There has been an excellent thread running in the EU forums laying out constructive criticisms of the new arena tie-break system. This system, if you hadn't heard, introduces a new buff called "The Crowd Chose You!!!
" which will be applied to one team or another based on two criteria: the team with the largest number of players alive, and the team that brought an enemy player closest to death. Blizzard EU Community Manager Taepsilum posted in the thread to clarify that this is still very much in iteration, and to calm the fears of the posters, but a couple of great points were brought up.
Firstly, if one team gets an enemy player to repeatedly low health percentages, for example, they get a mage to 10% then to 5% then to 8%, but their opponents get one of them to 2%, do their opponents win? It sounds, from reading about the system, like the team that got the player to 2% would emerge victorious, but this seems a little counter-intuitive. Another great question is whether the health for the calculation is percentage based or numerical. Percent would make more sense, it seems. Hit the break for Taepsilum's post.
There are some very good points being brought up in here, particularly the dispute between the "frequency in which each team gets someone from the other team close to death during the course of one game" Vs "how close to death the enemy was".
Some solutions could encourage turtling and we have to pay attention to that.
On the other hand, we also don't want to change players' behaviour too much because of this, for example, some healers might stop feeling comfortable playing at low health in order to keep the pressure on the other team during certain moments, and paladins/spriests/mages/hunters/rogues won't wait so long before they bubble/block/deterrence/vanish, etc...
Who should win between a team that nearly kills their opponent 20 times during one single match and a team that leaves their opponent at 1 HP and plays defensively during the rest of the game?
Maybe there could be a mix between both variables with some weight applied to each of them.
We'll make sure to share the potential issues you brought up with the devs, there are many solutions available to break ties and it's possible that we end up seeing some iteration on this.