Flaws with the current system
If you happen to avidly read The Moonkin Repository or Elitist Jerks as I do, then you've probably come across this great post by Hamlet. (Fair warning that the post in question is all calculus, but don't worry if you aren't a math person or if such figures make your head spin.) Though there are a few comments about the function itself, as there will always be on a site like EJ, there really isn't anyone disagreeing with the core concept that Hamlet shows: that barring further information, the current iteration of Eclipse is rather broken at this point. Although we don't know even the most basic of information as to exactly how Eclipse is going function, the results of the formula are still solid. Essentially, what Hamlet demonstrates is that our rotation consists merely of rapidly alternating between two spells. Going by the limited constructs of what we know of the system, actually switching the scale between the sun and moon is simply not going to happen. Instead, balance druids are going to get the scale to a specific point and then hover at that point for the duration of an encounter -- meaning the rotation would essentially break down into some version of Starfire > Wrath > Starfire > Wrath spam, with the only variation being additional spells that might be required to keep the scale in that location. It could end up being Starfire > Wrath > Wrath > Starfire or any variation thereof.
At the core of the concept, I would call this system broken. Although Ghostcrawler has stated that they don't want the balance druid rotation to fall into a systematic 1111222211112222 set rotation, that's exactly what we are being pushed into (although it would be closer to 12121212 or 122122122). To an extent, this isn't such a terrible thing. After all, many classes have somewhat simplistic rotations such as this. Fire mages, for example, follow a rotation that is more akin to 122232232232322312223 or whatever variant, depending on Hot Streak procs. It's a bit random yet overtly simplistic in nature, rotating only between two nukes and a single DoT.
From how the development team talks about balance druids, though, this doesn't at all seem like their intention for balance druids. They want us to shift between our nature and arcane spells, and to that extent Starfire and Wrath, but they want us to do so at semi-random intervals. There would be very little that's random about the concept as it stands, however. Artificial constructs placed on the mechanic can create some random variation to a degree. For example, if crits shift the scale by a higher amount than normal hits, then the rotation might end up being a slightly more reactive 12112112121121212112. The flaw is still there that we're not actually "switching" between our spells. All we end up doing is alternating between arcane and nature abilities in a predetermined, high frequency pattern.
There is still some hope for the system. Another part of the preview included a new spell, something that balance druids have been asking for since time began, that is going to have a special interaction with Eclipse.
The importance of Nature's Torrent
New Talents and Talent Changes
Balance druids will have a new talent ability called Nature's Torrent, which strikes for either Nature or Arcane damage depending on which will do the most damage (or possibly both), and moves the Eclipse meter more (details below). The improved version of Nature's Torrent also reduces the target's movement speed. 10 second cooldown.
Just as Eclipse is the leg on which balance druids stand, Nature's Torrent is going to be the crutch that supports Eclipse. From a perspective of balancing damage, Nature's Torrent is completely unnecessary. Balance druids could easily be balanced around a simple Eclipse model. Doing so would just leave our rotation rather flat. If there is to be any depth to Eclipse at all, then it is going to come from Nature's Torrent.
Let me clarify that slightly. Certain basic mechanics in how Eclipse should end up functioning could cause slight variations in casting sequences, but they will be minimal at best. One such mechanic would be if critical attacks move the Eclipse bar more than normal attacks, then that would naturally cause the rotation to periodically shift by, for example, only causing you to need to cast one Wrath in between a Starfire instead of two. It wouldn't be a significant shift, and certainly not enough to make the rotation "interesting" by any means, but it would be there.
The other factor that cause potentially cause small changes to the rotation is movement, or rather, Wild Mushroom. Wild Mushroom is another new spell for druids in Cataclysm
which is predominately going to be AoE-centric but have some single-target applications. The spell essentially acts as a natural landmine by conjuring a little mushroom (or maybe it's a big mushroom) that eventually activates and becomes an explosive trap of sorts. Although that itself has little to do with movement, the key is that Wild Mushroom will be instant and druids will have the ability to trigger the effect on their own, presumably by summoning another mushroom. This means that druids on the move can just spam Wild Mushroom to supplement their damage until they can free cast again. Since Wild Mushroom does nature damage, it would cause the Eclipse bar to shift and alter the rotation slightly.
That tangent aside, Nature's Torrent really has to be the crux of Eclipse if Eclipse is to be random or fluid in any way. Without this, our rotation will actually be just that: a rotation, a set sequence of predetermined abilities that has little variety. For any of this to work, there are several facets of Nature's Torrent that have to function in a specific way or the spell becomes, for lack of a better word, useless.
Nature's Torrent damage capacity
Damage is always key, and this is certainly not going to be an exception. How Nature's Torrent interacts with Eclipse is all well and good, but it is nothing without the damage of the spell. If the spell is weak, then the Eclipse effect has to be absurdly strong in order for us to consider using the ability. How the damage is balanced is largely going to be a function of how the spell is intended to be used and the impact it is to have on our rotation. Even still, utility spells (spells that are not predominately used to deal damage but rather to alter other spells) generally don't end up being all that useful or scaling all too well. That's not to say that they can't work or that they can't scale, just that history has shown in the past that these things usually end up being true. Spells that exist merely for the benefit they give to other spells or effects are generally viewed as a weakness or a flaw within the class/spec/rotation. Think of Scorch
. Go ask any mage about Scorch, and I doubt you'll find many who actually like casting this spell. The damage is terrible, and a mage would really like to cast pretty much anything else, but the utility of Improved Scorch
is too great to give up, so they cast it anyway.
Nature's Torrent should not fall into this category. It doesn't have to be our hardest-hitting spell that we have. The damage that it deals doesn't have to be so significant that we want to use it on every cooldown, as an elemental does with Lava Burst
, but it shouldn't be negligible, either. The damage should be just enough so that it and whatever utility that Nature's Torrent provides for Eclipse feel like they are worth using. If Nature's Torrent does 100 damage, even if it made your next Starfire deal 100,000 damage, it would feel crappy to use the spell. It would be worth the cast to use and a DPS increase; it just wouldn't be "fun" to use. It would be much better to have Nature's Torrent do 40,000 damage and have your next Starfire deal 60,000 instead. Then it would feel like the spell has an impact, that using the spell is worth it, that the spell is fun. Obviously, those numbers are just random examples and not anywhere at all what I think or suggest Nature's Torrent should be doing, but it gets the point across.