I see a lot of complaints about how the arcane rotation is suddenly very static and I just don't get it. You have 1 spell that you can use any time, one on cooldown and one as proc. There are no less rotations than we have today in WotLK. And there's no lack in dps/dpm tradeoffs either. Just looking at the mastery I don't see how you can come to any other conclusion than that the rotations will be dynamic. The way I see it the mastery alone dictates the mana management. Even without any extra tools, just using the different rotations and evocation/gem there's plenty of mana management to do just because of the mastery.

I guess the main problem is that the mastery is very mathy and most players don't understand yet its impact on how to play the spec optimally. The main thing to understand about the mastery is that in between mana cooldowns you will want to use cycles in order of increasing mps in consecutive order, starting with low dps/mps and increaseing to high dps/mps the closer you are to mana cooldown. Choosing at what dps/mps level to start and when to switch between cycles and how to react to rng will be the major thing that will distinguish an average mage from one taking the spec to its full potential. For more details on the math behind this you can check Cataclysm Mage Changes and some previous posts by Roywyn and Muphrid.

Personally, I don't mind slow being a prereq, but with conditions. Basically, I don't want to be forced out of key talents in order to get it. Looking over the trees, I don't really feel one point behind on anything truly important that I wish I had the point from Slow back for, and Slow is one of those handy spells which I haven't taken in wrath mainly because I couldn't afford it, not because I don't want it. If making it a prereq is what it takes to allow me to spec into Slow, I'm happy with that, and other classes should be so lucky as to have required filler as useful as Slow. That said, I'm completely in favor of getting rid of TTW.

As to mana management, I'll agree with the general tone that the current set up doesn't get there. In order for mana management to work, it will need for mana tricks to be something other than simply hitting Mana Gem or Innervate on cooldown. If mana management is simply using mana abilities on cooldown and maintaining the same rotation, that is an utter failure of the design. Theoretically one tweak could be new spells, but a new nuke would probably wind up being a bland one, and how many "X second cast, deals Y type damage, no frills" nukes do we need? Some thoughts.

1. Mana Adept as a negative synergy on a continuous, linear scale is utterly unintuitive, and requires spreadsheeting in order to determine not just perfect use, but above-average use. Adding some sort of breakpoint to it would make it understandable, and provide a target for mana consumption. Something like "Becomes less effective below 40% mana" for a continuous but non-linear bonus or "While above 60% mana, Mana Adept adds XX% damage, based on mastery" for a discrete bonus. A situation where almost any rotation above a mana target is superior to the maximum rotation under the mana target would be ideal, in the sense that it makes the intended use of the mastery clear. Such a scheme might or might not be a problem in PvP. Clearly, a well-timed mana burn could deprive the mage of a good deal of power, but that might not be a bad thing in the sense that any PvP spec should have some vulnerability or another, and if Arcane's is mana burns and the spec is more robust against other kinds of attacks and control, that could be seen as a feature not a bug. If Arcane is more vulnerable than other mage specs to various attacks and control besides mana burns, then there is clearly a problem overall. It just strikes me that having some sort of target value for mana that doesn't require a spreadsheet to calculate would be a good step toward an intuitive design for Arcane.

2. There will need to be unsustainable-but-otherwise-superior rotations in order for Mana Management to work or be interesting. I view this basically as a truism about Arcane. If the superior rotation can be sustained by mana regeneration, then all interactivity disappears as all mages simply use the highest. In other words, a rotation which is higher dps in the short term will lead to an eventual long or medium term dps crash, and something which seems to be a lower dps rotation will win out over a longer period. Think of it like a warrior in live, where particularly at lower gear levels, you couldn't maintain Heroic Strike on every swing (despite it being higher short-term dps) due to rage starvation. The concept or making a choice to do less damage in the short term because of medium term consequences is not foreign to WoW players.

3. The flip side of this is that there needs to be a mana-sustainable "ABC" rotation. If we can't Always Be Casting, then this too is a failure. I mean, they could redesign Arcane Mages to be like caster-Rogues, regenerating 10% of their mana per second, but with Arcane Blast costing upwards of 35% of your mana per cast, with wanding-while-casting so we don't get too bored, but I don't think anyone would really enjoy that.

4. The whole short-term dps, long-term dps needs to be balanced in understandable lengths of time. If we need to balance how much mana we use around the total fight length (several minutes), this is spreadsheet work and thus bad design. There's no way a player can get any sort of mental grasp on that. If we are designed to balance our mana consumption around how much we'll use in, say, 20 seconds, that's fair and something which a moderately-skilled player should be able to understand. That is, high burn mana rotations need to crash fast, and high regen rotations need to regen fast.

5. I think we need to take a close look at how other classes work and try to pull lessons out of those. The goal of Mana Adept seems to be to transform the Arcane Mage from a resource-independent spec (other DPS casters, Ret Paladins, and Live Hunters, to various extents) where you follow the same rotations and priorities at all resource levels, into a resource-dependent spec (like rogues, warriors, DK, and Cata hunters) where use of abilities will depend on what resources you happen to have available or unavailable at any given time. Each of those classes is designed around always having the potential to use more than you can generate, where that generation takes place over a short-to-medium time frame. Also, resource generation is not entirely predictable. Rage can spike due to crits (bonus rage on crit removed in Cata), can drop due to missed attacks, can spike due to incoming raid damage. Most rogues have various energy procs, and CP generation is not entirely controlled, with chances to generate extra CP showing up fairly frequently. Mutilate Rogues in live show this best, where the combination of Seal Fate and Ruthlessness force slight-but-interesting changes in the rotation every five seconds or so.

N.B. I don't consider spreadsheet optimization to be bad, by any stretch. However, I think that if a tree design requires more than napkin math to do something other than mediocre, then that is a problem. If it works out that simply mashing buttons or hitting the prettiest ability to use generates 65-70% of maximum dps, that a little napkin math or consideration brings you up to close to 90% of maximum, and you need to spreadsheet to get that last 10% dps, that's a perfectly fair model. Something where wild play gives 90% effectiveness, or where spreadsheets are required to get within 25% of theoretical maximum just seems like bad design to me.

///

The real question is how to put it all together. I favor a system more like a mutilate rogue, where we're altering the number of Arcane Blasts we cast fairly frequently. Something like 4 AB above 80% mana, 2 AB below 60% mana, and 3 AB in the middle, with ABarr forcing rather than proc fishing, and OMGWTFBBQ INERVAET NAO if you drop under 40% mana. Something which keeps me looking at my mana bar every few seconds and picking out in advance how many AB I want in my next cycle. It'd probably need Mana Adept sweet spots which are clearly identifiable, and some sort of mana proc to break the long-term fixed cycle problem. Make such a mana proc large enough to cause me to alter my behavior for the next few seconds, but infrequent enough to allow me to simply ignore my mana pool by counting on enough procs to keep me topped off. In other words, I think stack size management should be the tool by which we control our mana level, rather than the introduction of a new spell.

Basically, every class needs to make decisions about ability use every few seconds. Some people need to watch DoT uptime (feral druids, shadow priests), some need to watch a large number of procs (frost mages, fury warriors - well, at least Cata furies. instant slam procs, not always usable raging blow, not always affordable heroic strike), some juggle cooldowns (enh shaman, paladins), others combine them (demo locks, ele shaman), and some just do something weird (boomkin, but boomkin are weird in general). Arcane seems to lack those decisions right now. If it's simply fish for a proc like live, then it's a total failure to create a more engaged playstyle. If the method by which they introduce decisions into short-term gameplay is by watching exact mana values like a hawk, and they give us the tools we need to adjust mana on the fly, I'd be happy with that. No doubt, there are other ways to bring in mana management (or ways scrap mana management and come up with some other sort of short-term decision making), and this just happens to be the idea I favor.

The thing is, they've decided Arcane's focus was going to be mana efficiency. Our cooldown spell is fine in that regard, as it's not terribly expensive, and the proc is even better with missiles being free. The problem is the fact that our main spell is also something that stacks a mana-inefficiency debuff. If we use Barrage on cooldown, the earliest we can break stacks is at 2 (5 second cooldown Barrage, 2.5 base cast Blast). We can't reliably use missiles early to break stacks because now the base spell is a proc. At least on live we can decide to use missiles early to break stacks if Barrage doesn't proc. We can slow down Blast casts to break it at 1 if we're really desperate, but that also means less chances to proc missiles, so that's compounding the damage loss.

The kicker here is the fact that with a main nuke that directly opposes the goal of the spec, if we need to to use something else as filler, we're hindered by the 25% Arcane damage bonus from our specialization, and really have no viable option. Sure, we could sacrifice 25% damage for a mana conservation rotation and use Frostbolt or Fireball instead, but that's not really viable.

So, the best reliable mana-efficient rotation we have is AB2ABarr. If Missiles procs during that, use it instead. So, the only decision is, do I continue to stack it to 3 or 4, or do I have to leave it at 2 and break it right away? Mild control, at best, considering their description of a fun management game.

These two points reflect my own feelings.

Having a rotation that is reliant on RNG to cast one of its primary spells does not appeal to me. That's why I liked the auto-AM activating version of Arcane Barrage so much; it gave me a choice. Even if Arcane Missiles keeps it's relatively high 40% proc rate, I'll still run into frustrating instances of bad luck where I'll cast four or five spells before I am able to cast AM again.

RNG is like playing whack-a-mole blindfolded. Sure, you'll eventually hit one, but it's not because you're skilled. I can't take pride in that, and I find it more frustrating than enjoyable.

I had an insight earlier today.. Arcane is like playing a Cataclysm Resto Shammy with Healing Wave being a proc and no Greater/Lesser Healing Wave. I mean, you can heal pretty well with Chain Heal spam. Under certain conditions it might best to spam Chain Heal. You even have a cooldown instant that you want to throw into your rotation. But, if you want the big numbers or more efficiency, you have to plan your rotation around Healing Wave procs. If you want to conserve mana, your only choices are to stop casting or throw in more Riptides. I guess the analogy works somewhat.

So I was thinking, maybe we can borrow a little from the Resto Shammy design to make Arcane more interesting. Arcane definitely needs something other than Arcane Barrage to cast on a consistent basis. So add a spell, borrow a spell, or make proc-only Arcane Missiles go away. I personally, think making Arcane Missiles usable anytime is the best answer. I would then borrow the Tidal Waves and the Riptide buffing Chain Heal mechanics from Resto. It could look something like:

Arcane Waves - Arcane Missiles is now usable anytime. In addition, when you cast Arcane Missiles or Arcane Barrage, you gain the Arcane Waves effect, which reduces the [mana cost or cast time] of your Arcane Blast spell by x% and increases the damage of your [Insert whatever spell Arcane is going to end up using for aoe] spell by Y%. 2 charges.

Arcane Barrage - Launches bolts of arcane energy at the enemy target, causing X Arcane damage. Also, your next Arcane Missiles cast within 15 sec will have it's damage increased by Y%.

It gives you semi-dynamic rotations with extreme mana efficiency if you go for mana cost reduction in Arcane Waves. And, as much as I hate copying everything Resto has. Healing Rains might be a good starting point for an Arcane AoE:

Portal Storm - 2 sec cast - 10 sec cooldown - Calls forth a Portal Storm to blanket the area targeted by the Mage, Dealing X to Y damage to enemies in the area every 2 sec for 10 sec. [Potential Secondary Effect - Off-school AoE damage boost - Our single target spells to deal AoE damage when cast on targets in the area - When you cast Arcane Explosion a Mirror image of you appears in the Portal Storm and explodes as well]

Anyways, just some of my inane ramblings.

I think you´re forgetting why they changed it to a proc in the first place. I really agree that it can be annoying when something just refuses to proc. But the problem with AM castable at will is that it´s very hard to balance. I´d rather have a good proc and run the risk of bad luck streaks than have a bad nuke I never touch. Also with the new barrage it will be somewhat castable at will. Im thinking of it as a DPS version of swiftmend that can also proc and everyone loves swiftmend right?

Oh no, I definitely remember why they changed it to a proc. However, that doesn't mean they accomplished their goal.
And how is AM a nuke you never want to touch? AM (even un-MBAMed) served as the default, go to method to clear an AB stack pretty much throughout the course of WoTLk. If nothing else, we touched AM a hell of a lot.

Even so, that is only part of the story. Procs as a core mechanic are ok if there are enough of them which are varied enough that each proc produces some unique and interesting gameplay. But basing the entire gameplay of an spec around a single proc with a single effect (especially one whose chance to proc you cannot influence in any way) is just bad and boring design. I really don't know of a simpler way to put it. That being said, I most definitely believe that procs are not the only way to create good gameplay, which is why I am hesitant to just look for solutions to arcane's problems by introducing tons of procs.

To explain with an example, take a look at frost (I know I pick on frost and fire specs a lot but I think it is valid since they are close to home). Frost has procs, but it's procs 'work' since they are so many of them that there is pretty much always something proccing (hence giving you something to react to at all times, even if those things are different). Plus, these procs are supported by talents that allow you to improve your chance to proc them. Furthermore, these procs in frost are not the only gameplay to the spec, they are just a part of a much more complete package.

A spec whose only gameplay is a single proc is a boring spec to play. This is not something new, in fact, it is the exact reason why the general consensus is that Arcane is a boring spec to play right now on live, as well as a boring spec to play in beta.

Overall though, I think most of us have already concluded on these things. This thread, to some extent, is an effort to solve the "Arcane is boring" problem, amongst other things.

Swiftmend is one small part of the Druid toolset. The difference between that and Arcane is, Druids have a bunch of other cool healing tools in their toolset which add depth. People like Swiftmend because it's a fun spell - but also because it's one tool of many which a Resto Druid can choose how and when to use.

If The Resto Druid healbox consisted only of Rejuvination, Swiftmend and Healing Touch - people wouldn't like Swiftmend so much. They'd complain that Swiftmend is a very frustrating mechanic, as it's the only way to get heals off on targets between two extremes (Slow small Rejuv heals VS Slow big Healing Touch heals). Comments would include how boring it is to use this proc religiously on cooldown with little choice. The sentiment would be, "Why can't I have other fun tools to manage my healing? Too much of my gameplay depends on mindless use of this proc, that's boring!"

The latter is akin to Arcane mages currently. Missile Barrage by itself can be fun, but having the entirety of Arcane based around one proc that affects this one spell, is a little excessive. If Arcane had several other compelling Mechanics and means to DPS (Dynamic rotations, different spells to use, ability to modify existing spells etc) Missile Barrage probably wouldn't be such a big deal. But because the rest of the spec is quite bland, it just further exacerbates the problem.

Imagine if Missile Barrage was actually a utility proc.

When it procced, it modified the next spell you cast. Arcane Missiles gets the current speed increase, Arcane Barrage gets the ability to hit multiple targets, Arcane Explosion grants you X% of the damage dealt back as mana, Flamestrike/Blizzard got a +25 damage bonus (to compensate for having no specialisation), Arcane Blast got the ability to stack to a 5th stack, which was extremely mana unsustainable but great dps (Only a Missile Barraged Arcane Blast could push the stack to 5, not spammable or refreshable to 5 via normal AB cast means). Suddenly the Missile Barrage proc appears - and we've got a pretty interesting choice, "What should I do with it?". AOE? Use it for Mana Regen? Shift up gears into the (5) stack rotation? A Hasted AM? Currently the answer is: Use it for the only thing you can, Arcane Missiles.

One of the more satisfying aspects of playing Arcane in TBC was knowing you had to end a fight with zero mana - and manipulating your play to achieve this. It was a lot of fun to manage your mana carefully for a given length encounter, then as the end approaches - blow that out of the window and spam the rest away, timed perfectly to hit 0 as the boss dies.

Any mana you had left, was viewed as wasted DPS. It would be nice to bring this back in some form. You could call it "Arcane's Execute" if you will. Even if Arcane had a special ability to get a 5th stack and cast Arcane Blasts for a ridiculous cost (10k mana?), I think many would enjoy the last-few-second burn of spamming away your mana for a sizeable DPS reward - and go out with a bang so to speak. Mana Adept should be rewarding mages (in its current form) for keeping Mana High, but that doesn't mean we can't also have a mechanic that basically says, at the end of a fight "To hell with that!" - and let you spam it out the window with Execution style mechanics. That's fun! It could be quite interesting that "Mana Management" will involve a balancing act between Mana Adapt (Keep Mana High) and Execution ABlast type spams (Unload all your mana before you exit combat). The Cataclysm numbers for spells are not finalised, so it's hard to see whether ABspam is going to be this effective, or whether you're just better off sticking with the current rotation involving Mbarrage instead.

Anyway, some of this could help towards killing a bunch of birds with one stone: More decisions, more depth, an added burn cycle, a means for ranged Arcane AOE. The point is, we don't necessarily need to add complex new spells and mechanics to do it, we can just play around with existing ones. A Missile Barrage proc can still a core part of the Arcane tree, it just doesn't have to be the bland proc it is now - limited to Arcane Missiles.

Well phrased like that I definetley haveto agree, the idea of changing the proc to react diferently to several spells was very good. Having AM as a spell or a proc still only results in it being one single thing and won´t add loads of "fun" either way. But a proc like the one you describe would be awesome. Besides adding depth to the playstyle it still feels doable before launch. It´s a little late to be expecting completely new spells or anything requiering lots of new graphics and the like. But procs to make several spells behave diferently is another matter.

With the new power auras feature procs will be less annoying for a lot of people to keep track of, perhaps something like having all three effects you described as separate procs that all come up and if you like you can use all three of them before the timer runs out and then the trick is knowing when you actually benifit from mixing aoe into your rotation at all. Having several procs like a hand of cards and being able to save them a few secconds even if you use one might be interesting and make for very versatile gameplay in gauntlet type events.

Well whatever the proc gets changed to it should not be consumed by Arcane Blast. I mean you're already going to be at least .5s into the next AB cast before you realize it's proc'd. So you'll have to cancel cast to avoid using the proc in most cases.

I'm glad you guys liked my healer analogy.