Odd that Tomb found the two scarce ones and can't find the common one.
There aren't any real numbers to describe how many out of how many are
Wide AM, because there aren't any official numbers kept by the mint on them. Given the following possible scanarios, this might help...
There are at least two, possibly three different dies known to have made the 1998
Wide AM cents. There is but one die known for all the 1999
Wide AM cents discovered to date. In 2000, there are at least a half dozen, with more possible.
If a die makes it all the way through it's expected life, it mints around 500,000 coins.
Assume the dies all made it through their expected life, and assume the numbers of dies reported above as the total existing...big assumption, but this has to be done to even come close.
There were a million and a half of the 1998 coins made, 500,000 of the 1999 coins, and at least 3 million of the 2000 coins.
Compare this to overall mintage and your chances are the following:
1998 - 1,500,000 in 5,032,155,000, or about 1 in 5,000.
1999 - 500,000 in 5,237,600,000, or about 1 in 10,000.
2000 - 3,000,000 in 5,503,200,000, or about 1 in 1,800.
Of course these numbers are only a very erroneously basic idea of what's probably actually out there. Take away half of the number made for any given die because of shorter die life, and your odds double that you won't find one...very possible.
Reason? Die life can be rounded down closely by examining a broad group of the coins minted by a die. If a decent sampling are observed for die wear, the latest die state found can be assumed to be the latest die state made, thus helping draw closer to the actual number of coins minted by that die.
One example - 1960D-1MM-004 (RPM#4). It is scarce, and is only known in very early die state and early die state. No examples of later die states are known and probably do not exist. Given this, we can assume that this die minted only around 2,000-5,000 coins. Take that to the same math above, and we end up with a number like 5,000 in 700,000,000 (about half the 1960D mintage, accounting for large and small date). This deduces to 1 in 140,000. Pretty scarce.