Of all the statistics in numismatics, mintages are among the most important. Simply put, a coin’s mintage represents how many were actually struck. As one can imagine, this figure is an excellent indicator of a coin’s rarity—but it doesn’t necessarily tell the whole story. As this article will describe, mintages can be somewhat misleading both on the high and low side.

Generally speaking the United States Mint kept extremely precise records of how many coins it made. After all, the Mint processed a tremendous amount of precious metals and was accountable to the government (and taxpayers) as to how the metal was used. Keeping exact production counts was of vital importance to make sure balances were correct and employees were not pocketing metal.

Nonetheless, mintage figures were still inaccurate from time to time—especially in the late 18th and early 19th centuries. The Mint may have recorded, say, how many half dollars were struck in 1805—but there’s no guarantee that every piece struck that year displayed the date 1805. Sometimes old dies from previous years were “held over” into the following year.

On more than a few occasions, US Mint employees would also strike clandestine restrikes, one-off experimental coins and other issues that never made it onto the official records. These items were often made for collectors, or because a high-ranking official wanted to give a special coin as a gift. Some of America’s most famous rarities (like the 1804 Silver Dollar and 1913 Liberty Nickel) fall into this category.

Even if the mintage figures were 100% accurate, they still might not be a true indicator of how many pieces actually survived. Some issues were melted in massive quantities; some of the 1924-1933 $20 Saint Gaudens Double Eagle issues come to mind as examples. Many of these dates have mintages in the hundreds of thousands yet less than 1% survives today. This is because most of these coins were still sitting in U.S. Mint vaults when Franklin Roosevelt issued his gold ownership ban in 1933.

A similar phenomenon is seen with United States Commemorative coins. If an issue did not sell well, a common solution was to melt the remaining examples. At least with Commemoratives the Mint often kept a precise count of how many pieces were destroyed. For other issues numismatists must guess based on population reports, auction records and anecdotal evidence as to how many specimens survived.