The three different talents merits will depend primarily on the fight and secondarily on the spec, hence we could write a whole book about the combinations they could produce. What I get as a general sense is that the game appears to have created a tradition of sustained incoming raid damage hence in most encounters I suspect Incanter's Shield will proc. In encounters with very sporadic bursting such as Spine, all specs might benefit from Invocation, however, if one can find a way to reliably proc Incanter's Shield, they might benefit more that way. The Rune of Power talent appears to be designed for sustained DPS or for longer two-way bursts, such as the Zon'Ozz encounter, where one could drop a rune at two different positions (Chimaeron was also a good example). But, whether that would be preferable to Incanter's Shield or even Invocation (even on non Arcane specs) might be a dilemma requiring good math to derive an answer.

i.e. The 3 talents do appear to overlap in duties and I suspect in some encounters the choice will be obvious but in some others intuition will not be easily enough. This will be true for all three specs.

I don't quite understand how people could merit various talents as "too narrow". The whole point of the new talent system was _not_ to have a single "best build" anymore, but rather to give the player choices. Also, Blizzard already pointed out that they were planning to make the dynamic switching of talents fairly easy. Therefore, I'd rather assume that each mage can reconfigure for various boss fights like he sees fit.

Personally, I find this concept incredibly interesting. At least I know that if I walk through the FL and DS raid bosses in my thoughts, I could easily come up with fights where I'd pick different mutually exclusive talents. It's like taking the right tools out of a toolbox for the requirements of each fight. And this is something we don't really have at the moment.

The runes only last for one minute after being cast so they require re-casting throughout the fight. Regardless it will be the go to option unless the fight makes it not viable. Incanter's, though, syncs better with cooldowns. Lining it up with Combustion / AB spam could result in a bigger net gain.

And this is where I think the Rune of Power is going to run into problems. As a ranged AOE spell, we're going to have to expend time every minute setting the AOE reticle and casting a spell that does not do damage. From a time management and overall DPS perspective, it would be far easier to have RoP be a self-cast spell, because that then lands the rune where we are and allows us to continue casting unabated.

One thing that I don't think has been thoroughly addressed with Incanter's is how large the shield is. The tooltip states that it will return up to 30% of our maximum mana. Given that mana pools in MoP are likely to be static (and assuming that the number will be in the 100k range), that means somewhere around 30k damage can be absorbed over 8 seconds, and until that number is reached there will be no break of the shield to trigger the spell power boost. This sounds very much like an end-fight ability with some minor benefits on fights that consistently provide that kind of raid damage (Chimaeron, for example). Unless of course the designers are planning to empty healers' mana pools on every fight through blunt-force raid damage, which is certainly possible. They could also just make it a smaller shield with a multiplier for damage conversion to mana to increase the frequency of breaks, which would help it as well.

On a Patchwerk fight? Sure, we might wait the whole minute to recast. Any other fight? we will likely see a new "skillcap" in regard to placement and frequency of use. Think about it. On a fight like Ultraxion, you can simply re-up the rune at 58.5sec for 100% up time. On flights like.. maybe Morchokk, we can cast it as needed at our destinations without needing to wait an absurd amount of time for a cooldown. With only a 10sec cooldown (as it currently stands), the spell has incredible opportunities for flexibility, among other things. The only question I have at this point is whether they will introduce a mana cost? And what might that cost be?

I don't see a reason to introduce a mana cost for an ability designed to return mana. Evocation has never had a cost associated with it. The trade off is the time spent channeling/casting.

And ... finding the darned mouse pointer in a clutter of spell effects, Power Auras, MSBT text and doing so while moving is my biggest difficulty with both Fire and Frost specs. It isn't so much the time required as the detailed attention to a screen that is already drowning in data. Spamming targeted effects also requires a macro so that you don't instantly turn the effect off. Then spamming one key to trigger the effect followed by clicking the mouse button is very much a "clunky mechanic."

All those issues I mentioned are very likely not issues for excellent mages, but they are for the rest of us. I believe that Kyrilon was trying to state that it was a clunky mechanic and I agree with him.

In order for rune of power to be better than invocation (for arcane mages at least) the mana regen it gives would have to be at least as good as the regen we currently enjoy from Mage armor. If the regen isn't as good, invocation will be my choice, due to endless burn phases without the need for evocations cooldown.

On a side note, i can see incanters shield as being much more situational, requiring either a high incoming damage fight, as discussed already, or an "I'm standing in the fire to regen mana." it strikes me as a more pvp talent, except if glyph of evocation still regens health. With evocation having no cooldown with invocation, it probably won't survive as a glyph.

Edit: for another thing, raid leaders may require us to spec rune of power for prolonged fights to help healers/other mana dps classes regen mana. I'm assuming rune of power works on anyone who stands in it.

Edit again: yeah on second reading of tooltip the word "your" makes me think I'm wrong about others benefitting from runes

***
Edit by Heef: As discussed in the warning for this post, please do not post misinformation. I am leaving this post in the thread because it does include some relevant discussion, if not in-depth (we are not going to have the numbers to figure out the true merits of these talents for a while). But please be careful in using the informtion that we *do* have properly.
***

I spent some time calculating the strengths of the buffs to our level 90 talents (Invocation, Rune of Power, Incanter's Shield), and they all look to be roughly even after the beta went up.

During Invocation, you'll spend 6 seconds casting Evocation regularly in order to keep up the buff. The multiplier to your base damage done would be calculated as: (100% * (n - 46) + 130% * 40 + 0 * 6) / n. Where n is the number of seconds between evocations. We spend 40 seconds under the 30% damage buff, 6 seconds doing no damage while evocating, then the rest spend under normal damage. The maximum buff uptime would equate to n = 46, which means we'll have an average multiplier of 113%.

In a Rune of Power, if you are really skilled/lucky and only cast a new one every 58.5 seconds, your damage multiplier would be: (115% * 58.5 + 0 * 1.5) / 60 = 112.125%. Even if you have to cast it once every 30 seconds due to phase changes that invalidate your old rune of power, the multiplier would be reduced to 109.25%.

Incanter's Shield is the hardest to predict without knowing how regularly damage comes, if the spell naturally expiring counts as it being "destroyed", and if we can reliably force it to break with maximum damage. Assuming that you can break it on cooldown, which maximizes the spellpower multiplier, it would be: (1 * (23.5 - 15) + 1.3 * 15 + 0 * 1.5) / 25 = 112%. Whether the spellpower boost is better than the damage multiplier would depend on the spellpower coefficients, and whether the shield can be broken. However, it does double duty in protecting you from damage with a shield equal to 30% of your mana pool.

In short, assuming these numbers hold, whichever one suits the boss mechanics or playstyle would probably win out. Rune of Power has the strength of only needing to be maintained every minute or so, but it restricts your movement to the little rune you place on the ground, and ground AOEs may invalidate the buff for a short while. Invocation makes you unable to run out of mana under any circumstance, and you can sync it up very easily for more powerful cooldowns. But, it has a drawback in that if you start evocating at a wrong time and have to move, you don't get a buff and you waste more time recasting it. It'll heavily punish those who are unable to properly time evocations with boss mechanics. Incanter's Shield is highly dependent on fight mechanics before we can fully evaluate its power, but its protection may allow you to survive otherwise deadly situations.

Don't forget, Sunaseni: You may have 2 RoP's at one time, and there is only a 10sec cooldown. Movement fights will still carry penalties, and I would agree that even with this in mind, the value of RoP would be decreased slightly. I don't think the loss will be as great though, unless there are a lot of fights like p2 Marrowgar (1st boss in ICC) where the movement is constant, and not linear in terms of "move from A to B then back to A".

Quick-hit thoughts on beta Frost:

1. Shatter now increases Crit chance against Frozen targets by 2x+40%, and increases Frostbolt damage against frozen targets by 20%. So our Shatter Crit cap for MoP appears to be scheduled at 30%, assuming no Crit suppression.

2. Brain Freeze is now proc'ed ONLY by your level 75 bomb talent, so uptime on that will be critical. Brain Freeze no longer interacts with FoF.

3. Frost Orb's animation is slick. SLICK. However, 10k mana per cast, given a 100k mana pool, seems prohibitively high to maintain usage in a rotation. Level 90 mana regen talents may improve this outlook.

4. Deep Freeze no longer causes damage to Freeze-immune targets (tested on dummy).

5. Water Elemental actions are currently broken after first-time use. Have to resummon to get the buttons and CTRL bindings working again.

6. Fingers of Frost: Successful Frostbolts and FFBs have a 15% chance, Frozen Orb hits have a 35% chance, and Scorch has a 10% chance to proc FoF. Also, Freeze effects against Immune targets always proc FoF. FoF causes IL and DF to treat targets as Frozen, and IL damage is increased by 25% for 15 seconds.

That last bit intrigues me. Glyphed Fingers of Frost provides a third charge, and testing on the dummy shows that the 25% increase boosts FoF/Shattered ILs after the first IL triggering hit, so it looks like the rotation will be something along the lines of:

1. Keep [Tier 5] Bomb on CD (cast time + duration is currently shorter than CD, so no 100% uptime)
2. Frost Orb?
3. Brain Freeze
4. If 3 stacks of glyphed FoF, dump into IL
5. Frostbolt

Given this basic rotation, Haste appears to once again be the primary stat, with Crit being useful to 30% and DR thereafter. Mastery will likely take a back seat, though the average percentages of damage between Frostbolt and IL will inform that decision.

Edit: clarified FoF IL interaction, as the 25% damage bonus does not affect the first triggering IL charge. Also, all three Tier 5 bombs have the potential to proc Brain Freeze, and it is not perfectly clear which bomb will have the highest proc rate/damage output based on x number of targets. Testing ongoing.