spartakos

So you have a 27% chance to proc a 60% chance to give starfire a 30% chance to crit for about 12-14 sec of the bloodlust.
No, do not gamble with your damage.
I just spam starfire keeping moonfire dot up (have IIS) but still do not know if its worth to change for wrath when eclipse procs.

 

dukes

In a 25 man raid, assuming all classes, you get:

+10% crit (imp scorch/winters chill);
+3% crit (totem of wrath/heart of the crusader);
+4% crit from talents;
+5% crit from moonkin aura (+0.5% with the idol of the raven goddess); and
~1% crit from intellect buffs (arcane int/bok).

This totals ~23% over your base crit rate. Taking a base crit rate of 15%, adding in IFF and Nature's Majesty, you should be looking at approximately 45% crit rate on all spells, 55% crit rate on moonfire, and 48% crit rate on starfire while Moonfire is active (if you have IIS).


Personally up to now I have used wrath to proc eclipse even during Bloodlust. I haven't actually done any kind of theorycrafting behind it, but now that the question has been asked, I'm intrigued and will investigate it further.

 

Daylis

Another thing i'm wondering about and havent come across an answer yet is moonfire/is reaplication on eclipses. Say i've a wrath eclipse up, it's just started and will last for 15seconds. IS (glyphed and 2t7) and Moonfire (2t6) have just fallen off the boss.

Is it worth recasting those at the cost of a gcd each, or just spam wrath until eclipse ends. Lets presume 2/3 ISS talent.


IMO it's worth keeping dots up even during eclipses, i've no mathematical proof for it though. Just an educated guess.


This brings us to another thing.. Say you've both IS and MF rolling when [Sundial of the Exiled] proc's, adding 590 spelldamage for 10 seconds. At what duration is it worth recasting dots and clipping existing ones for them to gain extra 590 spellpower for their whole duration?
Pretty safe to assume Nature's Splendor with this. And, since the proc duration is 10seconds, it's also logical that dot's will be over half of their duration already, especialy IS.

When this happens to me, i usually reapply them, but i'm far from certain that's the optimal solution. Some math would be more than welcome, unfortunately, i suck at it.

 

Angelfire

@ Daylis - It was posted before that it is not worth it to clip dots in general. If you have to clip it (so that they will receive the proc buff for example) try and delay it as much as you can, preferably until the last tick.

@ Spartakos - If you just spam Starfire and ignore Eclipse procs, why put talent points into it? Just put your points elsewhere where they will actually be useful.
It has been posted by Thedopefishlives that the best rotation with Eclipse is MF/SFx3/Wrath to proc eclipse/SF till internal cooldown on eclipse ends/Wrath to proc and so on, applying MF as needed of course.
However, if I recall correctly, this did not produce much more damage than using SF to proc eclipse for Wrath, it's just cheaper mana-wise.
Both rotations give better DPS than not having/ignoring eclipse, but eclipse is not for everyone.

 

erragal

There's also the option of Perfect Purified Shadow Crystals. You lose 1 spell power and 1 spirit off the NYI Twilight Opals but it's an overall better option if you're hit capped.

 

thedopefishlives

Under Bloodlust, my math shows that it's typically best to switch to Starfire. It also shows that casting Wrath when Eclipse procs is still increased DPS, even with the GCD clipping.

 

Erdluf

A DoT refresh probably costs you one Wrath under Eclipse (could be two). That penalty is 20% of a non-eclipse Wrath. I think your DoTs will make that up (DoT DPS - Wrath DPS) if they get an extra 6s or so. During the first half of eclipse, DoTs are certainly worth refreshing. During the last few seconds, cast your last nuke, or two.

Raid scaling for IS with Glyph, 2T7, Splendor, MS, is about 1.5. Scaling for Wrath (after including crit, Moonfury, etc.) is about 1.25. If you cast IS, you will cast one fewer Wrath, meaning you only gain about 0.25*590 extra damage (about 150). Since IS ticks are > 150, it isn't worth clipping IS at all.

MF scaling (Splendor, Glyphs, no 2T6) is about 2.45. Subtract the scaling from the lost Wrath, and you get 1.2. Your benefit from casting MF during the buff is about 700. That should be less than your MF ticks, so it is not worth clipping an existing MF.

Scaling numbers:
IS = .76 (base) * 7/6 (Splendor) * (1+.3 (glyph) + .1 (2t7)) * 1.13*1.03 (E&M) * 1.04 (MS)
MF = .13 (base/tick) * (1 + .75 (glyph) + .1 (Moonfury) + .1 (IMF)) * (4 (base ticks) + 1 (splendor) + 3 (SF glyph) * 1.13*1.03*1.04 (E&M+MS)
Wr = (.57 (base) + .1 (WoC)) * 1.4 (crit) * 1.1 (Moonfury) * 1.13*1.03*1.04 (E&M+MS)

I did not include the IIS bonus to Wrath, and my Wrath crit number may be a bit low. Both would favor another Wrath cast over clipping one of your DoTs.

If you are moving (no opportunity to cast Wrath), it is probably worth clipping a tick or two of your DoTs, because refreshing while moving means you delay the next refresh (while not moving).

Edit: Math. (590 * 1.2 != 900). Also, it could be worth clipping one tick of either DoT if you don't have time to cast Wrath (there is less than 1s left on the buff).

 

Woodlum

The Eclipse-Twist idea that was brought up is an awesome thought, and really only for your DPS perfectionist (but that is also the thought of Eclipse itself to date) as you won't lose much by finishing back-to-back casts before spamming your opposite vs a twist method. The point that seems to be lost on the discussion for Twisting is that you only should "Twist" when the ICD (Internal CoolDown) is finished on Eclipse, then revert to repetitive casts/rotation. To complete the thought, when Eclipse finishes, go to your normal spell rotation until the ICD is done again.

You might choose your lead-off twist based off preference (non-Opener), top-end of Eclipsed Wrath vs Starfire (which I don't think is clear at T7 content which makes twisting a good idea in my mind).

I think "spell phases" is a term worth considering/coining here in the community:
- Opener (opening rotation on a standard boss fight, something like: throw up DoTs -> cast Wrath for quick E&M),
- Twist (alternate Wrath/SF until Eclipse proc),
- Eclipse (spam Eclipse empowered nuke),
- DoT Check (renew DoTs, depending on your preference to clip or not, leaning towards not),
- ICD Rotation (standard SFx3 or whatever)
- Repeat starting from Twist (ICD on Eclipse is over)

- Movement (cast instant spells while on move and pick up with Opener once settled).

A mod tracking the ICD is really going to be king in a twist environment and twisting may ultimately be easier or harder on the human and situational awareness factor in the long run which expands consistancy in damage, worthy of debate but will leave it for someone else to mull over for the immediate moment while I try gaming through the thoughts I jotted down regarding cast sequence and cast time management.

----

On a seperate note, I see that Owlkin Frenzy is something debated for this current stretch of content and I too believe in the counter-arguements in putting points in it. I also wanted to offer that I see many of the Moonkin here throwing points in IMoTW for their raiding specs and would humbly submit that 2/3 Nature's Focus for your standard resto-points investment would be the better option for some pushback assistance (wrath only) if pushback is a concern at all. And this is speaking to raiding only as I suspect we all have (a) Resto Druid(s) who take IMoTW as sort of a mandatory investment to get deeper into their tree, it would seem 2/3 NF is more useful if a Resto Druid is dependable for the buffing.

----

Also, it seems to me that treants are not getting a +hit conversion from us. Does this seem to be a correct observation from anyone else? And if so, this is something that needs attention from the devs (GC) as everything else pet wise is getting that benefit to date, and the improvement to treant dps (which is already very nice) by being able to connect more consistently would be appreciated.

----

And as an after thought with the Bloodlust/Heroism, I would recommend spam casting SF during this time no matter the situation, maybe even see if you can time it to lead up to a Wrath induced, SF favored Eclipse proc before. The easy way to look at the issue is... The cast timer is specifically the difference here and I believe Wrath hits a time-reduction wall so ultimately loses benefit where as SF doesn't so it ultimately sees more gains from the hasting effect. That ideology at least makes sense without having to math it out.

e: (sp)

 

Eilt

My treants usually end up with 6-10% missed on the night as well, would be nice to see if this is intended or what.

On misses though, most of us end up at ~14% hit, then there are those of us with spriests who do not have iFF that are pretty much depending on misery debuff. What I find is that this leads my initial MF to miss occasianlly (~1% on an entire night). Would it be worth it to give the spriest a few seconds to set up and get in an attack in to put up misery and for me to not cast MF while running in, or should I just recast the MF as it misses?

Right now, with mana regen being so easy, I think that getting MF up while moving outweighs the potential to miss by not having the misery debuff.

e: typos

 

◊ Lord BEEF

I typically do insect swarm then moonfire while running in to avoid the potential for an initial aggro pull with a big nuke crit. I see no reason to wait for the shadow priest to get misery up though. The worst case is that there's a 3% chance that you'll have to recast a spell.

 

• Arawethion

I've just switched to IS->MF on the initial run-in for a different reason (I don't think it's too realistic to worry about pulling aggro with the initial hit of a glyphed MF). By casting MF second, have time for one extra Wrath to try to get an Eclipse, before I have to switch to SF to extend the MF.

----

To be clear, my rotation generally looks like:
1) Start->IS, MF, W, W, W
2.A) if a "lunar" Eclipse procs, spam SF for the duration. When it ends, recast IS->MF->SFx4-5 (until Eclipse ICD ends), and we're back at the beginning.
2.B) if <3.5s left on the MF DoT, and no Eclipse, cast SFx3.
2.B.i) if "solar" Eclipse procs, spam Wrath for the duration while refreshing DoT's. As soon as it ends, cast SF and you should extend the MF you cast during Wrath. SF through the ICD and recycle.
2.B.ii) if still no Eclipse, refresh IS and you're back at (1), casting Wrath.

Since the goal is to get lunar Eclipses, you want to minimize the situations where you have to cast SF to extend an MF, while Eclipse is off ICD. So the ideal case is to have a fully-extended MF ticking right as Eclipse comes of ICD, so you can Wrath 9-10 times (refreshing MF and IS once each) to try to get an Eclipse, before you have to SF again. This often happens naturally if you refresh IS and MF (and start casting SF) immediately after an Eclipse ends, as Eclipse will come off ICD only a few seconds after you cast your 3rd SF.


Open questions after trying this for an evening (I've not yet played with Adoriele's spreadsheet):
--Cast MF if it expires during a (lunar) Eclipse? (I think, no)
--Cast IS (glyphed with 2T7) during a (solar) Eclipse? (I think, yes).
The idea here is that a GCD during a lunar Eclipse is very valuable, as you want to squeeze in as many Grace'd Starfires as possible.
--Is it even worth casting the SFx3 to extend MF if you haven't procced Eclipse with Wrath yet? People are assuming yes, but has anyone worked it out? Getting the "wrong" Eclipse in any cycle seems like a big damage loss (probably greater than the ~3k from 3 ticks of MF).
--Behavior during Bloodlust, as Beef asked above.

Answers to Moonkin questions:
0) Read the TTT/use the spreadsheet: http://elitistjerks.com/f47/t66856-moonkin_pve_dps/
1) Maintain high DoT uptime. Use WiseEclipse.
2) Nothing beats 2T8.
3) Yes, sometimes you cast many Wraths and no Eclipse procs. Deal with it.

 

Rosoo

I haven't been modifying my casting rotation based on lust. Raid buffs and debuffs should be enough to ensure you get lucky most of the time with a quick proc. I'm also gonna edit the first post to also have my WWSs, so if anyone wants to use a different casting rotation on lust and compare, they have plenty of fights to do so with!

 

Rosoo

I have been recasting moonfire during lunar eclipses. I have been suspecting this to be wrong, and I'll try not doing it for the rest of the raid week.

From my understanding, currently casting IS at all right now is up on the fence for whether it's worth doing while standing still. Casting it while solar eclipse is up seems like a bigger waste of dps than casting MF during a solar.

I currently don't cast starfire at all if the ICD on eclipse has finished, fight mechanics permitting of course. The last thing I want to do is get a solar eclipse and be stuck wrathspamming. Sure, it's not a terrible DPS loss, but hey, every bit counts.

 

zinc

i did some testing on the training dummies with both idols
used 2 different rotations self buffed, no innervate

Starfire spam rotation
IS > MF > Wrath (till Eclipse proc) > MF (if necessary) > Starfire (till Eclipse fades) > IS > Wrath (till Eclipse CD back up) > Repeat

Wrath spam rotation
MF > IS > Starfire (till Eclipse proc) > IS (if necessary) > Wrath (till Eclipse fades) > MF > Starfire (till Eclipse CD back up) > Repeat

Wrath spam with Wrath idol
1. 3 Eclipse procs 2797 dps
2. 3 Eclipse procs 2831 dps

Wrath spam with Starfire idol
1. 2 eclipse procs 2602 dps
2. 1.5 eclipse procs 2503 dps

Starfire spam with Wrath idol
1. 4 eclipse procs 2772 dps
2. 3 eclipse procs 2468 dps

Starfire spam with Starfire Idol
1. 2 eclipse procs 2367 dps
2. 2 eclipse procs 2390 dps

i dont really care for the Starfire rotation just cause i always have to refresh IS when eclipse procs to take full advantage of IIS whereas the starfire glyph fixes that in the other rotation

edit: wrong spell, doh

 

Marauding Master

And by Starfall, you mean Starfire.

Your Starfire rotation is bad because you're not optimizing it. Casting IS is a DPS loss and refreshing Moonfire during your SF Eclipse is an even bigger DPS loss. iFF is still worth it, but cast IS only during when you have to move.

 

erragal

IS with just the glyph (Not even including 2-T7) is better DPCT than Starfire at any current gear levels. Moonfire is nearly 50% more DPCT than Starfire if you're using Starfire + Moonfire glyphs. Even if Starfire(Eclipse) put you at 100% crit rate you'd still be losing DPS by not reapplying moonfire during it. The only time you would not reapply moonfire is if you know it won't tick the full 24/27 seconds.

Please don't give out incorrect information that's based on your assumptions rather than the actual math.

 

Eilt

I honestly thought Ihad read here that refreshing your DoTs during an eclipse was a DPS loss...Will look around and see what posts I can find

 

erragal

Insect Swarm can be depending on when you do it during the eclipse, and which type of eclipse. I don't refresh IS during Starfire Eclipses or while under bloodlust. Not refreshing IS during Eclipse:SF is purely a risk up until SF: Eclipse puts you at 100% crit (At which point it's always the right decision). You're gambling on the RNG before that point. I ran all these numbers manually at 2200 Spell Power, 20% haste 60% SF crit with the Starfire Idol to reflect good gear and give Starfire its' best scenario (Spell Power is actually a bit low favoring Starfire even more). At lower crit/haste levels 2-T7 IS will be a DPS increase even during bloodlust.

IS during wrath eclipse is a bit different because it also improves the wrath damage, but Erdluf has done those numbers at least twice now showing the range where it's a DPS increase. Refreshing moonfire during Eclipse: Wrath should always be a DPS increase.

Moonfire is still higher DPCT even in a bloodlusted 100% crit Starfire situation. That's something most people will probably miss in practice because it's very easy to get caught up in the big fast numbers.


One other observation from live raiding: I don't believe Typhoon is always a DPS increase when you move. In many fights you need only make small position adjustments such that you aren't done with the GCD from casting it by the time you are able to go into a wrath/starfire. Even something as low as waiting .5 seconds in GCD to begin your next cast ends up being a wash with Typhoon in terms of overall damage output.

 

Erdluf

The real answer is that it makes sense to refresh DoTs if they fall of early during Eclipse, but if they fall of late in Eclipse you should cast one more nuke.

Suppose you are at the end of Eclipse and your choices are:

SF (eclipse) MF

or

MF sf (non-eclipse)

By casting MF first, Over the course of the fight, you have gained at most one MF tick (possibly not that much). Your expected "loss" is about 30% of a non-crit SF.

If the refresh is earlier in Eclipse, your gain goes up (more MF ticks gained, also more SF getting IIS bonus), but your loss remains the same (still at most one SF pushed out of Eclipse).

Every time I've tried to run tighter numbers (for either Wrath Eclipse or SF eclipse) I see results that say always refresh during the first half of eclipse, never refresh during the last quarter of eclipse (unless you don't have time to finish your nuke anyway) and from 1/2 to 3/4 duration it just doesn't matter very much. The break-even point seems to stay between 4-8s remaining on Eclipse.

 

• Adoriele

I'm going to see what I can do to add a sense of this into WC, but it's pretty spot-on. You're really looking at comparing 1 GCD's worth of the benefit to your nuke (i.e. SFAver*.3*GCD/SFCast) to the number of ticks you would lose by waiting until the end of Eclipse.

 

erragal

I understand a bit better what you're getting at, but it goes decidedly agains the concept of always utilizing your highest DPCT ability at all times. I had to think of it as sacrificing the single tick of moonfire damage for the 30% crit buff on one Starfire and look at how that changes average Starfire damage to get a better sense of it. Apologies for that. Have to look at it from a different perspective to get that.

 

• Adoriele

It's not necessarily best to use the highest DPET ability at all times, though it comes close. For example, if Starfire is a higher DPET than Wrath (which it usually is), and Eclipse exacerbates that instead of switching it (you generally get more benefit from a SF Eclipse than a Wrath), you want to cast Wrath to proc Eclipse, even though it's a lower DPET.

 

lissanna

I almost wish we could go back to the days where we just did moonfire/starfire (repeat forever), and had a clearly highest rotation above all others. The way Eclipse works, combined with the GCD clipping problems with wrath, I really don't think there is a highest DPS rotation that stays the highest at all levels of gear. Maybe there is so much variability within a rotation based on the RNG that how you manage the rotation may have a bigger impact than what the rotation is. I mean, what I hear from people as their highest damage rotation seems to change on a daily or weekly basis, that I'm not even sure I trust damage meter reports or theorycrafting anymore.

 

zinc

i modified the rotations slightly to test

Starfire spam rotation
IS > MF > Wrath (till Eclipse proc) > Starfire (till Eclipse fades) > IS > Wrath (till Eclipse CD back up) > Repeat

Wrath spam rotation
MF > Starfire (till Eclipse proc) > IS > Wrath (till Eclipse fades) > MF > Starfire (till Eclipse CD back up) > Repeat

Starfire spam with Wrath idol

1. 3 Eclipse procs 2668 dps OOM in 2:19
2. 3 Eclipse procs 2540 dps OOM in 2:11
3. 3 Eclipse procs 2928 dps OOM in 2:33

Wrath spam with Starfire idol
1. 3 Eclipse procs 2811 dps OOM in 2:20
2. 2 Eclipse procs 2339 dps OOM in 2:11
3. 3 Eclipse procs 2475 dps OOM in 2:02

For some reason i prefer to proc the starfire eclipse, it must be all the big red numbers that start flying across my screen when it happens. With my current gear it kinda is pointing towards the highest dps rotation too, i guess ill have to do more testing on it

 

• Adoriele

Be glad you're not a Mage, where your DPS can swing as much as 1500 without any sort of user control, simply because of Hot Streak. Personally, I like the idea that I can have good and bad days, it keeps it interesting, and it all averages out in the end.