only open to museum workers.
n=10; Mean= 3.85; Mode = 3.95; Median = 3.935 But I agree that you know when you know. A few years back, J. P. Martin came to the annual Michigan State Thanksgiving Show, just for himself, not to work; but as happens, two men came up to him and one handed J. P. a coin, asking for an authentication. He looked at it. "It's fake." The guy was not happy. "How can you tell?" J. P. just paused and the other guy turned to his friend and said, "If someone put a Chevrolet nameplate on a F-150 would you know? Mr. Martin knows ancient coins." End of discussion. Ah! Like the Greeks... I wonder how long that continued. Carolingians, no doubt, but how far into the Middle Ages? It would make an interesting exploration.