i would like to see the populations behind your statistics please. I agree that the honor system has the potential to curtail the low to mid range amounts of negative behavior (by low to mid rang i mean people that are not insta trolls or just ugly people to deal with but are decent people that can get set off) but without some sort of carrot to cancel out the stick of the tribunal (which is a whole different subject in and of itself) i dont see how its going to drastically change the way how people are in game. If the goal of the system is to just provide motivation for that group of players that aren't permanently awful to not be awful and to reinforce good behavior, congrats on the implementation, and thank you for implementing something like this.
However one weeks worth of results may not show the true level of impact something like this has on a game the shear size and scale of League of Legends.
But with that being said i just wanted to know what you were measuring the statistics you mentioned in your first post off of, as they kind of leave me wondering what sort of player population you were pooling from; such as was it from a 100,000 player pool, or 100 players over a set span of games or what.
Some clarity on that would be appreciated, and if you could implement some sort of reward for the program that would be absolutely fantastic, even as it stands with just the little ribbons saying honorable opponent thats cool but could you make achievements out of this? just an idea