If you're having mana problems, maybe gearing towards crit rather than haste is a better idea. Haste and crit are usually pretty equal in terms of dps, but in terms of mana regen crit is the clear winner. I know I'm definitely trying to swich out some haste for additional crit. Not only for mana purposes but also to combat those unlucky times I get so frequently where I end a fight with 10% less crit than I should have.

Oh, I agree it's not the best idea, nor does it really give us 'management' per se. It just seems like the simplest/fastest way to actually _do_ something about mana management. I would rather the old spirit bolt idea that was tossed around during beta.

And I'll admit I do not really pvp (a bg once in a while) so I will leave that side of it to y'all. I just never really understood the need for a cooldown on evocate. Warlock don't have a cooldown on life tap or any of their regen abilities. As far as I know, actually, no other class has anything similar to our problem. As said above, the only reason it hasn't been a major issue is because it's constantly being masked by huge mana pools.

What was this idea? Don't think i've heard of this. Was it a kind of cooldown-based spell that scales solely from spirit, and gives a mana restore equal to the amount of damage done?

Well dispersion and innvervate are on cooldowns as well. Where according to the spriests I talk to they have more mana problems than we do (at least as frostfire -- as arcane or arcane/fire it will/may be different or similar).

epoh:
First, warlock lifetap doesn't have a cooldown because the nominal cost to lifetap is actually the health, not the GCD. People seem to treat this as not the case. (Personally, I wish lifetap would be redesigned so that the life cost is the most important part again. But, that has nothing to do with mages.) Lifetap has no cooldown because it's self-limiting because you can kill yourself. Evocation has a cooldown because it has no literal 'cost' other than the opportunity cost of using it.

As an aside, warlocks seem to have been intended to have the most organic resource system (of the mana and/or ranged classes. Warriors and DKs seem to have a comparably organic system). I doubt any other mana class will ever get a resource system as on-the-fly as ours. The side-effect of this observation, is that mage mana pools will probably require more planning ahead than warlock regen does. Ours is entirely reactionary; one could argue this is class-defining. Expect yours to require some amount of planning.
Actovision: Going OOM is supposed to be wand time. The idea is to have some way, somehow, of avoiding wanding, not something better to do when you hit that point.
Mana-using classes are not designed to go out of mana; they're designed that mana is a limitation around which we plan. If the regen system is working as designed, you only go out of mana if something goes wrong, whether your fault or the encounter going pear-shaped (as an aside: warlocks are supposed to go dead, not OOM). There's supposed to be a tradeoff between short-term DPS and long-term total damage, and another tradeoff between damage (of either sort) and safety margin.

This gets back to what I mentioned earlier, wondering why spirit and regen are treated as off-limits to min-maxing. People min-max DPS instead of total damage, and don't seem to realize there's a simplifying assumption lurking in the background that makes these two similar instead of diametrically opposed like they would be with a viable "burn" cycle. And then when people can't sustain it they whine on the official forums instead of taking what I consider to be the correct view of things (not you guys, wowtards in general). Min-maxing total damage over a given duration, rather than instant DPS, would give value to regen for high enough durations. Why is this approach considered off-limits? I think the psychological reason is actually that sustainability is almost reachable without some/all of the class tools, it seems like an oversight or miscalculation that it's not within reach, when I think the miscalculation is actually that it's so close to being reachable. If busting out all your tools barely got you to seven minutes in a frost spec and four in a fire spec, I suspect the total-damage rather than DPS would be the number under consideration in theorycrafting and regen would have some weight. Not that that would be good design, I'm just making a counterexample.

As a related question, giving up 3% or 5% crit seems to be treated like an untenable proposition (switching armors). Is that really the difference between viability and fail? I realize that mage armor currently isn't pulling its weight as a regen tool so I'm not asking about current opportunity costs. But in magical fairy-land where mage armor made your spells give you mana instead of costing mana and gave you a unicorn noncombat pet, is 3% crit too high a price to pay?
(nb: pre-TBC this would have been a rhetorical question. I understand nowadays FFB crits for an imperial asston so it's an honest question, since I don't play a mage.)

To answer the last part of your quote i'll paste what was typed on the last page:

It's pretty accepted that targetted mana-regen should have an opportunity cost. The issue brought up is that the cost ratio is not really weighted correctly at the moment, especially with regards to Armor spells:

- Spirit regen itself is poor
- Mages don't have much spirit on their gear (On 2 pieces of tier 7) and preferably to try to avoid it where possible anyway.
- Changing Armors invokes a nontrivial mana cost

Most people are not asking for a magical fairy-land answers. Obviously, losing 3% crit would be an excellent tradeoff for the ideas you suggested. People may exaggerate the "We lose Molten Armor Crit!" issue because it feels like you're giving it up for peanuts in return - and thats not fun. The equivilant would be me asking you whether you would still be happy with Lifetap if it restored only 20% of the mana it did now. You still *can* regen mana, but it feels more of a frustrating/silly thing to do. Thats what using Mage Armor (aka relying on spirit regen) feels like, especially for Fire/Frost mages.

Shadow priests may well have just as bad mana problems. It wouldn't surprise me. They are a dps class dealing with a healer-based mana regen model just like we are. The Spriests in my guild have not complained about mana regen at all, so I wasn't aware they had any issues.

Minmaxing total damage before running out of mana as opposed to DPS is only relevant some of the time. It depends on how it compares to a "burndown, go oom, evocate, burndown" cycle. Or if evocate is included, substitute "wand" for "evocate".

Currently 5% crit matters so much to a frostfire cycle that even though they can cast more bolts if running mage armor, it works better to just run yourself out of mana, recover a bit, cas another etc in the fights we're seeing. The extra time mage armor provides doesn't actually do more damage, it just takes longer to do same damage. This becomes even more true with the tier 7 set, because the gems become even more efficient at restoring mana and add spellpower that benefits from 5% crit, extending the burndown cycle even longer before hard OOM.

There is also a kind of circular logic here. If most of the time a full burn is the right way, you will hand off all the gear with spirit/mp5 to the warlocks (who get spellpower) or the healers (those that make use of 5 second rule or spirit, or anything with mp5 on it) because for them they either get more burndown DPS or more efficient sustained healing. This only gets worse as gear scales, because the return from Master of Elements (which benefits from 5% crit from molten and not at all from spirit or mp5) gets larger as you pile on more crit gear.

Unless and until those people in your raid are fully geared with the spirit/mp5 stuff you won't even roll on such gear. So in the fights where the equation is different you won't have enough spirit for mage armor to do much, and likely no mp5 stuff at all.

"Planning" is a fairly tired old canard in the mage mana discussion, in my opinion. It's usually used synonymously with the idea that you should know precisely how much mana you should consume for a fight of any particular length, which just isn't true for any fight that isn't Patchwerk. I think this is probably some of the justification internally at Blizzard, too, so I don't fault you for making the argument again. But it's simply not sound. Truth be told, nobody knows their mana consumption to such a degree of precision - nobody can know.

Let's take Sartharion as an example. If all I was doing was standing and casting, I'd have a pretty accurate picture of my mana consumption. But that's not what I'm doing. I cast a whole lot less when multiple Twilight Fissures get placed on or near me. I cast a whole lot less when Flame Tsunamis alternate sides, forcing me to move each time. When I'm AEing adds, if an add gets loose due to any number of possible reasons, comes for me, and I have to run it back to the tank, I'm casting less the entire time.

Running out of mana is only half the problem. Consider the flip side: what happens when I have more mana than I expected to have? Purely passive regeneration stats are only valuable if I run out of mana; if I wouldn't have run out of mana anyway, all that extra spirit/mp5 does is let me end the fight with more blue in my bar, which doesn't exactly score me extra points. That's basically a lost opportunity cost from not having something else on your gear instead of regeneration.

My point in bringing up those examples is to point out that the kind of planning we're being required to do is impossible. Not just difficult, but impossible. None of the examples I listed, nor any number of other random occurrences in any number of other raid encounters in the game, is the least bit predictable. And yet we're expected to predict all these unpredictable things, since our regeneration comes from gear which we can't change in a fight, and mage armor which, due to cost and time consumption, is also effectively unchangeable during combat.

The optimal mana situation for a DPS caster is to hit zero exactly when the fight ends, having maximized the use of their regeneration stat with no overage, and maximized the amount of mana spent. Warlocks are nearly always within one or two taps of the optimal situation, for fights of indefinite length (past initial OOM time). If you wind up with more mana than you expected to have due to any of those factors, which surely impact you as much as any mage, you just skip your next tap or two and you've caught up.

We are min-maxing with mana in mind. It's just that it isn't as simple as "Fight duration x seconds, y mana will be used, so I need this much regen to max my model." It's more along the lines of "a% chance this regeneration is totally ineffectual, b% chance that it meaningfully impacts my damage done", with a>b in current content. A small percentage of the time we'll run out of mana and be helpless, but is that worth gimping damage for 100% of the time? Not so much.

To sum up: we can't plan for the unpredictable. All we can do is run with a best-guess situation - but we have no tools to handle guessing wrong. You might call having a reactive mana system class-defining, but I call it necessary. Proactive mana management doesn't work when random events occur.

In a raid environment I find it nearly impossible to run out of mana. I can't remember the last time I had to evocate on anything. I've only used it on trash per say.

Epiphenom: I did not mean to imply that your mana management system should be entirely pre-planned. I agree with your post in that you have, at best, an approximation on random-stuff fights, making the task literally impossible. All classes still need oh-shit buttons, and if mages can go OOM under non-ideal circumstances you do need a reactive method of dealing with it. Having generally variable mana situations exacerbates the problem. Any mage regen model will need on-demand reactive mana. No arguement.

My point is not that you have to plan everything in advance, just that you will have to make one or more decisions somewhat in advance. This does not mean all your abilities will be proactive. You may even need more reactive abilities than you currently have in order to have a proactive model sufficently robust enough to handle encounter design. The choice of armors, and forecasting when to transition to a burn-cycle, are examples of planning forward that would fit into a mage regen model that wouldn't at all work under the lifetap model. That could be the total extent of forecasting, and still sufficiently differentiate your playstyle with regards to mana from other mana-using casters. Warlocks have absolutely zero proactive mana abilities, and will likely stay that way because of lifetap. Mages will (already do) have proactive mana abilities, and a much more limited, but certainly not empty, selection of reactive mana abilities.


Currently evocate is really your only reactive mana ability. It has a cost (channel time) so you don't use it unless things get rough, and it has a cooldown meaning you still have to plan somewhat to be in the situation where it's going to be enough to save you if things go wrong. Mana gems aren't a reactive abilities because they have no downside, so you chew them as often as you can (recasting mana gems could be a reactive ability). Is evocate enough to cover your reactive needs? I really have no idea. 60% of your mana pool is a big number, one could make the arguement that this one ability is big enough to cover quite a large catastrophe. If I were designer, I would probably prefer to put in a second smaller gap-plugger as well, to make it easier to cover for fight variance instead of just catastrophes. Again, something with a cost and a cooldown that you wouldn't want to use unless something bad happened. But this is starting to get speculative.

I personally cannot stand the choices we Mage's are having to make during raids for mana management. Yes, the change to the gem mechanic is better, but even with wasting a glyph position for the additional bonus its simply not enough.

As for Molten vs. Mage armor, the choice is clear, from a statistical perspective coupled with the glyph molten wins hands down. I've been experimenting with mage armor and glyph during extended boss encounters, and noticed that I no longer need to evocate, but of course I drop 2-4 positions on the meters.

Which is more important, sustained dps for the duration, or all out burst followed by OOM wand slings? (extended boss encounters only)

Again with this idea that we have a 'burn-cycle' or mana dump option. The ONLY thing even approximating this is Icy Veins and any haste use trinkets you may have. We have 1 cycle - gogogo. We stay on that cycle until we have exhausted our mana pot, mana gems and evo then we wand. The only real way a FFB mage can temper their mana is to not keep up LB (taking a way a serious chunk of dmg) and to not take advantage of Hot Streak (again, a dps loss.) Neither of those things do anything other than extend the time to oom.

Keep in mind that the #1 reason this is so hard to balance, is that if they make regen stats required to maintain DPS over a long fight and balance DPS around it, then DPS would be too high on fights where there's more than enough mana without those regen stats. If they balance us based on fights where we have more than enough mana, then we're too weak on fights where regen stats are needed. For warlocks, on the other hand, the duration of a fight where you wouldn't have to lifetap is so short, that the DPS lost by lifetapping has rather low dependancy on the fight duration.

I don't see how that can really be solved without somehow normalizing fight durations (and assuminng what buffs you have and don't have) before making any balance changes, and even then it'll be complicated. That is, as long as they don't want to give mages infinitely sustaineable DPS in some way or another.

We don't need infinitely sustainable dps* as there will never be a fight duration that is infinite, that's not really the problem either. As far as I understand the developers intention is to make classes with an infinite resource shine a little brighter on fights where the mana bar is struggeling and vice versa, which is an acceptable design. I think Ghostcrawler was implying it in some thread regarding rogue dps(1st post, 2nd quote, 3rd paragraph).

The problem with this is though that there is hardly any encounters where mages are struggling with mana but I think ghostcrawler(again) stated that this will change or they want this to change, as in we have to manage our mana and here's the problem.

We don't manage our mana, we make a proactive decision of using mage armour/molten armour, we use mana gems whenever we need them(or whenever we can with 2p t7) and we use evocation as a last resort and then we wand. We're not actively managing our mana but simply depending on the encounter design to not run us oom. The only thing you could really classify as mana management would be evocation as it's the only spell that actually has a dps penalty for being used and is based on an active decision until you realize you will need the whole lot of the mana it gives. But making mage mana management revolve around how many tics of evocate you should take before continuing to nuke just feels very clunky as well as restricting the fight conditions under which a mage can operate properly.

Now of course it would be fine if they never designed an encounter where a mage feasibly runs oom but that would not make us manage our mana, because we can't.

I might be taking ghostcrawler too serious regarding mana management but it is scary if we're supposed to manage our mana without any tools to do so.

*It would make it a lot easier though, can't argue with that assuming it's an active decision with a dps penalty.

Old Evocation:

When activated, Evocation immediately activated spirit regen (the OO5SR regen) and increased spirit-based mana regen for 1500% for 8 seconds.


I mention this not as a solution, but to head off anyone else's mention of it as a solution.
Old Evocation would bring more importance or use to spirit on our gear, but the situation would remain that Evocation is a button we would press once during a boss fight to regenerate up to a potential second whole mana bar.

It could be more of a likable tool, though, if Evocation was broken to a much weaker spirit-based active regeneration tool to be used often like Lifetap, with a short (under 1 minute) cooldown. This makes it so that it would be a Mage Lifetap that returned mana based on how much spirit you had rather than based on a tooltip value of mana restored.

Mages would just go back to having a +spi staff they swapped in, whenever they needed to evocate and then go back to their regular dps rotation.

On the design of mana dps classes in general:
Having unlimited mana in every fight is Blizzard would want to avoid, as it takes out mana managemnt entirely.
Having limited mana in every fight would require dps to be balanced around limited mana. Thus on short fights dps will be too high (using higher dps but lower dpm spells), and on longer fights dps will be too low (not enough mana).
Keep in mind Blizzard tends to make the "last boss in the instance" an "epic" fight which is exceptionally long and usually also harder than previous bosses.

I think the design should be such that some classes need to mana mana, and some dont.
Classes that need to mana mana will gain the ability to do more dps then they should do (compared to the others) for limited time. For example mages via 3xAB cycles.
The downside to having to manage mana should be that these classes do less dps than the others on very long fights or fights which have mana drain. On such fights these classes should fall back to higher dpm, lower dps spells.
All these spells should be accessed within the same spec for those classes. So these classes will not have to respec for one of short, medium or long fights each time.

Some examples:
* Ret paladin's spells could be designed in such a way that doing judment and crusdaer strike on every cd gives decent dps, with very high dpm, used on long fights. On "normal" (medium length) fights the paladin will also use divine storm for extra dps, but dpm will go down. on "nuke now" phases, the paladin will be able to also concerate (vs single target) to do even more dps, but the dpm will be very low so the paladins will go oom shortly after if he keeps this cycle.
The paladin's dps will be balanaced around the cycle that uses divine storm but does not use concenrate.
* A mage can have fire/frostfirebolt spam on long fights, adding living bomb on normal fights, and adding something else for extra nuke on very short fights.
Note: I am talking about a major redesign of abilities.

It is importatnt to mention:
1. Say a mage using is "normal" rotation should be about equal to the other dpsers (like hunters or warlocks). The dps balance should be around the "normal" rotation, not the max possible but unsustainable cycle and not on the "very cheap that can sustain it forever" cycle.
2. Managing mana should not be just a limiting factor. There should be an advantage to having to do so compared to the other classes, such as the abliity to do exceptionally high dps for low dpm.
3. If you don't suffer the penalities of managing mana (you dont do less when there is mana drain or very long fights) then you should not get the advantages of managing mana also (no extra burst abilites). Warlocks for example could potentially go either way. They could be a class that mana is not an issue, or they could be designed so that lifetapping will punish them enough so they will lose the same as say a mage when the fight has mana drain or is exceptionally long.
4. This whole post is aimed at PvE design ignoring PvP. Maybe a different model will be needed for PvP, this is beyond the scope of this post. Healers mana managment is similary ignored by this post.

I think it's an exaggeration to say we don't manage mana. Deciding which armour to use before the fight counts as mana magement to me (not that it's relevant in current 25-mans). There's also the option to switch reactively to Molten Armour if you have more than you expected (that's the only real reactive option for armour and TBH I think I've only really used it on Gruul, Prince, and Noth, and i stopped using it on the first two).

There are indeed some reactive choices we can make other than evocation; for example blink or run can be a straight DPS option on Thaddius, and a survival versus DPS option in some encounters, especially when you don't know them so well. And you must manage your mana when AOEing on Noth, for example, to make sure you are not OOM in the next boss phase.

I agree we need more options; I don't see what's so bad about Evocation, though, especially if it can be used more often in future. Yes it means you have to stop doing damage for a while; feel free to wand if you prefer, but spare us the exaggerated grief that any activity of a mage might temporarily stop DPS. Consider using Icy Veins (or even Ice Barrier) if you're concerned that Evocation will be interrupted, or else make good decisions (management again) about when to use it. Don't gear for more haste than you can handle; consider having swappable gear with more or less haste.

Since someone brought it up earlier: [Glyph of Mana Gem] really needs to be looked at again by Blizzard. Perhaps with 2 piece t7 bonus it's all the designers wanted to put on a glyph, for now, but when we outgear our t7 it's just a laughable glyph that in theory has some good potential. The upside then, would be players who care about gems/reactive mana would be able to make a choice and glyph to improve this area of their preferred mana-management playstyle. However, the poor quality of the glyph rules rules out that ever having to make that decision from any serious players mind until there are some significant improvements.

The obvious solution is to split the 2 piece T7 bonus and make the +40% mana component a major glyph, whilst keeping just the +damage component as the actual bonus for the tier gear. The tier bonus is great and i'm sure there will be a collective sigh the day we outgear and lose it. With a small change this glyph could help encourage meaningful decisions with regards to your mana management: Do you want to use a major glyph slot to improve your reactive/emergency mana restore ability - or not? Currently however, there is no choice or it's a silly choice at best (Lose a dps glyph for +10% mana restored?) - which echoes the sentiment felt on this topic in other areas of the mana management umbrella.

Give us solid options, varied viable choices to make - and make them fun.

Due to the trouble in swapping glyphs, I would hate seeing mana solutions brought through glyphs you have to take over DPS glyphs. That's as bad as forcing you to spec differently for longer fights. No other class needs to do those kinds of stuff.

All that does is compound the situation where we're making at-the-trainer decisions by adding yet another occurrence of "do I glyph for mage armor or molten armor?"

Mana isn't something you should have to primarily plan for at the trainer. It's something that happens on a fight by fight basis, and a dungeon-by-dungeon basis. Mages need fewer "at the trainer" and "before engage" options and more "in the moment" options.