Ranks based by K/D ratio only could be abused as well. You could get insane k/d when you start playing battle with a leveled char and after getting those you can only play siege.
My last generation longswordsman had a k/d of ~ 12:0 after the first couple of rounds on the battle server. Of course, it settled down to ~3.1 (which is not that high, I assume) after about two weeks of play.
AFAIK, my current generation lvl 18 longswordsman has 20 kills and 2 deaths (just played a couple of rounds on the battle server this Wednesday), so the KDR is 10.
What I want to say using these examples is that a system based on both k/d and the total number of kills is needed if you really want to see which player has the biggest e-peen. Still, the actual value of this "ranking" would be close to nothing (I saw the best US melee player messing around with a wooden sword/wearing no armor/doing other similar things, i.e. simply having fun on the battle servers numerous times).
