It doesn't matter how many people submit for Donkey Kong. That really should have no factor whatsoever.
If we assume that submissions are generally of strong quality and therefore also use overall rank on a title to avoid having to try to theoretically rationalize game scoring equality:
Quantity increase of submissions increases the difficulty of being ranked higher on a given game which weight a game with more quality submissons more heavily.
OK, well if you combine both tournament, and marathon submissions, MAME/Arcade I am 1st out of 50 submissions.
For Donkey Kong Junior Submissions I am 2nd out of about 100 submissions, and for Donkey Kong I am about 50th for about 200 submissions.
So me
1st out of 50 for DK3
2nd out of 100 for DKJr.
50th out of 200 for DK
For Mark Kiehl
1st out of 100 for DKJr.
7th out of 200 for DK
50th out of 50 for DK3 (Since he does not play the game)
So now let's make the positions equal to each other, and make them equivalent to DK3.
So for my ranking it would be 1st for DK3, 1st for DKJr(You cut the 2nd to 1st since there is twice as many submissions), 12.5 for DK(You divide by 4 since there is 4 times as many submissions)
For Mark it would be 0.5 for DKJr (Half of 1 is 0.5), 1.75 for DK (7 divided by 4 equals 1.75), 50 for DK3 (Since he has no score I know of).
So now we add the totals, and the lower total is the better one.
So for me, 1+1+12.5=14.5
So for Mark, 0.5+1.75+50=52.25
So even if we do it by ranking divided by submissions I still do pretty well.
Edit: if you look at what I did, I even did what Steve suggested. Give the DK ranking 4 times the value, give the Donkey Kong Junior ranking 2 times the value, and give the Donkey Kong 3 ranking 1 times the value.