I'm going against all form of convention and actually calling it viable. I've been swapping back and forth between 40/0/21 and 10/48/3 for the past month. We're in Hyal/BT and I put out almost the same numbers. Much of this I'm convinced is due to the EP +3% hit bug that frostbolt is getting, which makes a staggering difference in what I can gear.

At my level of gear (T4-T5 pieces), +hit is a bit of a missing stat, so I've got gems + a pair of "+hit gloves" which I swap out my T5 for to get to the hit cap on fire.

The main reason that I found it viable is the trinketable nature of Arc/Frost, with AB spam and AP frostbolts _at the right time_, I could control when I did damage.

Sure, fire has that sustained love, but my encounters don't have me standing still for 5 minutes nuking some poor static mob to death.

Yeah, I got WWS parses to back all this up.

We typically run with 2 mages. 3 from time to time, but it really isn't the norm. Thing is, if you were to go for the absolute best stacking, ignoring all else, you want ideally 3+ warlocks and 1-2 mages. There isn't much incentive to bring mages because we don't provide group buffs. Again, we're talking pure TC here -- in practice, in EJ, it is mostly a matter of who is available on the night x of raiding and somewhat vaguely a factor of attendance. As a result, we typically run with 2 warlocks and 2 mages. This means COE is very rare. I like to believe my DPS remains competitive despise this obvious lacking buff, but I guess that is mostly a matter of perspective.

What you need to understand is that what you ask cannot really be answered in a generic way. Warlocks do totally nuts DPS given that they don't need to lifetap (and given cod/coa). As such, your RDPS will affect the number of lifetaps they end up doing, as well as fight duration (or if you prefer, the boss HP). Assuming a standard warlock composition of 1 aff lock + x destro locks, the more warlocks you have, the more your ISB uptime should go up. Of course, this means increased rdps, particularly the more warlocks you have. Strictly speaking, crit should really help a lot destro locks, as such e.shaman/moonkin is logically the next step in raid stacking. However, beyond that, you need to think about your raid as a whole. Assuming you run with just 1x e.shaman, and just 1x s.priest, the logical conclusion is grouping them together. The next logical step is to give mages the s.priest because mages simply cannot dps without mana (although, admitedly, at t6 levels it isn't as required as it used to be). Mages are unable to lifetap and do something else than waiting for cooldowns to be ready for more mana. Thing is, in my experience, locks gain just as much dps from a resto shaman than an e.shaman due to decreased lifetap time. However, all of this assumes that the warlock would end up doing a considerable amount of lifetapping. If the warlock doesn't lifetap because rdps is too high, or fight duration is too short, then obviously e.shaman is better.

Here is my crude attempt at theorycrafting mana consumption for warlocks. Be aware that I don't claim to know well the class and its intricaties, but the numbers should give a good guesstimate nevertheless.

assuming 0 haste: 420 mana / 2.5 s
buffed mana pool: 10-11k
2400 mana / 2min (super mana potions)
41 mp5 from BoW (or 45 if improved, but let's ignore that)
--------------------
time to oom: ~80 seconds (32 casts)

24% mana from mana tide / 5min
--------------------
gain: +15.5s (+2625 mana, 6.25 casts)

37 mana / shadowbolt with JoW
--------------------
gain (on 32 casts): +7s (+1184 mana, 2.80 casts)

1000 dps s.priest = 250 mp5
1200 dps s.priest = 300 mp5
--------------------
gain (on 32 casts): +23.75s (+4000 mana, 9.5 casts)
gain (on 32 casts): +28.75s (+4800 mana, 11.5 casts)


So in other words, fight duration will greatly affect the number of lifetaps they might have to do. The number of time spent lifetapping (ie: number of GCD lost equalled in terms of dps loss) needs to be above the gains from the e.shaman.

This is mostly an endless debate and I fear that I cannot really give you a clear cut answer to this. What I am trying to show you is mostly that, given a fight long enough (and keep in mind, those numbers are for fully unhasted warlock, which is probably not the case), your warlocks will probably be lifetapping with no s.priest. Resto shaman makes this less of a pain. To be quite honest, I think it boils down mostly to who you have in your guild. Sure, if you want the perfect raid setup for a boss X, then you should have everyone in the raid be a leatherworker, probably run 5+ shamans, and most likely run with a bunch of rogues. Or 1 mage and 2+ destro locks. The math behind it is very complex, but no matter what TC dictates, TC will not take into account specifics of a given fight which will make certain classes/specs rise or sunk in dps. Case in point: generally our hunters gets really solid dps on teron, but you can't say they maintain their rank on other fights. The same applies to all classes. I like to believe that mages are rather 'versatile' in that matter and tend to do rather decently in most contexts where we are doing a dps tasks, but then again I suppose you could say the same of every class.

And as for what EJ does, we generally have 1-2 s.priest with 1 e.shaman. S.Priest always goes with e.shaman and mages, then destro locks. If not, then destro locks gets resto shaman (+leatherworker) and s.priest if available.

I will take the devil's advocate position. Hyjal favors mostly arcane due to short fights. Although with the recent fire/frost buffs the numbers should be somewhat very close. In other words, arcane in hyjal is in loosely what we could refer to 'a best case scenario' for dealing its DPS. Problem is, it takes the best case scenario, for arcane, to match fire/frost dps. I am not saying your dps will 'stink' as arcane spec, but mostly that the further you move into BT the more fight lenghts will increase somewhat drastically and you will see the very effects of poor DPM. Yes you can still get somewhat 'close' numbers, but you don't really have any advantages as arcane besides:

1- EP being bugged, making it easier to gear up for it
2- having 'burst' dps
3- reduced threat

Now here is the thing. Those are, in my view, the only advantages that arcane has right now. However, I can also prove to you that all 3 of them are not good arguments when you look deeper into it.

1- Sure, EP is bugged. Problem is, regardless of whether or not it is bugged, fire/frost dps should be matching arcane dps for the most part, and outdo it the better your gear gets.
2- Arcane burst dps is the argument I personally love to hate. For all intent and purposes, water elemental 'burst' dps is far far better than arcane power. This is particularly evident to see in Vontre' unreleased simulator that draws a nice graph where you see the evolution of dps in time during a fight. You can really see a huge spike when WE are being activated, and a very narrow spike for AP.
3- Reduced threat. If you have threat problem with fire/frost spec, its probably because you're on gurtogg (ie: threat pushbacks). In actual practice, this is the only place where threat might even come into play, and in that case you can invis and the issue is no more.

And yes, I also have WWS parses to back me up as well. Sadly, I doubt going that route will be productive. I know that no matter what I tell arcane mages they will not change much their opinion on the spec. It can work, for sure, but that doesn't mean I would recommend it generally. Every spec have pros and cons. I personally believe that every 'pros' of arcane specs are really overrated, and that arcane spec number of 'cons' is also a lot higher than the other specs.

edit:
after some thinking, I did find a good reason to spec arcane. AOE. That is pretty much it.

I'm not so much convinced that it's a devils advocate position, it's more about context.

Please don't think I'm advocating arc/frost as "better", it makes little theoretical/mathematical sense that it's competitive. I'm a very reluctant frost mage ;-)

I'm merely pointing out that in real world scenarios, for the encounters that I'm doing, with the gear I have available to me, and the bugs that currently exist with EP makes arc/frost competitive with deep fire.

That's a crapton of caveats to make it competitive. I'm a TC at heart, so I understand the math behind things, but I also understand that a model of events doesn't always match reality (or game reality in this case) and I strive to find out _why_ arc/frost is competitive to fire, when by all indication fire should be ahead.

I've already specced back to deep fire in my bi-weekly quest to try new things out, my "I've got WWS to back it up" wasn't a "you're all wrong, die in a fire", it was a "if you're arc/frost or deep fire where I am, they're both viable, and given normal random factors in encounters pump out very, very similar numbers".

Personally, I believe the 3% EP bug is the leading cause of this discrepancy, and the sooner it's fixed, the better.

Oh, and if you've got access to gear greater than T5, and +hit ceases to be a significant issue, then I can't see how arc/fire or arc/frost would even be close. I think I'm just at the sweet spot which I believe is an important thing to inform people of, given that there are almost certainly more mages at my position, than at full T6.

Well, my take on the matter was that arcane is giving you good numbers mostly due to (hyjal) short fight durations.

In regards to the braid for the +hit on it, couldn't you just use something else like hex and get +15 spell hit rating to gloves? I am sure most people have +20 dmg, but if it is that big of an issue, you can calculate the difference of hex trinket and +15 hit or braid and +20 dmg for the glove enchant. Just putting it out there.

There are a couple of things wrong with that list.

1. Winter's Chill shouldn't belong in a discussion of 40/0/21. You get it from another mage and if that mage specs Fire or 40/0/21 or misses the raid then your dps misses him.

2. Even if Winter's Chill was present it's not that much of a dps increase. Running the numbers on Lhivera's script I'm showing it being equal to 7.5% not 11-12%.

3. Empowered Fireball scales a lot better then Mind Mastery. I have pretty decent gear and with 5/5 Arcane Mind I would have a little over 700int fully raid buffed which would give me 140 additional spell damage. Empowered Fireball would give 202 spell damage from the 1347 damage fire damage I have when raid buffed, add in Wrath of Air and it gives 217. Also, unless something changes, Empowered Fireball will continue to scale better then Mind Mastery all the way till the next expansion.

4. Arcane Instability > Critical Mass (that one being in your favor - gotta be fair) for the same reasons that applied to Winter's Chill versus Improved Scorch.

1. I personally have 2 mages who always raid and who are and will always be deep frost.

2. Pretty sure you're not entering in the ~231% frostbolt co-effiicient, i get 7.5% for 10/48/3 winter's chill

3. it's 25% not 20% so 175 damage, and your int's a bit low i have ~70 or so more than you, still loses out when fully raid buffed though

4. Again, i don't see what i'm missing but they should essentially be the same, 1% crit is an additional spell doubled in damage (assuming 200% crit) every 100 casts, 101 spells/100 casts = 1% dps increase, correct?

Pretend you have a 25% crit rate.

100 casts is equivalent to 125 non-crit hits.

Add 1% crit.

100 casts is now equivalent to 126 non-crit hits.

126/125 < 101/100

I have a question about specs in 2.3.2. Theorycrafting has shown that 2/47/11 is better dps than 10/47/3.

However why not drop the pyroblast, thus you wont have blastwave either. Drop dragon's breath, and the one point put into improved fire blast and throw the points into the arcane tree for 1 point into arcane concentration?

The way I see it those other spells aren't being used in a raid enviroment unless something went wrong with the exception of fire blast (to increase dps at the cost of dpm). So it at least appears to me that at no sacrifice to dps, you can save a tiny bit of mana.

Well personally I would have a lot harder time killing parasites without db / bw

Not to mention the trash clears (esp hyjal) where aoe is required. I really don't think it'd be worth it for the ~ 20 mp5 I would gain.

I've never done Illidan and I never really thought of that, but on hyjal trash flamestrike spam tends to work well for me, though I do use Dragon's breath on occasion should I happen to pull aggro on a mob. So what I guess you're getting at is that a 6/44/11 can be viable for raids, however it isn't as practical as the 2/48/11, correct?

Well there still are a lot of fights that involve some aoe in it. Hyjal and Illidan come to mind.
And while losing full arcance concentration will definately be noticeable, I don't believe dropping some versatility for a very small mana regen improvement is worth it.

It's the opposite - at 0% crit one would see a 1% dps increase from 1% crit increase since 101/100 = 1.01

At 99% crit - 200/199 = 1.005. That's of course using a simple crit modifier (double damage)

A more wordy way to think of it would be this. when you increase spell damage or spellhit or spell haste - ALL of those things can also crit, But you can't Crit a Crit.

Edit - I may have misread what you posted - If you were talking about the crit modifier The % dps increase for 1% crit would go something like.

Your average fireball = FireballHIT + Critmodifier*(Crit%/100)

((FireballHIT + ((Crit%+1)/100)*FireballHit*CritModifier)/(FireballHIT + (Crit%/100)*FireballHit*CritModifier)-1)*100

Where say a fire Crit modifier would be 1.10 in that method of calculation

Obviously as Crit modifier tends to infinity - we get %dpsincrease = (((Crit%+1)/Crit%)-1)*100
Subbing in 25% crit we see that if the crit modifier was tending towards infinity - the %dps increase from 1% crit would be 4%.

You drop 4 points in fire just to get some arcane +hit that you don't need and 1/5 clearcasting that is 15 mp5 at most?
I'd consider it a pretty bad deal.

If you need mana, get CC and drop IV. If you don't, get IV and drop CC. You can't have your cake and eat it there, really.

Crit co-efficient isn't crit %, yes it's also 1% dps at 0% crit, but also if your crit co-efficient is infinite (aka your crits do an infinite amount of damage), or in more general terms, going from 99-100% crit by adding 1% crit is a .99% dps increase if you crit for 100x what you hit, but only a .5% dps increase if you crit for 2x

I'm going to have to interject with my usual notation to try to make this clear.

To find the relative (percentage) increase in damage per cast (and thus, DPS) we have the following...

If we want to see where a 1% crit chance increase is a 1% DPS increase, we have...

With Ignite, that implies that c = 1-1/b = 1-1/1.1 = 9.09% crit. Below that, it's more than 1% DPS increase. Above that, less. As the crit bonus increases, the crossover point rises, trending toward 100% crit (but never reaching it for any finite crit bonus).

I was just trying to explain the effects of the crit bonus tending to infinity :P
Don't need the other coefficients to explain that since they'd be divided out anyway

In other words: it all comes down to how much dmg you can deal with AP, IV, 2x Trinket and Destro Pot running.

With 2.3.2 I will give 40/0/21 (+ Winter's Chill) a try. We cant realy say how much more dmg these "paired trinket combos" are worth but at least it has the chance to be realy good

Taking the points in reverse order

4. Others have already addressed that point.

3. Sorry - I was a bit quick on the calculator but the main point if certainly valid. Current scaling as one's gear gets better keeps a relatively static quantity of Intellect while the amount of Spell Damage increases dramatically. This means that Mind Mastery is an incredibly powerful talent when you are in Tier 4 gear but rapidly loses strength compared to the Empowered Talents as you gear past that point.

2. Not sure exactly what you are saying with the "7.5% for for 10/48/3 winter's chill" but if you are saying that Winter's Chill is giving a 7.5% dps increase then that's exactly what I said in my post. 7.5% does not equal 10-11% and so winter's chill does not equal improved scorch.

1. Final point - YOU may have 2 dedicated Frost mages who are willing to boost YOUR dps above their own by playing Deep Frost and giving YOU winter's chill but that is not a common set of circumstances. Back before TBC there were WC bitches who would suffer through lower dps in order to boost the collective dps of the other frost mages. If you could show that having one Deep Frost mage and one 40/0/21 mage would be collectively better then 2 Deep Fire mages then you would be showing us something relevant to everyone. Even if it took 1 Deep Frost and 2 Arcane/Frost mages to exceed the dps of 3 Deep Fire Mages that would be worth theorycrafting over since it would be a new way to look at raid structuring.

Saying that 40/0/21 is superior to Deep Fire when supplied with Winter's Chill from another Mage is the equivalent of saying 40/0/21 is superior to Deep Fire while in a group with a shaman while the Deep Fire mage is not. You are making an assumption of a buff from another mage in your comparison without taking into account any decrease in comparative dps from that other mage.

If you want to have a real discussion of the 40/0/21 build and bring Winter's Chill into the picture then you have to average the dps of the Arcane/Frost mage AND the mage that is giving him the Winter's Chill buff and then compare that average to the other builds.


I really have nothing against either Deep Frost or Arcane/Frost. I think they are valid specs and they are worth discussing in this thread on theorycrafting, but if the mage is incapable of generating the buff without another mage then you have to involve that other mage in the discussion. A Deep Fire mage who doesn't have to refresh Imp Scorch will certainly do more dps then one who has to refresh his own buff. The real question should be whether that is relevant or not? Would having one mage keep up the debuff increase the collective dps of all the mages. That's the question that is worth theorycrafting, not the simple fact that if he didn't have to keep up the debuff he would do more dps.

So - I was thinking about hyjal trash today - and aoe caps etc. Of course crit is the ideal aoe stat when damage capped.
I was wondering if anyone has a darkmoon card: Wrath laying around to test this - Or if somebody knew how exactly the mechanic works with aoe'ing.

For instance - Say I was aoe'ing a pack of 10 mobs - would it be giving me an effective 170 crit rating on each aoe given that they all strike simultaneously? Or would it just be completely useless.

If you're damage capped the only way to increase your damage is to go for spell haste