Yeah, I don't know if I'm quite that dedicated until the patch comes out. Maybe tomorrow when I could be working instead of being at home playing WoW

But as to earlier about Crusade vs the Aran Trinket (god I wish I had one now), it'll probably come out to be better than Crusade in any situation where you would have gone OOM. 3000 mana will provide you the chance to do 4 more casts (no JoW no clearcasts) over a 2 minutes period. At 6000 per channel thats 24,000 damage. Crusade at 80 damage

80 * 1.875 * 1.13 * 1.05 = 178, with crits at 34% thats 223.39 damage per channel

2.5s cast speed, 2 minutes is 48 casts or 10,722 damage.

Crusade gets left in the bag , or on the AH to buy more super mana pots instead

I looked back at where the 2900 mana per pendant thing was brought up and I think an assumption was made that after the 20 second buff duration, the mp5 would continue to tick for 20 seconds. As you stack it, the buff duration does not refresh from what I've read (I never got a pendant myself). Was that taken into account? It wasn't obvious to me if it was unless I just missed the post that outlined that.

I'm pretty sure it was assumed to refresh.

From comments on thott you have 20s from the moment of use till the buff is removed. Perfect use would be immediatly after a spirit tick and the instant a non-consumable MSD proc. 12 stacks by tick one (252 mana), 20 stacks by ticks two/three/four (420 mana * 3) => 1500 mana. Half as good as previously thought.

There was one interesting comment posted by a druid. He claimed that an innervate applified the effect 400%. If true, evocate would probably work similarly.

MMmmmmmmm that would be interesting. Someone should test that. If true, it would be a full mana bar plus some. Not sure if it'd be worth a trinket slot. But it might be. It'd fill up the extra 800ish mana you'd get raid buffed from that 40 int, too.

I can likewise confirm that the buff duration does not refresh (at least on Live); some pallies in my guild were quite disappointed with the trinket when they realised this.

However unless the trinket works differently to other mp5 gains you will still get mana each tick i.e. every 2 seconds, so there will be 10 ticks rather than 4.

Assuming you popped the trinket during a 2.5s hasted AM and managed to roll the haste for at least the first 10s (and had Godlike reflexes), you should get, over the course of 10 ticks

Tick 1/2s/4 stacks/84mp5/33.6 mana
Tick 2/4s/8 stacks/168mp5/67.2 mana
Tick 3/6s/12 stacks/252mp5/100.8 mana
Tick 4/8s/16 stacks/336mp5/134.4 mana
Ticks 5 - 10/10s - 20s/20 stacks/420mp5/168 mana

or 1343.6 mana total (equivalent to a passive 56mp5 trinket if you activate every time the cooldown is up).

This is off the current topic, but I thought it interesting to provide this finding.

Until now I was modeling AP as a straight 2.5% damage/mana increase on all cycles. I've finally come around to properly model AP in it's own right and I've found some interesting results. I've been under impression that it would be always wise to use AP in combination with AB spam, even if you otherwise would not have enough mana for AB spam, such as an example where combination of ABx3+AM+Sc and ABx2+AM+Sc would be optimal. However this was not confirmed by the optimization calculations.

First thing I found (which should be obvious) is that you should always use AP whenever it is up (barring special encounter considerations). Even if you don't have mana to spare you should still use AP.

Next is that using AP with AB spam is not always the optimal solution. If you would normally use AB spam, then definitely make sure to coincide it with AP. In most other cases it is better to use AP with whatever cycle you would be doing anyway. In certain situations it's still better to increase AB usage during AP, because of it's effect on AB dpm. For example if combination of ABx2+AM+Sc and AB+AM is optimal, then you should use AP with ABx3+AM+Sc.

Looks like just assuming straight 2.5% damage/mana increase isn't that far from optimal after all.

Wands are most likely a dead end if you're using one that does shadow damage, because you are absolutely going to massacre your raid's ISB uptime.

Even from a personal DPS perspective, that's going to reduce the amount of mana you're getting back from your SP so the mana gains are going to be lower than you're expecting.

Unfortunately, all of the non-arena epic wands do shadow damage.

A Flawless Wand of Arcane Wrath might be your best option to get around that, although you're giving up a reasonable amount of wand DPS compared to the s2 arena wand.

Could invis+drink be a viable option?

Actually the issue may be more of whether it is possible to sustain 1800 or 2000 DPS that you guys are reporting without pulling aggro even with the 40% reduction to threat that arcane has.

60% of 2000 DPS is still 1200 DPS in terms of aggro generation ... Have we finally found a use for invis? *shocked*

To be fair your 1050 dps as fire and proposed 1800 dps as arcane are probably not a very accurate comparison - Manly et al testing on the PTR are in tier 6 gear or equivalent, and from the sounds of it you aren't nearly as progressed (no armory link but I'm guessing tier 4/tier 5).

Also I'm not sure if you accounted for haste in your Arcane scenario, which has a non-trival effect on the timing (faster burn).

I don't think we can draw any meaningful conclusions without using equal ilevel gear and accounting for haste effects.

Maybe Manly should do some testing and tell us what his DPS and DPM in full fire spec for fireball is like. Then compare that with his DPS and DPM for full arcane missile spam.

The way I see it. If AM spam DPM is 30% lower than fireball (example), but AM spam DPS is 60 or 80% higher than fireball. Then it is still better to spm AM until you are oom rather than try to keep from running oom using a lower DPS rotation.

Haste doesn't change DPM. You run oom faster. But you still did the same 200,000 dmg with your mana pool when you are finally oom.

Its a bit like a race. Say I unload mana normally spamming AM (non hasted, non skyfire diamond procced) and I last for 2 minutes. Now, with the meta gem, I unload all that mana in 1 min. I would still have done the same amount of damage, except I did it 1 min faster.

Now I am ahead by 1 min and have that luxury to wand back to full mana before continuing.

The reason why this can't be done with AB is because AB spammed has a much much worse DPM than AM or fireball. Its almost half the DPM that fireball spam can achieve. So, I end up using up my entire mana pool to do only half the damage I would if I used it spamming fireball. So, for AB spam followed by wanding to be better, AB spam would have to achieve at least 100% more DPS than fireball spamming. And quite clearly, it doesn't. AB spam DPS is at most 20 to 30% higher than a deep fire spec spamming fireball DPS.

Is there any WWS from BT / Hyjal raiding? I browsed this thread fast and only found dr.boom testing.

Depends. Manly was saying haste gear didn't have as much impact so much as mystical skyfire diamond. That meta gem and the changes in it and how AM works with it are the key things that make AM spam workable now.

It could be just from normal gear and mystical skyfire diamond procs. In which case it is accessible by just about every arcane mage out there.

Well, back when MSD was lasting 4 seconds, you really really wanted spell haste to keep the proc up. Thing is, at 10 seconds, its just downright absurd how long you can keep it rolling. Up to a point where you're better off just forget about spell haste and try and maximize upon your DPM.

I have to call out a few numbers given.

1- 8000 mana regenerated in 40 seconds (!). 8000 mana is NOT your entire mana pool. If you're arcane specced, and raid buffed, that is not very realistic, even in full glasscannon gear. Maybe 9-10k for low end gear, and 13-14k on high-end gear (and I haven't checked how much gnome racials would affect this).

I understand that using my own gear for example, and using elixir of draenic wisdom, I do have over 350 out of combat MP5 raid buffed. I can assume it only goes higher with better gear. Still, no matter how I try and calculate things, I don't see how you can claim to regen 8000 mana under a minute ! 350 mana * (60s / 5s) = 4200 mana / min.

2- Wanding eating up ISB procs. No doubt, a good point to raise. However, since you're in a full wanding cycle, it is very realistic to assume a mage can swap any fast-attacking non-shadow-damage wand for the duration. It should not be an issue.

3- The 2 cycle theorem does not explicitely state that going OOM and having a 'regen' cycle is out of the question. However, if you think about it for a second, the theorem is all about 'efficient resource management', and since the regen cycle is, by 'common sense standards', very sub-par compared to the resources used in the DPS cycle, I don't think you will be able to convince me ever that it is the proper solution.

The time it takes to generate mana is like far worse efficiency than even AB spam DPM. I know it's not 2 things that are comparable, but I think its pretty obvious here. I don't know about rank 3 DPM, but it should be ridiculously good, with the one downside that it does have terrible DPS. My point remains tho - I just flat out can't see how AM spam and rank 3 AM spam not be better 2 cycles than full rank AM spam/regen cycle.

4- Fire spec DPS? Well I could mention fire spec DPM, but it is pretty absurd to do so for this purpose. You won't and can't go OOM as fire spec. Yes, even on 20 min fights. Yes, probably even if your evocation gets interrupted. Yes, shadow priest are that ridiculous. DPM aside, using similar gear, with no buffs and no raid support, with the proper trinkets, it's more in line with 1200-ish DPS *. I did not take into account raid buffs because 1800 DPS was gleaned using no buffs, so to be fair I did the same for fire.

*. Figurative number for now. I will try and get a proper number tomorrow.

Hmmmm, I might have miscalulated the mana regen part. Sorry. But I am curious to know if AM spam followed by wanding to regen mana might result in an overal cycle that is better than what can be achieved currently.

Comparision to fire is only because its easy and fire is very consistent. Fire can't go oom, but there is a limit to the amount of damage it can do within a certain cycle. (like 2 minute).

Our full mana pool if we constantly cast, plus mana gem, plus pots is finite. It might add up to 20k, or it might add up to 30k. But it cannot be stretched. The amount of damage we can do with that finite amount of mana depends on our DPM.

So, for example. If by doing all we can, eat gem, pot every cooldown, evocation, we have a total of 30k mana. Then if our DPM is say 7, our total possible damage will be 210,000. DPS will determine how fast we can do all this damage. Now, if I usually take 5 min at full burn to do this 210,000, and now, I only take 2.5 min (because the meta gem proc has effectively halved my AM to 2.5 seconds). Then I have a full 2.5 min left to regen mana, wand for more damage. And use up whatever mana regened during wanding to do even more damage. No spell rotation, no matter how efficient will be able to regenerate as much mana as outright jsut wanding I feel.

If you have already done your fair share of damage in half the time you took. Anything extra after that is a bonus. We have to guage ourselves of course. We might not wand up to full mana. Depending on the length of the fight, and our mana regen during the fight, we may restart AM spam before we have wanded back our full mana.

Besides, if fire spec does not go oom because of a shadow priest, then that means that in a group with a shadow priest, even with wanding you should be able to get back your mana pretty fast.

I skimmed the gear discussion part earlier in this thread, but haven't been able to find an answer to my question in it. Apologies if I still happened to miss it somehow.

I'm not entirely certain which gear set would be optimal for my current 48/13/0 arcane build. I've so far been under the impression that with the obvious need of the 2 piece Tirisfal set, it would be more beneficial to also stick with the Tirisfal 4 piece, rather than replacing those items with Tier 6 pieces for a marginal stat and big hit rating upgrade at the loss of the set bonus. I see Manly for instance uses some Tempest pieces, however, does the two piece bonus and the stat upgrade work out as better for damage after all?

To return briefly to the discussion on Pendant of the Violet Eye; someone earlier in the thread mentioned the Darkmoon Card: Blue Dragon and suggested it might be more effective. If we assume

1) Each AM volley has a chance to proc the card effect and
2) The card has no internal cooldown, and further

3) Average AM is a 3.3s cast (to shamelessly steal Manly's guestimate from earlier)
4) You have 396 spirit (112 spirit from gear (full t5), 145 base, 50 from DS, 33 from MotW, 20 from food, +10% from Kings) giving 279/126 mp5 spirit-based regen outside of the 5SR/while casting with Med+Mage Armor

Approximately 90 volleys will be fired per minute, 1.8 of which will trigger the card. Let's say this gives 20s/60s uptime (accounting for overlapping procs - my math skills aren't up to modelling this elegantly). In those 20s you will receive an additional 612 mana or 153mp5. With 1/3 uptime that's equivalent to a ~51mp5 passive trinket.

This is a bit worse than the equivalent mp5 calculated for the pendant, but that assumed ideal conditions where it was clicked every 2 minutes _and_ this click coincided with at least 10 seconds of hasted AM. Additionally the Blue Dragon will of course scale with any extra spirit you get (however minimal the gains may be).

But of course it all depends on the first two assumptions!

Well. wanding means you are regenerating mana at 100% while if you are casting you only regen at 45%. And you do some damage with wanding, though I would admit it is extremely low. So you are gaining a lot of mana while still doing some damage if you wand.

The main point is that before you even started to wand, you would have spent your entire load of mana, and with a decent DPM, that would have translated to a huge amount of total damage already. So, anything after that is a bonus.

Think of it like running. With a certain stamina, maybe a person can run 10 km in 2 hours and then he needs to rest. But ow you have a person who can run twice as fast. His stamina stays the same. So, he runs that 10 km in just 1 hour. Then he needs to rest too. But since he has already run his 10 km, anything after that is a bonus. He could rest half an hour, while walking another 2km, then run another 5km in half an hour.

So, in the end, by the end of 2 hours, this person would have covered 17km.

If our mana regen is the same whether we cast spells or not, I would agree it doesn't make sense to wand if we can switch to a mana efficient rotation. But our mana regen is higher when we wand, substantially higher.

I admit I think I made a mistake in my mana regen calculations. But I think in a raid situation, its quite fast. It certainly won't take a whole 2 minutes to regen to full mana while wanding. Remember how on that rejaxx fight, sometimes we would tank the last add, and the spell castors would wand their mana back to full? That didn't take that long in my recollection.

Just got fixed in tonight's patch push, move along folks. Nothing to see here.

Not surprised in the least. It's still amazing. Just not 'WTF retarded' good.

Hmmm, so with the changes, I wonder how it compares in terms of DPS with other rotations and specs and now that its not up permanently, can we sustain spamming AM the entire time?

Honestly, good.

While the synergies were interesting (to say the least) if they stayed then it would have made us even more dependent on extremely situational or at least extremely supported situations.

As I said to Manly in a PM, even our "insane" level dps was still really not that insane. It was (for us) impressive as all hell but I'd rather look forwards to another solution. Preferably one that doesn't also include off the hook -threat and a range restriction. Those just don't balance at all in my book.

It reminds me of the old Beta situation where AM could be channeled while moving. We all knew it was a bug and yet very few of us bothered to use it. I mean really, 68-70 was far easier just as Frost than Slow+AM kiting crap.

I'll let the number crunchers get to work. But from my personal experience on live, AM 'should' be sustainable with the current am/focus mechanics on PTR with a normal shadowpriest/shaman group, mana gems/pots and evocation. Barring any crazy focus streaks, of course. It's really going to come down to trying it on various encounters with the new mechanics. Theory crafting can get you only so far. It can tell you if it's worth trying. The encounters themselves tell you if it works.