Looking at a small database of 1305 games
(my own in this case). I find.

1) No knights on the board at the end in 517 of the 1305 games
2) One white knight, but no black knights in in 451 games
3) One black knight, but no white knights in 425 of 1305 games.
4) Both sides having two pawns on the end in 90/1305 games.
5) 0 pawns on the board in 36 of 1305.
6) Both sides having no material at all (hence a draw) in 4/1305
7) White having two biships, but black no bishops in 73 of 1305.
8) Both sides having two queens in 0/1305
9) White having two queens, but black only 1 in 2/1305 games.
10) White having two queens, but black 0 queens in 9/1305

Just for curiosity, could you look into your database the number of
games out of 1305 which end with (a) and excess of n bishops over
knights regardless of their color, for n=1 and for n=2; and (b) the
other way round?