Making the news today in the World of Warcraft is a level 80 protection paladin going into Mogu’shan Vaults (a level 90 raid!) and single-handedly defeating encounters that require (at least) 10 players that are (also, at least!) 10 levels above him. Apparently, he managed to take care of Stone Guard, Feng the Accursed and Elegon before Blizzard caught wind of it. The video below shows his Elegon kill:
Of course, since I play a protection paladin myself, this interested me quite a bit (in terms of “wait… WHAT?!?!”).
Key to this involves a bug with our Shield of the Righteous ability (our primary damage reduction ability), and how it interacts with Vengeance (a passive ability that increases our attack power based on unmitigated damage received). We’ve known about this for a while over on Maintankadin (and I’m guessing that Blizzard knew about it as well, but just never fixed it) — you can click on the link for all the details, but to summarise: due to a quirk of the Vengeance calculation and server/client latency, timing Shield of the Righteous with a melee swing from the boss could result in the Vengeance calculation grossly overestimating how much unmitigated damage we received, resulting in a massive spike of Vengeance, and hence attack power, and hence damage we do. Not only that, but many of our self-heals and absorbs also scale with attack power (Sacred Shield, I’m looking at you here — also Light’s Hammer and Execution Sentence, but since we only get those abilities at level 90, they’re not really applicable here), and one can see how our level 80 guy was not only able to deal enough damage, but also keep himself alive.
But that’s only part of the story. Matt Walsh over on WoW Insider has filled us on the rest: there is some bind-on-equip rare gear found in the levelling zones of Mists of Pandaria (intended for players levelling from 85 to 90) that can, strangely enough, be equipped by a level 80 player. (Here’s an example.) By equipping those, our level 80 paladin managed to gain amounts of secondary stats (dodge/parry/mastery/haste/etc.) that were not originally intended for a level 80 player (when gear item levels were in the high 200s, not the low 400s in the case of the gear he had equipped). More importantly, because he was level 80, he was cleverly avoiding the combat rating drop-offs that occur at levels 81 and 86 (those levels coincide with players encountering Cataclysm and Mists of Pandaria content for the first time respectively). This resulted in absolutely crazy character stats: Matt mentions that our level 80 paladin had 95% haste and 88% mastery, which one simply wouldn’t see on a level 90 character.
All of this came together to ensure our level 80 paladin could survive the raid bosses. The haste and attack power from Vengeance meant that Seal of Insight was regularly overhealing him, and Sacred Shield was fully absorbing his attacks (not only that, but the insane haste percentage meant that it was constantly refreshing as well). Plus, all of that mastery meant a damage reduction boost with Shield of the Righteous.
In a nutshell, all of this was simply our player using Blizzard’s design choices, and creating unintended consequences out of them.
Blizzard’s reaction to all of this has been to hotfix Vengeance so that it caps at a player’s maximum health (until a more elegant solution can be found); Vengeance previously had no maximum limit. While I recognise that this is a temporary fix, I don’t believe it to be ideal: different tanking classes (protection paladins, protection warriors, blood death knights, guardian druids and brewmaster monks) are not created equally with regards to maximum health. It remains to be seen how this will affect encounters that have the strategy relying on the tank having higher than usual Vengeance (10-player Wind Lord Mel’jarak coming to mind straight away); I’ll have to see what difference this makes in tonight’s raid. (Although I may not notice much difference, since my personal gearing strategy seeks to maximise my health pool while still keeping an acceptable amount of secondary stats.)
I’ll end by throwing in this image, because it’s strangely relevant: