I'm sure there's bad Mages somewhere being reduced to wanding, but the actual situation is that we go to a less mana intensive, lower DPS rotation (skip Living Bomb, maybe skip Pyro if desperate) if we sense trouble on the horizon. And use Evocation.

The issue is that Evocation is only an "Oh shit" button basically, and Mana Gems are a no brainer. There's no strategy involved really, and the new Pyromaniac does nothing to change that situation (or to make Spirit anything less than a waste of item budget.)

Please remember that Naxx is a reused instance and based on old encounter design. During those times, Blizzard did punish melee a lot but also gave melee a huge dps advantage over ranged. It's not the case anymore with new encounter designs since the middle of TBC.

Keristrasza in the Nexus is closer to the new raid design than Naxxramas. In that fight, melee can move without impacting their dps, while ranged loses some cast time, even if they only move the tiniest of inches. In the second half of TBC, the number of silence and pushback mechanics balanced out the PBAoE mechanics.

I am usually of the mind that player suggestions are, well, terrible, at all times, even when they're great. I seem to believe that most ideas fall into various categories, such as either being overpowered, situational, short-thought, biased, or a rehash of other ideas. However, this thread is doing a fine job of addressing issues without delving into the problems I've listed.

In fact, it seems like this thread is clad with ideas that Blizzard could at least make a stepping stone from. The problems with Evocation, as outlined here, include that it is 8 seconds of non-DPS. Eight seconds is a lot of time to not be DPSing, especially when 'burst' phases cover 10-20 seconds at a time when trinkets proc or are used.

Evocation then would benefit from any change that relieves the channeling function of it. While 60% mana back is a lot, it goes without saying that the mage class is built around have mana to deal damage. You cannot be a mage without a mana pool. I see a few ways of changing Evocation's mechanics to address issues of both mobility (having to move out of the fire and ending evocation early), interruption (self-explanatory), and DPS stopping.

- as an instant mana refill, with a Mana-Over-Time component, Evocation would function as a mana gem every 4 minutes, with 30% of your mana instantly being returned to you, and the other 30% being returned to you over X number of seconds.

- functioning under the same channeled mechanics, evocation would work like a reverse Curse of Agony. It would return a higher percentage of mana back in the first two ticks, and a lower percentage of mana back in the last two ticks. By example, the first tick would restore 45%, the next tick would restore 20%, the next tick 20%, and the last tick 15% (4 ticks would return 100% of 60% of your mana then, but the first tick would restore 45% of 60%). These numbers are just for demonstration, keep in mind -- the mechanic is what I'm addressing.

- a cooldown based on the duration of your last evocation. If your last evocation lasted the full 4 ticks, you would have a 4 minute cooldown. If you received only 3 ticks, your cooldown would be reduced by X number of minutes. The idea is that if you were only able to receive one tick of evocation, the cooldown would be short enough that you could re-evocate some time later. This would have to differentiate between being interrupted by damage (spell pushback) and being interrupted by other mechanics such as stuns, school interruptions, silences, fears, and so forth.

The first addresses the mobility and non-DPS-time downsides of evocation. The second addresses things such as evocating for only as much as you need it (the goal being to receive enough mana to continue DPSing for the remainder of the fight, but not so much that you have excess mana at the end, which would imply that you lost DPS time). The third addresses the frailty of evocation.

I don't get what your argument is. I'm not comparing melee to ranged dps, I'm comparing mana to non-mana using dps. Ranged have to move or adjust and deal with things just as much as melee do, in addition we often have to dispel debuffs or what have you, regardless that isn't relevant or what I'm arguing.

I'm saying the way mana is setup now, it's either a non-factor or crippling depending on the duration of the fight. If the fight is short enough that mana users don't have to worry about running out of mana, we use our return abilities like gems for mages or innervates for druids, or fiend for priests, and keep on dpsing at 100% ignoring mana altogether other then to pop the cooldowns when they are up. Theres no real thought to it other then we don't want to run out. On fights that last long enough to exhaust all these options or otherwise drain us of mana faster, once we run out there is nothing we can do. So while a rogue or warrior can be balance around the controlled rage or energy gain factor, there is a corner that they get backed into when trying to balance caster dps in my opinion.

We can be balanced assuming that there are times when we will have to use a lower dps cycle during mana regen but that is broken if we don't end up needing to use any of that regen as the raid gears up and the fights get shorter. Since we are balanced around having a period of regen and lower dps, if we are suddenly on 100% full burn output in every fight because they are shorter, our dps is going to be inflated far above classes like a warrior or rogue because we aren't limited by mana like we are supposed to be. If we are balanced around the 100% full burn scenario, during new progression or fights that last a long time where we exhaust all our mana return options, our dps suddenly has a huge penalty that a warrior/rogue/feral does not.

I'm not saying that a ranged should have more dps then a melee or vice versa, I'm saying that when considering how to min/max dps, fight duration really has minimal effect on non mana using classes compared to mana using classes and that affects how hard it becomes to balance dps and all these different mana regen mechanics that exist.

I think a better mechanic than this is to change Evocation into a 2 second casting time spell that restores 15% of your mana (with a low or no cooldown). This lets you evocate for as much as you need to without wasting the rest and risking a mana drain (or something) leaving you with options for the rest of the fight. With no cooldown it basically turns Evocation into Lifetap with the risk being standing in place instead of losing HP.

1) From an immediate perspective, the argument is entirely academic. Mages do have useless stats covering most of their gear, but they are *balanced* around this fact, and don't run out of mana. If Blizzard suddenly gave us 30% of our spirit as spell damage, ala locks, they would probably have to nerf us in some other way. So from that perspective the problems with spirit and regen are completely academic.

2) But from a playstyle perspective, this is a minor problem. It makes mages a pretty boring class to play on raids, as they are entirely focused on rotations, ignoring things like whether I have enough combo points, whether I need to burn some rage, etc. I think most classes require some on-the-fly decision making that allow the best players to really shine. Mages don't have that as much, and it is a minor gripe with the class. I know I enjoyed my mage in pve the most when I was arcane/frost and was playing in exactly that fashion.

3) But to my mind the biggest problem with regen is that Blizzard seems to feel compelled to shove it down our throats. We jump back and forth between Blizzard not giving us enough mana to do dps, and having so much mana that we can ignore it. And a perfect point doesn't exist due to the way mage's are structured. There *is* no perfect balance Blizzard can strike. We don't have viable mana dumps, so either I have enough mana to execute the rotation around which I am balanced, or I don't have the mana and I am a gimped class. This is *not* academic. If Blizzard's love/hate relationship with mana forces us to come up with solutions that will stop them from screwing around with us, so be it.

I, for one, do not think that the finite vs. infinite resource problem is actually a fundamental flaw.

Fights do not have an infinite variety of lengths. They're not even ranging from 30s to 30min. They're ranging from like 2 minutes to like 6 mintues. And game balance is not inseparable from encounter design. It's obviously true that a class with a finite resource* is going to be more powerful on a short fight and less on a long fight. This does not make them under/overpowered, or broken. Since blizzard's explicit goal is that DPS classes are within a few percent of each other, and their implicit goal is that the rankings shift from fight to fight, it is reasonable that having finite resources can shift your DPS by 10% as the fight length varies from 2 to 6 minutes. In fact, it might even be a good thing: it's an easy knob for the devs to turn when tuning fights. Mages are underperforming? Cut the HP and relevant timers in half. And, I think that given the tools currently available, mages probably are within that range. When looked at from this angle, 8 seconds of lost DPS to extend DPS time by two minutes is well within the range of acceptable variation.
[e] This does assume that the class has tools that make its DPS on a six minute fight considering more than half that of a three minute fight. The availability of mage armor dampens the effect of finite mana by giving much longer OOM-time for comparitively little DPS loss. Having a higher-DPM slightly-lower-DPS cycle available would also solve this.

Now, that said, blizzard has pretty much confessed that mage regen is currently under a band-aid fix. They wanted to give you guys a good system, but they couldn't in time, so they cut your costs to give you a functional system (ie ignoring it completely) that does balanced DPS while they shelve the issue. Please realize, having enough mana is a bandaid fix.

As has been pointed out, evocation is not meant to be your baseline sustainability. It's meant to be your emergency sustainability. Its cost, whatever that may be, should be balanced to discourage you from using it whenever its up. Given that, it can probably remain unchanged. If any redesign occurs, it should probably be just to make it more reliable so that you can count on it in an emergency. For example, a self-silence buff that persists through damage and movement.

This also means you're missing a baseline sustainability option. And by option, I mean something that is actually a choice, not like mana gems in which your options are use them or get /gbooted for being an idiot. The question at hand is, do you need one? Many other classes have one, but that in itself is insufficient. Arguably, mages are unique in that they rely on passive measures and decisions made at the talent and gear level (and mage armor) plus an emergency ability rather than having them be an active part of combat. If you want an active regen model, you have to show this is bad. Balance arguements will likely not work, gameplay arguements probably will. The arguement overlapping the two, about how you need metagame knowledge to plan the amount of passive regen you need, is probably effective as well.

Spirit is an endmic problem. It seems like the current philosophy is, you should need spirit for the mana, with SPT->DMG conversions being more about making the bitter pill easier to swallow than about being the premier reason to have spirit. That's how it is with warlocks. Say all you want about the 30% on fel armor, analysis of the 4T7 bonus has shown the spirit bonus is better spent on more from the mana regain on lifetap than on the spellpower for offensive spells. Damage conversions come second, after rehauling the mana system. They are meant to address the problems of people not getting the spirit they need, or being upset about doing so, not the problem of wanting spirit in the first place.

At some point I figured out that ice lance should return a multiple of your spirit in mana when it crits (or proc spirit tap). This amazingly unintuitive result solves the PvE shatter-combo problem without disrupting PvP, as well as giving frost mages a use for spirit (assuming costs raise to otherwise-unsustainable levels). I was floored and do not expect this to actually occur.


*This does not just mean academically finite, this means realistically finite. Either a mana dump or a realistic ability to rum OOM in 3-4 minutes by not using class tools satisfies this definition.

I think Evocation would be a lot more enjoyable to use if it was affected by Arcane Stability (ie, immune to pushback).

My 10-man raids generally run without Replenishment (our resident shadow priest is only available 2 nights a week), so on long fights like Kel'Thuzad my mana situation can sometimes get very dire. It's pretty frustrating to specifically wait until after a Shadow Fissure to use Evocation, only to lose a good chunk of mana return to a Frostbolt volley (or worse, lose most of it to a frost tomb). There are other examples but this is probably the most annoying I've found.

Well, I think the problem is that we can't have it all. Ultimately, from a player's perspective, the biggest problem mana issues have is that the difference between 'OOM' dps and 'full' dps is too big. If we had a real 'OOM' DPS, or at least some way to dps during evocation, then it wouldn't be so bad. But it could also mean implicitely that to do that would require lowering slightly our base damage to even it out. Do I really want a lower base damage under most circumstances ? Well, its somewhat debatable.

Ultimately they'll never be able to change the fact that eventually we will go OOM, and that timers or not at that point there is nothing I can do about it. While the overarching goal might be to have mages manage mana, the end result is that 'managing' mana, as it is now, means mostly sacrificing dps to allow for longer sustained damage. And the very important thing to point out here, is that the 'longer' part of this statement isn't a whole lot. How much longer does 1 mana pot allows me to hold ? Not much longer. How much for mage armor using non-spirit gear ? Well, I doubt its much. How about mage armor and changing all my gear to use spirit gear ? How much longer can I damage, and most importantly, how much dps do I lose ? And non-arcane specs hardly have any options spell-selection-wise to manage mana (/increase DPM).

I think if we had a common scenario out of which we could give concrete results as to what the net benefits (costs/gains) are out of 'managing' mana, we might get something interesting. I have a rough idea of what it looks like, and its not too great.

But even if all of the above shows with numbers that we can't really extend the damage by much longer even if managing mana, the fact remains that eventually we will end up fully OOM, and it has got nothing to do with managing mana. And once we get to that point, the dps difference between OOM dps and 'dps' is simply too staggering.

What if, in the world of temporary band-aids, blizzard would be to give us a spell that we could cast that costs no mana and would allow us to do some baseline damage ? Or retake the old idea from Vontre; to have arcane missile generate mana (!) (of course, that was proposed before missile barrage) ?

The main problem with our mana is that we can go OOM with all the CDs expended and nothing to do. I was in a 10 man KT we had no replenishment or JoW and he (KT) liked me a lot for mana bomb, I was OOM started to Evo. and had to run out of the void zone so I ended up with evo. on CD mana gems on CD and 0 mana. I was helpless, the only thing I could do was wand or sit down and look at the flashy lights.

There are many suggestion on how to solve the mana problem not sure how blizzard looks at them, and their lowering the CD on evocation to 4 minutes is just ridiculous as it doesn't solve the problem when you're oom, it just helps with chain pulling as manly said.

My proposal is to make the CD variable dependant on how many ticks you were able to get, that is if you took all 4 then the CD is 4 minutes, 3 -> 3 minutes and so forth. That will make evocation a lot more usefull and will not give us more mana then blizzard intended.

EDIT: now that I look at my proposal I like it even more, would be great if anyone with an account on wow forums could drop a line for the blues to read, or Manly to place in his signature as it seems to have an effect :)

1) On OOM DPS Design

Mages having about zero OOM DPS is the fundamental design flaw, from an academic point of view.
Priests, druids, shaman and perhaps paladins are in the same boat.

The idea is that you as player ensure that you do not run OOM.
This requires proper raid encounter design, having most raid buffs and being smart about using your mana.

Right now, the mage class works pretty well in raids most of the time.
Mana is usually enough, but it can get very rough on lengthy fights without JoW/VT.

If everything aligns, there are situations where you inevitably run out of mana, regardless of what you do.
Those situations however are extremely rare and only when you lack Replenishment and Evocation gets interrupted.

As far as I know, the other classes mentioned above consume less mana than we do and simply do not run out of mana in normal raid encounters.

So, while having low OOM DPS may be a problem by default, it is solved by designing stats and encounters that make you not reach your OOM timer most of the time with proper play.
It is something that I can more or less accept as design.



2) From a practical point of view, mages usually do not run completely out of mana.

Load up Rawr with a decent gear set, have Replenishment and Judgement of Wisdom as buffs, and use 3.0.8 mode.
(Rawr actually underestimates your longvity, since you can and should recast Mana Gems for fights longer than ~9 minutes. That's irrelevant for current content though.)

Frostfire Bolt is infinitely sustainable with Molten Armour, Potion of Speed, full Living Bomb uptime.
It just requires using Evocation, and you can actually cancel a few Evocations for running out of fire.
You can stand and spam until your fingers bleed or your gear breaks, but you won't run out of mana.

Fireball with Molten/LBomb, Speed Pot and Evocation is sustainable for 14 minutes.
17 minutes with 2/3 Meditation, 24 minutes if you use a Mana Potion.
That's still with Molten Armour and full Living Bomb uptime.

Frostbolt spam is even easier sustainable than Frostfire. 20 minutes spam before needing to Evocate.

Arcane with a 1-Blast-1-Barrage cycle as base (2-Blast during haste cooldowns) lasts 4 minutes without Evocation.
That's with Molten Armour, and using AP and MBAM and any procs I might miss. No Frost Channeling/Master of Elements.
Arcane is infinitely sustainable with Molten Armour due to the 2 minute cooldown on Evocation.


It gets rough when you have neither JoW nor Replenish in a small raid. Or if your ret paladin excels at suiciding on the pull.


Frostfire drops to 7 minutes sustainability with only Evocation.
It's infinitely sustainable when you drop down to Mage Armour, Mana Potion and no Living Bombs.

Frostbolt is nearly the same. 7 minutes with just Evocation, infinitely with Mage Armour and Mana Potion.

Fireball needs double Evocation and a Mana Potion for 6 minutes.
It is infinitely sustainable with Mage Armour, 2/3 Meditation, Mana Potion and no Living Bombs.
But it does less DPS than Frost specs during those long fights because it sacrifices so many damage options.

Arcane also needs double Evocation and a Mana Potion for 6 minutes.
With Mage Armour it goes up to 11 minutes.
After that, it has to use strict AB-ABar which involve pausing to wait for the ABar cooldown (if I read Rawr right).


3) What does that mean?

For all practical purposes, sustaining mana is not a problem in a balanced raid in encounters not longer than half an hour.

It does get rough when you're in a raid without mana regeneration.
It can be done with good class knowledge and a spec that cost uses little mana. I.e. you're defaulted to Frostfire.

It heavily relies on Evocation though. If that gets interrupted, you're out.


4) On Evocation.

So, for a mage on a chaotic progression encounter, that means either:
a) always have Replenishment and Judgement of Wisdom, or
b) make Evocation more reliable, some "try again" mechanic if it gets interrupted/you have to run out of fire.

A shorter cooldown might work for b), but you risk making Evocation too strong.
With 2 minutes, you'd always use Molten/SpeedPot/100%LBomb, because Evocation is the best Mana Regen option of those three and enough to cover the fight.


Evocation with it's 8 seconds channeling time is fine. It's powerful enough.
Allowing DPSing during it would make it even more of a no-brainer as your first option when you need extra mana. There's also PvP to consider, whether we want to or not.


What could work is removing the cooldown and changing it to a debuff that reduces the mana gained from Evocation by half or doubles the channeling time or something.
Penalty might perhaps have to be even harsher (1/3rd or 3 times), would need some number tuning.


That way, nothing changes most of the time. You want to use Evo every 4 minutes (2 as Arcane) when the debuff drops because it's good mana-per-damage.
You don't want to chain Evocation because it's bad mana-per-damage during the debuff.
If a critical Evocation gets canceled and you really run completely out of mana and cooldowns, you could still use a penalised Evocation since anything is better than OOM wanding.


The mana system more or less works otherwise. If you have to use Mage Armour then both, spirit and intellect are better than crit and nearly as good as spell power.

The crucial part is Evocation, that will always get interrupted by Murphy's Law at the moment you need it most.
If it could be made less fragile or less punishing, then I'd be more or less happy with the system.

They did. It's called "Shoot."

I'm completely serious about that. I believe pretty strongly that the original design intent of a class can be deduced by looking at the talents and spells the class was given. For example, there's a big thread on the DPS forum now about whether or not it's a good idea, as GC has suggested, that Hunters have better melee abilities. People have argued that, to paraphrase, "that's not what hunters are supposed to be!" Well, respectfully, that seems to be pretty obviously wrong. They have mail, they have melee specials, they have talents that affect their melee capabilities. It seems clear that they were in fact originally intended to be a ranged/melee hybrid, not as good at range as a Mage or Warlock, and not as good in melee as a Rogue or Warrior, but reasonably competent at both, and thus more versatile. Obviously, it didn't actually work out that way in the course of development, but it seems to me that GC wants to redirect Hunter development toward a point closer to the original design concept -- not away from it, as some people seem to believe.

Similarly, take a look at how Mages started out. They had no real way to increase DPS. They did, however, have a way to increase duration of casting, and to reduce regeneration time. And they had a wand, and a talent to improve the use of that wand.

Based on these facts, it appears to me that Mages were actually not intended to be casting full-time. They were intended cast part-time for high DPS, and wand part-time for low DPS and regeneration.

The problems with this plan, of course, were: first, the classes that could DPS full-time performed at a higher average level, and second, encounters were not designed such that having sustained extra-high DPS on-tap for only part of the encounter was a useful thing. Plus, even if they had been designed that way, Mage DPS while casting wasn't really any higher than what rogues produced all the time.

So the route they took was to change Mage design to provide sustained casting, and the mana mechanics have never been revamped with that fact in mind. But perhaps they should instead have tried to fix the original concept. To borrow again from the hunter melee discussion: What if, while casting, Mages did 130% the damage of Rogues, and while wanding, they did 70% the damage of Rogues, and in most encounters, they had to spend 50% of their time wanding -- and lack of spirit would increase wanding time beyond the average? (I'm sure there are many things wrong with that specific idea; I'm just trying to illustrate that they could have taken a development path that didn't involve full-time casting.)

In short, I'm reasonably certain that Mages were in fact designed to work with an ability such as Manly describes, and in fact were given that ability in the form of wands, but the developers moved away from that design rather than trying to perfect it.

Tying mana gems to spirit really is a solid idea and makes a lot of sense. It's been brought up a few times, but it's worth really exposing as a potentially great mechanic.

Possibly making the actual mana restored scale with spirit, or gems passively granting a temporary +dam buff (like the 2pce t7, built in) whose power scales with spirit. Or spirit could reduce the cooldown of your mana gems. Theres all sorts of ideas to tie them in, considered sweeteners. You're not going to stack spirit because of them, but it's just a small extra reason to appreciate the stat.

It's an elegant way to give spirit a more 'tangible, fun use' for mages (mentioned in first post) and further consolidate mana-gems as a strong class feature. Rather than just handing out a generic Warlock type buff that converts Y spirit into Z damage - that feels uninspired to apply to multiple classes and a little cheap.

We could even get to the point where mana-gems would restore such a respectable chunk of mana, that you would want to use them much more strategically than the 'click every 2 minutes' scenario. More choices, more fun and logically tying stats into class-themed abilities - all from one fairly simple change.

I wouldn't approach this scenario as if it were the end-all solution to spirit issues, but something thats merely intended to help. A stat that is deemed acceptable, but only the mathematically gifted or atypically astute can understand why, still has room for improvement (However, I admit this is subjective between players). It's back to the TBC Ret Pally vs TBC Shadow Priest in a group scenario. Having fun, tangible and transparent mechanics is important. Even if spirit is in the background performing its primary role, there still is a compelling reason to give it the aforementioned attributes and tie it to something that fits our class theme.

I think looking at the original skillset to determine intent of the class is about as useful as reading the original manual. MMOs change and evolve, it's completely irrelevant what skills and abilities classes had at release.

Randomly tying things to spirit would be annoying and not particularly beneficial by itself. It's only worthwhile if it bridges the gap here choosing the equivalent item with spirit is sometimes the right choice.

Otherwise, we're left with a situation where mages are completely ignoring or actively avoiding spirit on our gear (which we are) and lamenting the fact that one of our mechanics is tied to a stat we dont have/want.

No, the biggest difference is that energy and rage function closer to a casttime/cooldown replacement than mana does. This is the same error Blizzard continues to make: "Giving casters more mana is just like giving melee more energy..... right?"

Wrong, giving melee more energy/rage results in something like a haste equivalent for specials, versus more mana for a caster which, on everyone other than an arcane mage, results in no change at all in their normal DPS.

They want melee to be mobile and able to continue DPS'ing on the run, so rather than giving them formal cast times or putting everything on cooldowns, they balance the 'cast time' around energy/rage cost and expected rate of energy/rage gain on a raid. So give a melee more resources and he does more damage because his cast time shortens (aka haste.)



Our first Sarth kill with three drakes was more like 10 minutes -- which isn't an unheard of length for a fight in WoW, even if the bulk settle around 6 minutes. As a mana-constrained class we need to care even about those outliers that are an extra 60% longer.

And since it includes significant AE, it very definitely was a strain on my mana.

All the calculations in this thread assumed that there was no AE in the encounter -- an assumption which is probably flawed, given that they gave almost every class in the game, even rogues, incredibly strong AE.

Great post Lhivera. Sorry if my following reply isn't up to par with your own comments, but you said it all quite well, there's no need to rehash it. But your comments on wanding got me thinking... Why isn't any emphasis placed on wands? Well, they have low DPS output -- that is a given. Their global cooldown issue was addressed. And now we're looking at ways to play around with mana regeneration.

It seems to me that the two could be blended. By this I mean that, when out of mana, your options would be: gain 60% of your mana over 8 seconds through Evocation, under the same mechanics it is currently under. Or, resort to wanding to generate mana. How? Convert damage to mana at a 1:1, or some other, ratio. Take the DPS of the two most common wands used in sets: [Rod of the Fallen Monarch] and [Gemmed Wand of the Nerubians]. This would directly translate into mana returned to you. I'm not talking about DPS being converted into Mp5, though in a way, it would be.

I'm saying that if your wand shoot does 617 shadow damage, you gain 617 mana. If it crits for 925, you gain 925 mana. You are sacrificing thousands of DPS for mana returns. This addresses several issues:

1) Wanding is off the global cooldown. You may return to casting at any point in time. This circumvents the numerous problems that can result from evocation's channeling (such as taking damage, or having to move out of the fire).
2) The caster is in complete control of how much mana is being returned, relatively. With a mana pool of 20,000, evocation is returning 3000 mana every 2 seconds (7500 Mp5). With wanding, you could have mobility as well as being able to return upwards of 1375 Mp5, or less if it wasn't a 1:1 ratio (which does seem incredibly high).
3) Zero interruptions. This is probably the most alluring factor. How often I've used evocation and had it inadvertently shortened due to AoE damage. And this is after all my intentions to use evocation at the most opportune moment.

When attempting to determine how they should continue to change and evolve, I think it's worthwhile to look back at the history of the classes, and the original design intentions, and ask if there's anything there worth using as the game evolves further. Were ideas and plans abandoned for good reason, or just for the sake of expediency? Sometimes, yes, the old ideas were just plain unworkable. Sometimes -- and this is clearly what's being considered for Hunters at the moment -- perhaps deviating from the original plan as much as they did was not such a good thing. Obviously, if Hunters did 20% of Mage DPS from range and 20% of Rogue DPS in melee, they wouldn't be viable...but perhaps abandoning their melee capabilities so completely was not the right move.

The developers can't be slaves to the past of the game, but they can't assume there were no good, unfulfilled ideas there, either. And certainly, when Manly specifically asks, "why don't we have a zero-mana attack," it makes sense to point out that we do, that we've had one from the beginning, and perhaps if the class had been designed around actually using it effectively, it would be a more interesting class.

I've got pretty colourful memories of doing 10-minute Morogrim and Hydross fights without a shadow priest!

Sustaining AE totally wrecks your mana pool, true. I had something written which was eaten by a client crash.
I decided to leave it out because not every class can AoE well, and all AoE is different and differently sustainable.
Same for trash, where a mage who spends 1/3 of the combat time drinking is still better than a rogue right now.

There are rogue and hunter changes in 3.0.8, and further contest is announced to be not just "AoE all!" style trash.
And mages are usually the only ones who spam drink after every or chain pots on trash when clearing contest fast.


AoE and farming mana are a really huge can of worms which I didn't want to open yet.
I think the mana management at hand offers enough controversy for now


AoE and trash zerging are another dimension of headache indeed.

The gear is what I stitched together more or less as best-in-slot for FFB a while ago with Rawr 2.1.2 or so.
I was using Rawr 2.1.4 in 3.0.8 mode for the test.

4T7, Gothik's Cowl, 535 haste rating, only item with spirit is T7 Robe. I actually need to reshuffle for some more hit.
I picked all the common raid buffs, i.e. 13%CoE/ToW/WoA/FI/iMKA/Misery and Flask/Food.
Judgement and (normal) Blessing of Wisdom, normal Mana Spring Totem, 10% Blessing of Kings, Replenishment. AI, GotW, Spirit from a Felhunter (that one is debatable).

For the numbers without JoW/Replenishment, I removed just those 2 debuffs.


As for your second question - Mage Armour doesn't have any use in a stand-and-nuke fight where you have JoW/Rep and Evocation never breaks.

Mage Armour is only useful if Evocation and Mana Potion do not provide enough mana for you.
You need a combination of several of "lots of AoE", "long fight", "no mana battery", "no mana buffs", "mana burns", "unreliable Evocation", "no running time" (you could regen a lot during Felmyst for example).

Which means that if you buff Evocation or reduce/remove its cooldown, Mage Armour (and mana potions) will never be used for progression fights.
I could perhaps see a use for trash if you always sit behind and drink though.


"Mana vs. Damage 3.0" - What to do when you run OOM?
Currently for most specs, you should work down this list in the given order to determine what to do when you run OOM.
It's valid for Frost, Fire, Frostfire. Partially for Arcane too, but you have to juggle your cycles as well.

1) Use Mana Gems. No-brainer with 2T7 even. Reconjure them mid-fight for extremely long fights.
2) Use Evocation. Plan ahead for multiple Evocations on long fights and watch boss timers to not get interrupted mid cast.
3) Use a Mana Potion. Speed Potions are really good now, try to avoid having to use mana potions.
4) Use Mage Armour. You need to know before the fight that the first three options are not enough.
5) Skip Living Bomb. Right now in Naxx gear, you'll want to cast it as long as you can. Efficiency may drop in better gear.

Evocation has gone from "terrible" to "alright", and our mana has become easier to sustain.

I'm not sure how you squeeze 2. I mean theres one you can use pre-pull, one in combat. With this said, theres never any disadvantage to not use one pre-pull, unless you somehow absolutely needed that mana potion in a sub-2min fight (read: pretty much impossible).

That's what happens when I try to squeeze to much info into one sentence.
The on-pull speed pot is always an one option of course.
But in combat, you should try to plan around using a speed potion instead of a mana potion, even if you have to Evocate for mana.

Another, much more practical problem about Evocation is that you can't determine how much mana you actually need when you channel it.
Master of Elements and/or Clearcasting is a lottery on top of everything else.
When you OOM when the boss is on 5% because you skipped one tick of Evocation, it feels pretty bad.

What has always baffled me is the fact that melee classes have always benefited tremendously from base stats (STR and/or AGI for AP, and AGI for crit), while DPS casters have not. Melee gets much more crit out of Agi than we do out of Int. Sure, Int also gives us mana, but Agi also gives some classes AP in addition to crit, as well as dodge and armor. Strength converts to AP at a 2:1 advantage for the classes that don't get AP from Agi, and gives a good amount of block value to paladins and warriors. Really, if the design had been that spirit converted to spell power all along, as str and/or agi do for attack power, we wouldn't be having this particular spirit discussion. Spirit would have been as good for cloth and leather casters as agility is for rogues. Instead, base stats have always been less desirable for us. Spellfire had nearly nil for base stats, and was for a long time better than two tiers of gear, and the market for Deathchill cloaks is a current example of how bad base stats are for us.

Of course, it's far too late to make this kind of grand change; it would require an overhaul of the entire item database. Instead of doing this, they've very recently (relative to the game) added spirit to spell power talents/abilities for warlocks, spriests, and moonkin. This reduced conversion must be extended to mages and the game balanced around that so that spirit is never as worthless as it stands. Making us want to use spirit gear has to go beyond the threat of running out of mana. If it doesn't increase our DPS on the majority of fights, we won't use it.