So far, Arcane Missiles seems like the only exception here, and I imagine it's because a number of us complained that it felt weird in the rotation when everything else was getting faster and AM wasn't. It really feels a whole lot better with a reduced cast time than with the extra ticks thing.

Everything else I've tried on my Warlock and Mage has gained extra ticks once it had enough haste to do so, including Evocation. This has the somewhat odd effect of making Haste sort of a part-time efficiency stat.

I'm going to go through some of the other items you raised one-by-one and provide further explanation. Sorry for the multi-quoting, but it seems the clearest way to answer this.


This is correct. Ice Lance has been totally redesigned:

- Instead of 1/3 of an instant-cast coefficient, it now has about 73.5% of an instant-cast coefficient. (Maybe 75% with some rounding errors, but on my initial tests it came in at exactly 75%, and now it comes in at almost exactly 73.5% with the same gear, so I think it was a deliberate reduction.) In an earlier build the base damage was roughly doubled as well, but that seems to have been reverted. Result is of course that the spell now does significantly better damage against unfrozen targets, which is presumably intended as Frost's version of the mobile DPS they're handing out to all ranged DPS specs.

- Instead of 300% damage vs. frozen targets, it now does 230% damage when Fingers of Frost is up. 230% of its new damage is a good deal more than 300% of its old damage, especially with it now being included in Empowered Frost. However, because Fingers of Frost no longer treats the target as Frozen, you can no longer count on it critting all the time unless the target actually has ice around its feet.

- They're trying to make Ice Lance do a few different things here, and it seems to be working. They want it to fit into a Shatter Combo against an actual frozen target. This works, just as Shatter Combos used to work with Fire Blast before Ice Lance existed. They're trying to make it produce decent DPS on-the-run. It does (where "decent" is a value somewhere below stationary DPS but substantially higher than current mobile DPS). And they're trying to get it into raid rotations, which requires that it be worse than Frostbolt with FoF down, and better with FoF up. The numbers seem to work out for this as well. They're killing a lot of birds with one stone here.

Yeah, they've thrown in the towel on AM for Frost and Fire as anything other than a leveling spell, if they ever intended otherwise. This is OK. Fire and Frost feel like they have plenty to do without AM.

That's essentially correct. You can think of it as an aggregating version of the Warlock spell Conflagrate. I think, but am not yet certain, that this is based entirely on DPS (tick size and frequency), so that you can squeeze extra damage out of a DOT that is about to expire in the process of aggregating it together with the others. Basically, you guarantee that the next spell you cast is going to crit, but instead of producing an Ignite, it takes that Ignite damage and rolls it up with all your other DOTs to create a new Combustion DOT, leaving the slate clear for you to restack your normal DOTs alongside it.

Burnout may be worthwhile for mobile DPS, especially if you talent Fire Blast down to 4 seconds. I dunno, it might be more of a PvP thing. Most people seem to think that using a Fireblast to encourage a Hot Streak proc is unnecessary with the very high crit rate on Scorch. I dunno, I think Fire Blast's damage looks to be high enough that it may be worthwhile to help guarantee that proc.

This, I think, is the primary intent behind improved Fire Blast and Burnout. If people get into raid content and a bunch of Fire Mages who haven't specced Burnout start complaining that they're doing poorly in movement fights, that may tell us something.

I don't know, with crit modifiers currently being broken, talks about making casters have 100% critbonus by default, and flashburn essentially copying the ignite functionality I Just don't see it staying.

I'm a huge fan of naming the mastery bonus ignite over flashburn if it indeed is going away. For nostalgias sake.

Flashburn doesn't really supplant Ignite. Flashburn is a flat percentage increase to all your direct damage; Ignite is a critical strike damage bonus increase. Without Ignite, the only crit damage increase in the Fire tree would be Hot Streak. We'd have the same problem with disparity in crit damage that we've had in Wrath, except now it'd be Frost on top and Fire lagging well behind. I'm dubious at this point that it's going to be removed.

I may have missed this, but does ABr reset the AB debuff countdown? If so, then yes, a 3 sec CD ABr really would be quite troublesome in PvP.
However, if ABr doesn't reset the buff, making it so that while it doesn't clear the stack, you would still have to cast AB to maintain the stack every 6 seconds or it falls off, it would be perfectly fine, imho, to have ABr remain a 3 second cd.
All that the mage would be gaining with this setup is the ability to fire one extra ABr at the same stack size before the stack expires, which I don't think is a large enough issue to really warrant an increase in ABr's CD and forfeit a large number of possible PvE rotations and cycles (especially the low MPS high DPM ones).

Indeed. I am not really worried about the new ABr not allowing us to cast what today is considered the "standard" rotation, i.e. ABx4 ABr MBAM.
Instead, I am curious as to why this design path for ABr was chosen, one which is in direct conflict with the hypothesis put forward by Seonid on the previous page (one which I agree with and find to be interesting).

The hypothesis being (if I may take some liberty in attempting to formalize it):

"Part of the mana management gameplay for Arcane will be born through the ability to manipulate the AB stack to achieve a variety of DPM, DPS, MPS cycles, that range from not casting AB at all (i.e. a ABr->MBAM cycle), to increasing the number of AB casts to 4+. This gameplay hinges on the ability to reliably cast the MBAM at any point of choice in the cycle, through ABr (since MBAM is the only way to 'end the cycle')."

A key concept of this hypothesis is; the total number of possible reliable cycles available to arcane are numerous and varied enough, so as to realize the 'control' aspect of the mana management game.
I believe the change to ABr directly conflicts with this key concept, since it effectively reduces the possible number of reliable cycles to ABx3/4 ABr MBAM (or perhaps ABx2/3/4 ABr MBAM with reduced amounts of haste).
Essentially, some low MPS high DPM cycles are forfeit with this change, since you will not reliably be able to end the cycle at an AB0 or AB1 stack.

N.B. just a quick note on "reliability". AB0 MBAM is still a possible cycle, since MBAM can proc off of AB, however, I would not defined it as a reliable cycle.
I would like to define "Reliable cycles" as all cycles that do not need a proc to execute, i.e. all cycles tied to the use of ABr. The importance of reliable cycles comes from the assumption that a core tenant of the mana management gameplay of arcane will be that the player has full control of the majority of his tools, i.e. that mana management of this form is not linked to procs. Naturally, this assumption could be wrong, but I think it is a fair one to make.

Just curious, was this change to AM's haste mechanics made before or after the reduction of the base AM going from 5 to 3 volleys over 3 seconds? Also, do you have any insight as to why AM was changed to be a 3 second base channel with 3 volleys?
While not outright against a 3 second AM (or 1.5 second MBAM), I do think that it will be weird having MBAM be a 1.5 second (+haste) cast. The longer casted MBAM 'felt' right in the Arcane playstyle, kind of like a respite from the furious pace of casting fast ABs. Like a 'breather' almost. I feel a 1.5 second MBAM would make playing arcane feel a little too "spammy".

Both changes were made in the same push. There's been no discussion as to why it was reduced from 5 to 3. I can say that it feels more elegant in a rotation without Missile Barrage that way, so maybe that had something to do with it. That's just speculation, though.

So far, I have a few questions about Flashburn. But basically, how much does it mimic Ignite?

Complicatedly, will a new Flashburn overwrite the previous Flashburn DoT, act as a separate DoT, or add the remainder of the previous Flashburn DoT to it? I tried to go through and did not find an answer already posted in this thread.

The worst thing to happen for Flashburn is it to overwrite the previous DoT. The next worst thing is having the Ignite munching effect.

I imagine all of the DoTs the fire mage has are longer than before. Otherwise for moving situations, it can be hard to reliably keep at least 3 DoTs up as often as possible to take advantage of Critical Mass and such, thus fire DPS would be significantly lower in such battles.

This behavior may be it actually behaving exactly as ignite does, you just have significantly more opportunities to observe it now if every non periodic fire damge event can trigger it.


With fire becoming more focused on DoTs tracking when to target switch will potentially become much more important. To the point where it might definitely be worth it DPS wise on a fight to cast a non fire spell if there are no other targets available to allow DoTs to run without pushing them back (them being Ignite and Flashburn, that is).

As far as I can tell, Combustion isn't working yet. I get the guaranteed crit, and it seems like some DOTs are wiped from the target, but not all of them, and no new DOT is created.

What's the cooldown if I may ask? The wording used in the tooltip suggests - to me - that it should be like 1 min tops.

Fireball and Frostfire Bolt DoT Questions
According to tooltips, they scale, with 0.6s scaling over all ticks totaled. I.e. Fireball should get 0.15s (~4.5%) every second 4 times, Frostfire Bolt should get 0.2s (~6%) every 3 second 3 times. Can anyone confirm that th DoT scales (no excessive testing needed, just qualitative)? According to some reports, the ticks could crit and proc Ignite at some point. Do they still do that?
I reckon DoTs just get overwritten when spamming these 3 spells? And is there an information on haste interacting with the DoTs, i.e. making the DoTs tick faster (which would at least reduce clipping issues)?
Also, could somebody test/confirm Pyroblast scaling and the DoT behaviour?


Thoughts on Fire specs
Scorch is a 1.5s spell with 1.5s scaling and +39% crit from talents. Fireball is a 2.5s cast spell with 2.4s scaling and +15% Empowered, +9% Critical Mass, +6% Spell Impact and and +6% Torment bonus. Plus you get 1 of the 4 Fireball DoT ticks.

Critwise, we're looking at 5% base/intellect from gear, 3%+3% paired Focus Magic, 3% Molten Armour, 5% Wintershadowscorch, 5% Moonrampagepack, 8% Fire Mastery. That's 32% without any crit rating whatsoever, and's 71% for Scorch.

The timer-based Hot Streak is a tad ugly to model (as Markov-chain or whatever) due to how a previous Hot Streak proc interacts with your time frame. And when your time frames change at haste breakpoints, but that happens rarely with the current cast times. I'll give an example, assuming that we can fit 2 Fireball casts in 4 seconds (so 3 spell hits in a 4s frame):

FB0 (crit, procs HS) - FB1 (chaincast) - Pyro0 - FB2 - FB3 - FB4 - FB5 - FB6 - ...
Due to the way we chaincast, we'll cast FB1 before using up the previous HS proc with Pyro0. So, we get our next Hot Streak proc either after when FB1 and FB2 crit or after FB(n) when 2 of the 3 spells FB(n-2), FB(n-1), FB(n) crit. FB1 and FB3 critting doesn't trigger Hot Streak since the time between them is 6.5s/haste due to Pyro0, usually > 4s. The same situation for Scorch spam, where you need 2 crits in 4 casts in general, or 2 crits in 3 casts right after a proc, which will be the most common situation due to crit rates.

Besides Fireball spam and Scorch spam, there is a third option - cast Scorch after a seeing Fireball crit. In a sequence, FB1 (crit) - FB2 (chaincast) - Sc3 (reaction on crit) - Py4 (?) / FB4. This brings up another set of questions.
If FB2 crits, do we get the proc early enough during the cast of Sc3 that we can cast Py4? If we don't and the first 3 spells crit and our chaincast FB4, can we cast the first instant Pyro before the FB4 crit triggers the HS proc, or will we lose one HS proc to overlaps? I don't even dare building the time frames for the following possible HS proc.

This seems to lead to vast possibilities of correct and incorrect timing options. The old Hot Streak was candy for TCers compared with this.

Also, I don't think you can balance this without a cooldown on Scorch. If the balance is even just a bit off, then forcing procs with Scorch is either worse than Fireball spam, or it's better but Scorch spam trumps it even more. They might balance this with DoT ticking, but they always ride an extremely fine line then. Probably too fine.


[Edit]: Vontre, a fire mage will have only 5% spell haste outside gear in single target fights. So they'll spend a large amount of time not haste-capped. But they are GCD-locked, which is likely bad for performance. Still the balancing issues above remain. With a 75% crit Scorch in pre-raid gear.

Lhivera, thanks for the correction and information.

[Edit]: AoE Scaling Penalty
Is it still -50% like it WolLK or down to -66% like in Vanilla/BC? I'm asking since most melee AoE got taken down a notch or two and they want AoE to be less attractive againt few targets.