Thursday, March 12, 2009

Rebuttal of "Exhaustive Tank Comparison"

This post is aimed at being a counter-point to Parris analysis of the state of tanks, which some have already pointed out quite aptly should have been renamed ”The problems of block scaling”. Firstly as has been noted, the analysis only deals with the mitigation of physical damage. This means the results say nothing about the current disparity in magic damage mitigation. The reason this is a problem is that he attacks druid health quite hard in the post, whilst conveniently bypassing the reason druids were given such large healthpools, namely their lack of magical mitigation. As he later points out, mitigation always scales superior to something which is essentially a fixed number, such as healthpools are, as they are dependent on gear in the same way that block is.

Parri states clearly that the reason high magical burst encounters are not included is that he considers them gimmicks. He instead opts for a 20k dps physical boss, starting with a 1.0 swingtimer and ending with a 2.5 swingtimer. This sounds quite reasonable, but upon closer inspection the swings have increases to 50k per hit by the time he reaches his 2.5 swingtimer. Furthermore he boosts the incoming damage to clarify the disparity, ending up with a boss hitting for pure physical damage at 75k per hit. That, my dear friends, is a gimmick fight that favours scaling mitigation and avoidance. Pure and simple, what at first seems a reasonable test has turned into a biased attack on high-armor classes.

A more reasonable approach is to assume a far lower amount of physical damage, as any purely physical boss, as has been shown, turns into a gimmick as the swingtimer increases(or, indeed, decreases, as block mitigates a larger and larger amount of ever more frequent hits). Therefore a more reasonable test of tanks physical damage capabilities would be a more restricted(1.0-2.0) swingtimer boss who's physical damage contribution is something in the range of 10-15k dps and where magic makes up for the rest of the damage. Now if we ignore the magical damage portion and focus on the physical portion, we'll find block fairing quite a bit better and I suspect overall the DTPS numbers would be within Blizzards guidelines of viability.

At this time I'll appoligize, I'm more theoretically inclined as opposed to empirical. However, I believe Parri's analysis provides enough proof for me to state that perhaps a more theoretical approach is overall better, as empirical evidence only tells us about specific circumstances which can easily be manipulated and influenced by personal bias.

As some have also pointed out, the analysis was made before the inclusion of recent DK/Druid nerfs, and these will change the outcome to some extent. In particular for druids as we've incurred two scaling nerfs, it is quite possible that our advantages on even the gimmick side of slow hard hitting bosses will have been cut quite a bit. Whilst I wouldn't argue this is unfair at this time, the coupling of armor and hp nerfs is in no way compensated by a mechanic such as SD, and furthermore we're in a very bad place now in terms of surviving high magic bursts. Needless to say that the polar set was a gimmick that is unlikely to be useful for any future fights, as it coupled high magic damage bursts with low physical damage and no requirements on TPS.

Furthermore whilst Parri shows off avoidance great ability to lower DTPS, no mention is given to the fact that druids are at a point where further scaling is hindered whilst the more diversified parry+dodge classes will still be able to continue scaling in T8 gear.


amorob said...

Your are completely correct. If Druid and DK Health pools and armor are going to be nerfed, something must be added so we do better against Magic damage.

Melthu said...

I'm confused, what exactly makes us bad at taking magic damage? We get 12% reduction from Protector of the Pack, Death knights get 15% from Frost Presence and the possibility of 5% more from Magic Suppression, Warriors get 16% from Improved Defensive Stance, and Paladins get 15% from Improved Righteous Fury, Guarded by the Light, and Shield of the Templar. We have 4% less magical mitigation than warriors and 3% less than paladins, but we have considerably more than 3%/4% more health than those classes, making us superior magic damage tanks.

Anonymous said...

Something to think about is cooldowns and abilities. Warriors have spell reflect, and Shield Wall (60% damange reduction). Our barkskin is only 20% damage reduction. Given ours has better uptime, its importent to note that magic damage is usually predictable. Druids have 2 cool downs, Barkskin and Survival Instincts, and one spell interupt on a 2 minute cooldown. When it comes down to it the other 3 tanks have better and more ways to deal with burst magic damage then we do. Hence our larger health pool.

Shamad said...

Anon summed it up quite nicely, yes beyond having the worst base magic mitigation we've also got lackluster cooldowns, therefore the statement that our magic mitigation is, overall, the worst.