(2) Set: Your Insect Swarm deals an additional 10% damage.
(4) Set: Your Wrath and Starfire spells gain an additional 5% critical strike chance.

The T7 set boni look very interesting...i wonder if it will be still worth to wear 4t6 or even 4t5 along with this. Just imagine your noncrit starfire hits for 5k in a raid. I haven't played beta but i think that sounds even a little too low bc i'm already hitting for 4,5k with owlkin frenzy. Now to compensate 4t5 you would need an equivalent of ~400 spellpower (or a mix of spellpower/crit/haste) when mainly casting starfire. I don't know if that's easy to come by.

Reaching back for this one, but my sheet does deal with this problem, capping the GCD (and hence wrath cast time under NG) at 1s. What it doesn't do, however, is account for the fact that you can't queue under GCD, which means NG'd wraths actually suffer from double latency like everything used to back before 2.3. In fact, now that I think about it, this could actually make NG a DPS loss at high latencies. I'll pop it into the sheet along with some of the other fixes I'm working on and see what happens, it might warrant poking the Blues about.

[edit] Forgot the point I was going to make. Even with the hit from capping the GCD, Wrath still has a higher DPET than any Starfire without Eclipse. I'll see if that ends up still being the case.

I'm pretty sure Owlkin frenzy does not have an ICD either. I just respecced for this talent and I am enjoying it. However without the last 4 sunwell bosses' WWS reports I do not know how much my DPS was increased.

Thank you for that info Treehugger.

Nothing on T8 yet? Since maybe like in tbc, T4 didn't last long. And many random items before T5 were better then T4.
So with T7 IS will have a max benefit of +40% from glyph and set. And most likely people will keep T6(2) as long as possible for that extra moonfire tick if the glyph will remain that way.


All that aside, the Blue's should be all over these forums instead of the official whine forums to start with
- Wrath doesn't feel right/fun to be casting.
- The glyph problem needs an alternative solution.

It has no ICD (or a trivial ICD)

00:38'55.281 Erdluf's Owlkin Frenzy is refreshed.
00:38'55.672 Erdluf's Owlkin Frenzy is refreshed.
00:38'57.406 Erdluf's Owlkin Frenzy is refreshed.

The four-piece bonuses won't be worth the stats lost by wearing them. Not only do you lose a significant amount of spell power, but the differences in crit/haste/hit rating values of the pieces will quickly surpass the additional 5% Starfire crit rate (especially when you consider that wrath is going to be a part of our cast rotation). Certainly if you have all the sunwell T6 pieces there may be a point in gearing up where you have both four-piece bonuses without sacrificing stats for it, I'm sure I'll be at that point for at least a little bit.

The more interesting bonus is the T6 two-piece. I can see myself using bracers + belt and upgrading them last as the paucity of sockets in non-set pieces makes these items very sustainable and quite a bit less of a stat loss. One thing you can't forget is that the 25-man Naxx pieces have almost double the amount of intellect and spirit on them which does contribute to our DPS as well as mana efficency.

In agreement with a previous sentiment expressed regarding the Starfire glyph changed: while it was a significant nerf, the devs have consistently expressed their intentions to regularly introduce additional glyphs. Unfortunately, no Starfire glyph would ever have been able to surpass the original one without introducing some serious power creep. I believe the mistake they made was not making the original glyph (OG?!?) a talent in line with SPriests and Affliction Locks; perhaps we'll see this in the future. Not trying to wishlist here, but I can't help but think that IFF should have been made a glyph with moonfire refreshing taking its' place (Adoriele: I wonder if there's still time to suggest this in the last week the class devs are really responding to concerns...I don't expect this level of feedback/communication to be maintained into live)



-


Some interesting thoughts I had last night during Sunwell about future raid encounters . Finally got an opportunity to make some attempts at Mu'ru. I'm not sure how many full balance druids have actually DPS'd this fight, and don't know any that regularly post here so I hope my impressions will be a new perspective.

Phase 1 Mu'ru as an encounter is almost custom-designed for all of our strengths in Wotlk, with multiple targets to cast dots on and regular usage of aoe. What really left an impact on me is that I was running out of mana with Mu'ru at 50% (Where I actually had to use my innervate), and that's with mana return on crits still working on live. This is also a fight where you simply aren't going to get the benefit of JoW or most mana tides.

I firmly believe that we're still going to be using money to respec regularly. Any real raider can see the huge benefit of using our dual-spec for resto just to make it through content (We barely need to switch gear anymore to be effective, and resto needs that free second glyph loadout). But if we truly want to min-max our performance we will have to have multiple (Probably three) DPS specs. There's no way I could make it through an unnerfed Mu'ru style encounter at 80 without all of my mana regen talents, and you have to consider gale winds + brambles are exponentially better under these conditions.

Now that we're starting to establish some solid thoughts on single target fights, I just wanted to explore how much our playstyle and mana consumption changes as the encounters differ in nature. Genesis in particular is a talent with almost no value in normal situations that becomes enormous if we can keep dots up on three targets (Potentially doable in a fight like Four Horsemen).

Eclipse is nearly impossible to use properly on a fight where you switch targets or have to go immediately into AoE mode. We've looked at the numbers enough to know that eclipse is already an ability that borders on marginal if you cannot make use of every second. One particularly frustrating thing is the non-100% proc rate for the Starfire buff. I was wanting to utilize wrath to proc Starfire for mana regen purposes (As well as DPS) but the mana inefficiency of having to cast 5-10 wraths for a proc really puts a damper on that.

There is a haste inflection point, but it is so seriously high that the only time you're going to hit it is full Bloodlust+drums. I played around in Rawr to see if I could get it to switch over, and indeed it does, at about 45% haste. In other words, Wrath is a higher DPS spell than Starfire under any circumstances short of almost fully clipping the GCD.

Haste isn't necessarily the limiter, though. With haste, you're (likely) never going to see non-NG wraths get capped. If, on the other hand, you increase your crit, more of your spells on average get capped out.

If Wrath is becoming the basis of what we cast, would it be worth speccing out of NG?

Drums no longer work at level 80, and there are no new drums.

No.

Nature's grace is a massive boost to wrath damage even though the lowest the global cooldown can go is 1 second.

I tried two test cases. One had high crit, and relatively lower haste. One had high haste and relatively lower crit. The crit chicken was unable to push it to the point where Starfire was superior, even with full raid buffs, Bloodlust, haste pot and drums. The haste chicken was able to hit the Starfire inflection point with Bloodlust alone. The difference in crit between the two is approximately 5%. I should add that the crit chicken did also have lower spell power, so that may also affect the results.

Lord BEEF: I know drums no longer work at 80. It was simply to illustrate the extremity of the haste inflection point. It's really way out there, depending on your gear.

[EDIT] I have to correct the record: The exact differences between the two test cases are as follows.
Haste chicken has 12.5% hit, 28.85% crit, 47.95% haste, 1486 spell power in the full test case.
Crit chicken has 14.6% hit, 29.85% crit, 46.03% haste, 1405 spell power in the full test case.

So the crit chicken only has a little more crit, but loses nearly 2% haste, which may make the difference. My apologies.

[EDIT 2] Eclipse is really messing up the math, as well. It seems that Rawr highly values the Starfire Eclipse proc, to the point where it's clinging to the IS/MF/W rotation a lot longer than I would expect. At any rate, there's too many variables in play right now for me to say with any finality what's going on. All I can conclude is that it takes a WHOLE LOT of crit and/or haste to make Starfire worth it.

Ah, okay. I knew there'd be problems with both, but hadn't napkinned which was worse yet. Also, Beef, it's actually quite possible that NG ends up being a DPS hit with high latency, and a wash with low. I still think it's worth speccing, unless someone really loves their brambles, as you're going to have periods where you're casting SF no matter what, and NG is a great help there.

Perhaps my modeling is off, but I can't get this to occur even when I manipulate the config.

I changed the default raid_80.txt config to have two Balance Druids. I changed the end of each action priority list such that one ended with Starfire and one ended with Wrath. Procs from Ecclipse were ignored, under no circumstances did the Starfire Druid cast Wrath, etc.

Starfire Druid came out on top easily.....

I thought perhaps the high levels of haste in the config was unfair, so I set it to zero. Gap reduces but still large.

I removed the Starfire Glyph. Gap reduces but still large.

I set lag to zero. Gap reduces but getting smaller.

I remove all the other players from the raid as well as the Pally buffs. Gap still there.

I set infinite_mana=1 to take mana regen out of the equation. Gap reduced to 3%.

The Starfire Druid always comes out on top........

If this is against the prevailing wisdom, I would certainly appreciate a code review of starfire_t and wrath_t within sc_druid.cpp

I do M'uru as full Balance Druid. I am focussing on the Void Sentinels there. So basically I start nuking M'uru, when a Void spawns I switch to the Void and when he is dead I AoE the little ones, then back to M'uru. In this fight I cannot use Eclipse because of the switching of targets and nuking/AoE'ing. I didn't run out of mana, I have that problem for the first time in the KJ fight. I had to slam a pot even in that fight . But I agree, M'uru is fun, and when you start dotting multiple targets, and AoE in between, mana can go fast.

direct_power_mod = ( 1.5 / 3.5 ); Line 935


That's incorrect, you used the modified cast time. Should be 2.0/3.5.

For both Starfire and wrath you don't define Master shapeshifter or Earth and Moon. I examined mage spell code and their spell talents all seem to be there, and I couldn't find anywhere else it might be defined.

base_multiplier *= 1.0 + std::min(p -> talents.moonfury, (int8_t) 3) * 0.10/3;

Should be one similar for Earth and Moon and Master Shapeshifter talents under Starfire, Wrath, IS, and Moonfire. I looked at the Earth and Moon code for triggering the debuff, and it doesn't appear to be rolled into that. You might also want to adjust it as a 3 point talent worth 1/2/3% spell damage. Master Shapeshifter is 1 point for .02. These are definitely multipliers and not direct power mods. It's weird because you have Moonfury updated to be 3 points, but in your Earth and Moon debuff code you have it at 5 points still.

I think your overall Moonkin dps numbers are significantly low because of this. You may also want to add an event for the T7 10% insect swarm 2 piece bonus. I -believe- it is a base_multiplier, but I haven't seen anyone here that tested it.


Hope that helps. I'm not a coder of any sort but about twenty minutes looking through everything gives me an idea of how you have things set up, those were the things I found that stood out. I also looked at the NG code and double checked your spell code to make sure it has a min GCD. I really can't tell you if NG is breaking the GCD because I can't actually run the sim, but I imagine you'd know if it was doing that

In any post-3.0.3 rotation, it seems like there should be starfire and wrath both being cast at different times. The benefit from the starfire idol makes it worth trying to cast starfire three times after a moonfire to get an extension before you have to recast moonfire. In an Eclipse build, you'd have to cast both wrath and starfire (at different times depending on what Eclipse effect you want). Murmur's math on the Live forums seems to support using starfire to proc the Eclipse effect, and then casting wrath until you have to refresh moonfire or the cooldown of Eclipse comes back up (in which case you'd starfire until Eclipse procs again).

Is there an effective balance build in 3.0.3 that doesn't use both starfire and wrath?

Issues I'm noticing:
-You're making a lot of rounding errors or just plain lazy calcs. The benefit from E&M isn't talents*.13/3. It's {0,.04,.09,.13}. Granted, any Moonkin worth his beans will be fully talented or not talented at all, but it's still not the best way to model it.
- There is no OoC cooldown, proc is PPM-based, likely 3.5/60*untalented cast time.
- Eclipse is 15s buff, 30s cooldown
- Insect Swarm's coefficient is ticktime/15 * .95, to account for affecting miss chance. Also, you're multiplying Genesis and the IS glyph. They stack additiviely.
- Your coefficients for Moonfire are close, but not exact. It's enough to cause deviations with the amount of spellpower an average 80 will have. You've got iMF and Moonfury stacking additively here, but you're multiplying in the effects of the glyph which, again, is additive.
- As Err said, Wrath's base coefficient is 2.0/3.5, not 1.5/3.5, which is likely causing Wrath's low damage.


Err, Master Shapeshifter and E&M's personal benefit are included in player_buff()

[edit] No. About the only one I can think of is if you don't have the mana to sustain a wrath-centered non-eclipse build, you'd use Starfire due to its higher mana efficiency. However, this should only be necessary if your guild is full of idiots who only have one replenishment source in a 25-man and pallies which won't judge wisdom. No, I'm definitely not bitter about this.

[edit2] Having two moonkin whose names start with Er is a baaaaaaad idea.

I'm looking at r819 source. Both the personal E&M, and MSS boosts are included in player_multiplier.

Good catch on the Wrath scaling.

dedmon, you might want to add commented-out explanations of "log=1" and "debug=1" to your default input files.

Edit: Adoriele, credit for the first code review goes to Eregal, not me.

Edit2:

MF coefficients.

DOT: r819 use 13%. Wowwiki formula for the DoT gives 13.023%. My testing puts the value between 13.048% and 13.063%. simulationcraft should be extremely close for reasonable spellpower levels.

DD: Simulationcraft uses 28%. Wowwiki formula gives 14.95%. My testing bounds the value between 14.77 and 15.15%. 28% looks really bad.

I have one question here to raiding moonkins.

How high is your threat compared to tanks. For example dps races like brutallus. Given the new tanking patch, at least when I'm playing feral, no-one can even come to close me with threat.
If tanks keep scaling so well in Lich King (and I think they do). We could free 2 talent points from Nature's Reach!
That would free one point to Moonglow and one further the tree.
The range buff +6 yards shouldnt be a problem, its hard to be in max range anyway if the target is moving.

The aggro is A lot better now. Before patch 3 I was mostly blocked in damage output because I had to stop dps. Now I can just go full out without having to watch omen for the boss duration.

Nature's Reach is more about the extra 6 yds. There are boss fights that require you to be a certain distance away (Anub'rekan and Sir Zeliek come to mind, although I've mainly done them at 60). I personally am not getting very close at all threat-wise in Sunwell, but I am still using 4 pieces of resto gear so my crit rate isn't where it should be.