Also, Nature's Grace seems to be applied just by hitting an Eclipse state (meaning that when you hit Eclipse, or when you hit the celestial alignment button, Nature's Grace is applied instantly). You don't need to cast moonfire or sunfire to proc the Nature's Grace effect. I didn't see that earlier.

The currently implemented version of Incarnation feels better than what the original tooltip on the calculator said (I was posting from my office with no beta access earlier today, and then only got a half hour at home before game night to test stuff, and now I have a half hour til podcast, lol). You basically have a bunch of things going on while it's up, rather than dragging out Eclipse, which feels better (both DOTs & nature's grace up thru the duration on training dummies). The in-game tooltip for incarnation has actually been changed to match the spell effect I was reporting earlier. I just haven't been able to dedicate solid blocks of time to mechanic testing.

Since my original argument was that I felt shorter times between Eclipses are better, this new Incarnation isn't actually a bad choice for when you know you will need burst DPS cooldowns. I still like soul of the forest, however, so we'll have to see how things actually work out with various encounters to see what ends up working best overall. I just feel bad for our poor little treants. They made them so much better but now we'll never use them. lol

Simc apparently has some spell data (they aren't saying they do, but it appears to be new in the MoP branch of their source control).

http://simulationcraft.googlecode.co...spell_data.inc

If I am right (and reading it correctly: Spell name->ID, then later ID->Coef) I see spell coefficients

Spell	MoP	Cata	Ratio
Wr	1.109	0.879	1.26
MF	0.227	0.18	1.26
SF	1.553	1.23	1.26
SS	1.548	1.23	1.26
Hu	0.262	0.095	2.76
Sfl	0.117	0.247	0.47
WMD	0.349	0.603	0.58

Single target spells ~26% stronger (to make up for lost talents?)
Hurricane gets much stronger (about 2 targets to match Wrath DPET).
Starfall and WM get much weaker (if nukes are your baseline, these lose more than half of their cata strength).

Single target, pre-Haste:
MoP: 10s of Starfall ~= 4s of Hurricane ~= 3 WM ~= 1 Wrath
Cata: 3.6s of Starfall ~= 9s of Hurricane ~= 1.5 WM ~= 1 Wrath

Speeding up Eclipse may be important just to get the mana to Hurricane, although MF still remains strong for multi-dotting.

MoP coefs copied by hand from link above. Cata coefs copied by hand from a January Wrathcalcs.

Edit: MoP Mage spell Blizzard appears to have a coef of .231, about 12% less than Hurricane. However, unlike Hurricane, Blizzard has a 50% snare. Blizzard costs 5% of base mana, compared to 75% for Hurricane.

I haven't paid attention to see if CoE is being removed like E&M is, but we're losing 10% spell damage right there as well as 6% from master shapeshifter. I *think* the new insect swarm is to compensate for that (if it ends up being implemented). The nerf to Eclipse's base damage was compensated by the increase to the mastery coefficient, so that's pretty much a wash. It looks like they'll definitely succeed in nerfing our AoE while increasing our single target DPS. I hope they address the Hurricane mana cost, though. Granted, especially with the new Eclipse mechanic I think mana won't be too much of an issue.

Just to clarify the mana gain from hitting Eclipse, each one grants 35k mana now, from a maximum pool of 100k.

Yeah, I understand about the new Incarnation, and I imagine it's probably a lot more fun too.

I'm still not seeing how you're modeling get that kind of a DPS increase from SotF. Let me try in more detail. While none of the following is exact due to the granularity of Energy gains, it's not going to be off by a factor of 2-3:
Non-eclipse phases are cut by 20%. Eclipses remain the same. If they're of similar length normally (which they are), you're in Eclipse 5/9 of the time now instead of 5/10: around 5% increase. A ~5% bump in Eclipse uptime can't be a ~5% DPS gain; Eclipse is much much less than a 100% DPS buff.

You're also proccing NG 10% more often, NG uptime will go up by a couple percent (<10%). So maybe 1% more DPS (to nukes) from there.

Mmm, with DoT uptime I can see getting to the 4% ballpark. I guess I should get my proper model working for these, even with L85 spells (that might be a few days though).

A pattern I've seen coming out of all this is increasing DoT uptime is a big deal. In reality we're going to have to evaluate single vs. double-DoT rotations in the first place so it's going to start getting a bit messier. Maybe I should see what I can start doing with the sheet now though instead of these estimates.

-------

For others not in yet, here are updated talents reflecting current beta:
Mists of Pandaria - Talent Calculator - World of Warcraft!
Spec abilities aren't in yet, but I know of no changes besides the vanishing of IS.

Here's probably a dumb question.

If something costs 75% base mana (ie Hurricane), is that before or after the 400% mana boost is applied from spec abilities?

I ask this because from what I understand with mana pools in MoP, pure caster classes (Warlocks, Mages), will have a very large base mana pool, as mana is no longer a function of intellect.

We, as druids, (Paladins as well), will have a very small mana pool to accomodate ferals, and our spec's abilities will compensate with the 400% boost.

So 5% of a mage's base mana pool is considerably larger than 5% of our base mana pool.

Just doing some napkin math here, I believe Tecton stated he had 100k mana, so our base mana pool is 20k.
Let's assume a mage actually has a base mana pool of 100k.

We're talking 15k mana for Hurricane, versus 5k mana for Blizzard. I'm pretty sure we get a ton more mana back over mages though, given how much Eclipse gives back.

Natural Insight increases the size of our total mana pool, not the base, so an ability that costs 100% of base mana costs 25% of total mana. Other caster class abilities cost comparatively less (about a 1/4th) base mana, but similar amounts of overall mana (e.g. Frostfire Bolt costs 2% of base mana, while Wrath costs 8%). The Eclipse explanation is probably the rationale behind the disparity between Hurricane and Blizzard, since Hurricane costs close to four times the amount of Blizzard, Rain of Fire, etc. I'm conjecturing that this increased cost is in order to remove the desire to stay in one Eclipse state for continuous AoE dps; without relatively frequent Eclipse cycling, you'll run out of mana.

Is Tranquility back on an 8-minute cooldown for resto druids now?

Probably an obvious answer, but hope dies last. Dream of Cenarius doesn't proc from white hits, does it?

My modeling is done via a simulator (I guess you've been gone for that part). It's not nearly as sophisticated as Simcraft, but I feel like I've done a good job capturing the rotation priority with it. That simulator generates a cast distribution table that indicates the percent of our rotation that we spend casting each spell in each possible state of NG and Eclipse. Those distribution tables get fed into Rawr; I'm using level 85 spell numbers and a T13-Heroic best in slot profile to compare numbers, since that's about all we have to go on so far. So while I agree that my modeling could well be off, they're not just numbers that I'm pulling out of a hat.

No, I know you were working on a model (and are probably further along than I am). Like I've said before I like having people working on sim-based approaches around as well the spreadsheet because it really helps cross-check various things. But whenever I see results I try to ask what causes them--I know it can be tricky to interpret from a sim but depending on what data it spits out maybe you can figure out how much of the benefit of some talent comes from Eclipse uptime, NG uptime, DoT uptime, etc.

Are you (or anyone else doing any of this) getting very short rotation lengths compared to what we're used to? I'm seeing well under 40s at even modest levels of haste when I make some sheet changes for new mechanics (more energy from W/SS, much less time spent casting DoTs, and assuming Euphoria is fixed). Leads to really high NG/DoT uptime, among other things.

Yeah, that makes sense, assuming your model is using one DoT cast per cycle (which so far seems a lot better than re-casting uneclipsed DoT's for 100% uptime--this is one of the effects of the short rotation). Something around 150% overall DoT uptime (Moonfire+Sunfire) just from your one cast each Eclipse.