Shifting Perspectives: Total Eclipse, part 2, page 2
Gaming the Eclipse bar
As I have talked about a lot, Eclipse is based on spellcasts; its uptime and damage contribution is balanced around this expected average. In numerous single-target situations, this turns out to be a fairly balanced system. We get no more out of Eclipse than any other spec gets out of its mastery in most cases. Issues arise, however, where encounter mechanics force balance druids to find ways to "extend" the duration of Eclipse procs.
Taking a quick glance as WoL, and the trends are rather obvious. On single-target encounters, there are a few really good balance druids who hold some of the top rankings, usually without using gimmicks. This is expected -- even not having any of the top rankings would be expected. Single-target DPS should be equalized across all specs within an acceptable range that can slightly favor pure classes. In such a system, you would expect to see mostly pure classes holding top rankings, with a few hybrids who are highly skilled and geared picking up a few as well.
We see very obvious shifts during numerous encounters, however. In a single-target encounter, there is no incentive to game Eclipse save for the possibility of Bloodlust or trinket cooldowns, which are minor factors. In dual-target and AoE encounters, balance druids are forced to work Eclipse toward their favor in a way that the spec was clearly not designed to support.
The balance of Eclipse
The balance of Eclipse hinges upon two very basic principles. First, it will have an average uptime during an encounter that allows the damage contribution to equalize close to that of other masteries. This we hold as true, as evidenced earlier. Second, those abilities that benefit from Eclipse yet do not adjust the Eclipse bar are low enough in DPS that they aren't worth using for extended periods of time, meaning that spamming Sunfire would result in lower DPS than using Wrath to drop a Solar Eclipse and move to a Lunar Eclipse. In single-target situations, this is true; in all other situations, this is false.
Previously I have said that haste has no impact on Eclipse uptime because Eclipse operates on the number of spells cast, not the frequent of their casting. This is true in only the strictest of senses. DoTs continue to benefit from Eclipse even after the buff is lost; therefore, haste has a direct correlation between the uptime of Eclipsed DoTs. The faster that you can switch between Eclipse procs, the higher uptime that you have on Eclipsed DoTs. This is because each DoT would have to be refreshed less frequently, which would usually only be done twice per cycle -- once in Eclipse, once outside of it.
In a single-target situation, this synergy doesn't hold too large of an impact; it takes a large number of sequences for haste to hold a significant impact on Eclipsed DoT uptime. The more targets that exist, however, the larger the impact haste has on Eclipsed DoT uptime, and the longer Eclipse lasts in general.
While there is evidence of this in math, it simply isn't needed, as it has been clearly shown within the game world as well. Parses from Twilight Ascendant Council, Maloriak, V&T, and several other encounters show how much of an impact this holds on our DPS, where we see a clear over-representation of balance DPS.
The arguments of high DPS are largely theoretical. Certain encounters will favor certain spec mechanics in different ways which can lead to over representation of that spec. This is an unavoidable consequence of game design. The primary issue with this is the cumbersome nature of the system itself.
Is gaming the system fun?
Toying with Eclipse procs to either delay or retain a proc for a specific encounter mechanic is excessively annoying, and I know that I cannot be alone in this. While I can agree with sentiments that it makes balance fun, challenging, and dynamic, by the same token, it also discourages a large sector of the playerbase.
Not being able to properly game Eclipse isn't a matter of minute DPS differences, as should be the case for skillful play, but rather it creates gaps well into the thousands. Having skill account for 10% of a player's overall DPS -- which would fall in the range of 1,500-2,000 in current content -- is acceptable. The DPS disparity between being able to game Eclipse for AoE or dual-target encounters can easily account for a difference of 5,000-6,000 DPS or more.
This is far too great, and it is further evidence that shows how broken Eclipse really is. Balance druids hold way too much control over Eclipse in the sense that we are capable of gaming the system in this nature to create extreme damage differences. It isn't a terrible thing, but it is rather telling that for one of the first times ever, we're seeing many of the top guilds run with two balance druids in their core raiding groups.
Two/too much fun?
Two? How absurd! I know, but consider that even in Wrath, it often wasn't considered to be beneficial to even have the one balance druid in the raid. Yes, our damage in the last expansion was more than fine. Yes, we didn't have to struggle claw and antler for every single raid spot that we got, but those druids who did raid as balance often did so because they earned that slot. There were few to no encounter mechanics that favored our DPS or for which our level of DPS so outstripped other players that guilds would go out specifically hunting for a moonkin. If your raid had one, great; if they didn't, they didn't.
To an extent, this is still true, but we're still seeing trends in the top guilds that prove how absurd Eclipse gaming can be. Sinestra was the prime example in this raiding tier. The world-first kill included two balance druids, both ranking in the top three in DPS. A vast majority of the following Sinesta kills by other guilds also included two, sometimes more, balance druids. In fact, multiple guilds have used two balance druids to handle all of the AoE on Sinestra and nothing else.
This speaks volumes to how far balance can -- and will -- go to abuse Eclipse. When a balance druid has over a 90% Eclipse uptime and DoTs/Wild Mushroom account for over 80% of their damage done on an encounter, something is likely wrong. And when you have this situation and no other AoE from any other player on the encounter other than the tank? Something is very clearly not working as intended.
Every week, Shifting Perspectives: Balance brings you druidic truth, beauty and insight ... from a moonkin's perspective. We'll help you level your brand new balance druid, tweak your UI and your endgame gear, analyze balance racials and abilities, and even walk you through PvP as a balance druid.
