Shifting Perspectives: Can Wrath remain a 1.5-second cast?
Every Friday, Shifting Perspectives explores issues affecting balance druids and those who group with them. This week, we are looking at how to handle Wrath in a world that can't handle it. Also, a notice that I am in need of some screenshots of pro balance druids doing whatever it is that balance druids actually do. Please send them to tyler@wow.com; put something akin to "Balance screenshot" or "Moonkin screenshot" in the subject line, and don't forget to list the name you would like to have credited for the shot.
A fair warning, if you do not like reading math or wild speculation, then this post is not for you. I will do my best to keep everything as simple and easy to read as possible, but no real promises.
With that out of the way, it's time to get into the core of the matter. Right now, balance druids have a serious problem when it comes to Wrath and the GCD. Although the main culprit of the issue is Nature's Grace, which causes us to cap the one-second GCD at 400 haste, that isn't our only real issue. Haste effects have scaled a bit out of control in this expansion as a whole, and it seems that in Blizzard's attempts to make us utilize both of our main nukes, they have forgotten to take that into account. Just as an example, even without Nature's Grace, Heroism makes Wrath practically unusable by blasting through the GCD without a worry. Even back in Trial of the Crusader, when we were still trying to cap our haste as close to 400 as possible, it was an issue, so it isn't a wonder that Icecrown Citadel has been much worse.
The gear scaling issue is one that Blizzard has said they are going to address in the coming expansion; however, will that really be enough for us? Even back in Naxxaramas it was more than possible to get well over 400 haste, which was before the gear scaling issue even took effect. I've already talked about how Blizzard has stated that Nature's Grace is going to have to be changed come Cataclysm, but is Nature's Grace really the core issue here, or it is Wrath's cast time itself? Ask any shadow priest and they'll tell you how troublesome it can be to work with a 1.5-second spell even without all of the complications that balance druids have. For a cooldown-based spell, it might be more acceptable; when dealing with a main nuke that's constantly being chain cast, it becomes a more complicated issue.
Functions of haste
Getting into the basics, the first thing is to take a look at how the haste system works in general. Haste as a stat follows a very simplistic formula in how it reduces spell cast time; where there become hiccups within the system is with haste effects from talents and buffs. Talents such as Celestial Focus and buffs such as Improved Moonkin Form or Wrath of Air Totem operate in a slightly different way than haste from gear. Unlike buffs that increase critical strike chance, which are additive, effects which increase haste are all multiplicative. When you have the three mentioned effects active at one time, you don't have 11% haste; instead, you end up with 11.3945% haste, which theorycrafters generally express as 11.4%.
With the current raid buffs, Wrath ends up with a cast time of 1.3465 right out of the gate. At that point, it would require 1137 haste rating in order to reach the one-second GCD. This ends up working out to be 34.6752% haste; remember that Heroism is a multiplicative stacking 35% haste buff. Should balance druids really be forced into a position where one of their main nukes is clipping the GCD every time Heroism is used? On one hand, Heroism is only 45 seconds out of every encounter, some of which can last a very long time, so the overall impact in the long run of things can be rather low; it's a common statement that Heroism is actually worth a lower DPS gain than things such as Demonic Pact and Earth and Moon. That is true in some instances when you equalize out the gains across the entire encounter. On the other hand, Heroism is often used during periods that require high amounts of burst where the equalized DPS gain isn't as relevant. Phase 3 Professor Putricide, the second transition phase during The Lich King, and phase 3 Anub'arak are examples of this.
There is no correct answer to whether or not Wrath should ever be capable of clipping the GCD. All of it depends on your perspective and the design philosophy. I'm of the opinion that no spell should be able to break the GCD save for very specific examples under very specific circumstances. With the frequency that Heroism is used in encounters, I don't see it as specific enough to warrant it the ability to cause a non-cooldown-based nuke breaking the GCD before haste from gear is taken into account.
Fixing the haste issues
Removing Nature's Grace, carefully structuring haste effects and tweaking gear values are all well and good, but are they really solutions to the problem at hand or are they merely band-aids trying to close a rapidly widening wound? Changing all of these aspects to where Wrath will no longer break the GCD is certainly possible, but is it really the solution that we should be look for? Perhaps the issue isn't so much the myriad of haste effects that we have, but rather Wrath itself. Going out of the way to create a specific and sensitive system just to accommodate the cast time of Wrath seems to be a major headache. Change whatever you want -- the fact remains that every single haste change in the game is going to have to take Wrath into account. That's a level of depth which the development team may or may not always have time to consider, or they may not always have a viable workaround for it.
The bottom line is that the game shouldn't have to be designed at a level where the development team is constantly having to doublecheck everything against a single spell. To that end, there have been many suggestions as to how to fix the GCD clipping issue. A very common suggestion that has come up is having all haste effects that would dip a spell, in this particular case Wrath, below the GCD increase the damage done by the spell instead. I want to say here and now that this is not a good idea.
Allow me to rephrase that. It is a good idea in theory; it would just be excessively wonky in implementation and balancing. Haste is a highly fickle stat in many ways that can change over the course of an encounter multiple times over. We've seen loads of trinkets that have haste procs or on-use effects. If Nature's Grace isn't outright removed, then there is another haste proc to factor in. Movement can cause totems to become out of range of you. And then there have been various potions which temporarily increase haste. Essentially, players are asking the system to calculate the variable degree below one second that a spell is cast and then modify the damage based on some formula to that effect. It is entirely possible for the servers to do this, but they won't do so without a decent load placed upon them. The impact it would have on the servers is of a minor concern, though; the bigger issue is in the balancing.
Essentially, this system would turn all haste after the GCD into spellpower, which is our strongest scaling stat. Given how well haste already scales for Starfire and that it will finally end up increasing the damage of our DoTs, it would be very, very easy for this change to grossly overvalue haste. There are already break points, even in Wrath of the Lich King, where haste can become more valuable than spellpower. Even if such breakpoints didn't exist, haste would easily out-scale both crit and mastery. Although all three stats don't have to scale at the same rate, going back to the TBC model where haste held a scalar value of .9 (where 1 is neutral scaling value of spellpower) and crit had a scaling value of .5 isn't a good system. Favoring one stat over the other is fine, but the scalar difference should only be around a factor of .2 between each of them; otherwise, you run into a case where players actively avoid a stat, or even where a high item isn't an upgrade at all because it simply has the "wrong" stats on it. In this case, if haste held a scaling factor of .9, crit one of .5 and mastery one of .7, then balance druids would seek out every piece of haste/mastery, haste/spirit and haste/hit that we could find. A new item that was mastery/spirit, mastery/hit or mastery/crit would still be an upgrade, albeit a minor one, whereas items that were crit/spirit or crit/hit could easily become a downgrade.
It would be fine if crit/spirit or crit/hit items were a minor upgrade in comparison to an older haste/spirit or haste/hit item, but we should never run into a situation where changing to a "better" item is actually a DPS loss. Obviously this is all speculation; we have no idea how the scalar values of haste or crit will be changed in the new expansion. But it would currently take a major overhaul in order for crit to be even remotely close to haste in terms of DPS value. This is especially true if Nature's Grace is removed. Even with Eclipse and Nature's Grace functioning as they do, crit only exceeds haste in terms of scaling because of the cap on Wrath. With that cap removed, there's currently nothing at all to create any form of scaling paltry between crit and haste.
