Originally Posted by PogoPogoPogoPogo
Elo rating is a measure of how good you are at winning matches.
Your system gives Elo weight to something other than winning matches, thereby making Elo less accurate.
Someone give this man a medal.
Anyways, to add to what he said: your teammates are drawn from the same pool of random players as your opponents. If you go 10/0/15 every game but are losing most games, then you are a much worse player than the guy going 0/7/4 that wins most games. There is really no debate at all possible about that.
Look at it this way. You say that a team that wins has a good kda so getting kills and assists is the best way of measuring performance. It's pretty easy to see that if KDA actually was the best way of measuring performance, this whole thread would be unnecessary: people with good KDA would already rise while people with bad KDA would fall. The mere point that you think that players with good KDA shouldn't get stuck at a lower Elo proves beyond any doubt that their KDA is not contributing strongly to their team's chance of winning.