Those results are indeed startling. So: there's really not much point at all in working up a full debuff stack inside of a regular rotation. You either spam AB or not, no middle ground.

Edit: on second thought, even AB spam is looking dubious. It's heavily outdamaged by both barrage and AM on missile procs.

Sorry about mis-reading your crit numbers before, I should have seen that as soon as I looked at it really.

my napkin math is showing *slightly* better numbers for an ABar/Fireball rotation with 2-minute AP/PoM/Pyro/trinket than a purer arcane ABar/ABla cycle. Of course as you stated you havn't TC'd stacking cooldowns yet so it may still work out in favour of arc/frost or pure arc.

Edit: although my money is still on good ol' fashioned PoM/Pyro over IV.

The latest changes look interesting, we can debate of how gimmicky the fire talents are, but i think it is a step in the right direction. Actually the only issue with the fire tree at this point seems to only be raid/single target DPS.

Blizzard could easily just touch 1 or 2 talent in there to make it viable. I could see boosting burnout to make it a 50% increase instead of 25% something that could make deep fire more acceptable, obviously it would need a bit of help somewhere else but i think that could be achieved with tunning.

Seems we are not that far to have a viable fire tree. the only change that i m worried about is the knock back on blast wave there is a lot of fights were we HAVE to gather adds before AOE. From what i have seen in the video though it seems the knock back is very small so not really intended to remove us from mele range ( Any beta tester can confirm this ? ). So i was wondering woudl this be intended to be more of an AOE interrupt ? Does the knock back interrupt casting ? I could see the use there.

OK, corrected yet again with the right Spell Impact effect. New results:

DPS
2695.07: Deep Arcane w/IV and Incineration, Barrage/3xBlast/AM on proc
2689.14: Deep Arcane w/IV and Incineration, Barrage/Blast/AM on proc
2647.20: Deep Frost w/50% WE uptime

DPM
15.52: Deep Frost w/50% WE uptime (Note, does not account for WE cast as we have no level 80 value)
14.70: Deep Arcane w/IV and Incineration, Barrage/3xBlast/AM on proc
14.46: Deep Arcane w/IV and Incineration, Barrage/Blast/AM on proc

For comparison's sake Lhivera, can you include a 33/38 Fire build and a Frostfire build (no idea what the optimal setup is for that).

Hey guys, I've been looking at the new regen mechanic and is the only change that it's continuous instead of segmented? Anyone know?




EDIT:

Oh and lhiv, could you please run the numbers for deep arcane (54/3/14) using ABar/Frostbolt/MBAM as a rotation, for comparison's sake to ABar/AB/MBAM?

its simply changed to refresh every 1/10 seconds rather than every 2 seconds, making it feel more fluid. the actual amount being restored is unchanged.

As pointed out before, I don't think 50% on Burnout will be near enough to rescue deep fire. In fact, two changes in the recent talents make me believe Blizzard actually wants 33/38/0 to be the fire raiding build. Instead of making the deep talents desirable, they changed Empowered Fireball to 3 (from 5) talent points and Arcane Potency to 2 (from 3). Incidentally, these were the two "interesting" talents that couldn't be maxed in the 33/38/0 before the patch.

I find this a very curious decision, as it eludes me what the purpose of deep fire is with these changes (I don't really see anything in deep fire that screams PvP). Anyway, if deep arcane with ABar will be the new raiding build, that's fine with me. Would make for a nice change in scenery.

Braal

Lhivera, just to clarify, your top DPS rotation is ABar>AB>AB>AB>ABar etc correct? With AM on proc? Just trying to get it sorted in my head XD

the nice thing about this rotation is that once you reach 33% total haste (very easy with tallents+WoA), you can change to AB/ABar without losing the debuff. Ok, it becomes a mana whore and we'll need to see how the DPM plays out to see if it works out in the end, but its deffinately a spec to keep an eye on even if it has horrible, horrible synergy with moonkin procs.

Think about how the debuffs/cooldowns work.
The ABar cooldowns means that the time between your AB hits is at least 3 seconds.
The AB debuff timer means that the time between your AB hits must be at most 3 seconds.

That means that you have no leeway at all. You have to cast them exactly every 3 seconds. 1ms lag will throw you out of your rotation.
If the buffs/debuff however have some lag/delay, then you might get a small time window to pull it off.

What you can do however is get a little bit more than 33% haste and go ABar-AB-AB.
You're not limited by the ABar cooldown that way, but burn more mana of course.

You can keep up 3-stacked Blasts at >33% haste with ABar-AB-AB. But you asked about Missiles.

2.5s Missiles are a ~33% DPS boost for 2.5 seconds for 3-stacked Blast.
After that, you have to cast a 0/1/2 stacked Blast, that's equal to three 1-stacked Blasts for 7.5 seconds.
1-stacked Blast is 1.15/1.45 ~= 80% of a 3-stacked Blast.

So you gain 33% for 2.5s, but lose 20% for 7.5s. Not a good deal for pure DPS. It saves mana though.
Since you have the debuff drop, you can do a ABar-AM-ABar, which is slightly more damage and saves you some more mana.

In order to actually get a DPS benefit from Missiles, you'd need AB-AM (5s unhasted) to cast within 3 seconds, which requires >66% haste.
At that point, you're GCD capped and Barrage will be similar or slightly worse worse DPS than Blast, it becomes worse with more haste.
So you can spam AB and use AM on procs for maximum DPS, or try (ABar-AB*2) for similar DPS (check individually) and less mana cost and use AM on procs.

Summary on highest Arcane Burn DPS:
<33% haste: AB spam.
=33% haste: ABar-AB spam if it works
33-66% haste: ABar-AB-AB spam
~66% haste: ABar-AB-AB spam, AM on procs
>66% haste: AB spam, AM on procs

Blast spam at 3 stacks will likely not be sustainable though.


Wrath of Air and Ret Aura add 14% haste, I used them in my settings. That shouldn't make any impact at all though.
I used ~14% base crit, and Molten/ToW/Judgement getting it to ~22%. So about the same as you in the low setting.
I did forget the 2% damage from Ret Aura and omitted the 5% partials, but they don't change relaive values either.
Same with taking Malediction and Misery.

I also used CSD + EO, which have a noticable impact on crit modifiers and crit-heavy specs.


Changing my gear (and crit buff) setup to your numbers (keeping CSD/EO), I get:
3.5k for Barrage cycles with Blast fillers, 3.2k for Arcane/Fire and FFB specs.
Frosts is 3.2k without frozen rune weapon, and 3.5k without it.
Now that would look very much like your balance if I removed the crit scaling CSD/EO.

When scaling up +spells and particularly +crit in the model, I see frost slowly sink back and arc/fire and FFB catch up.
That's why the numbers look so much different at ~30% crit and 3.2k spell power before talents.

Deep Fire with 50% on Burnout would land roughly halfway between current Deep Fire and Arcane/Fire.
It doesn't change enough when Arc/Fire can get the 15% Empowered Fireball now.

Fire simply lacks the ability to convert talent points into DPS right now.
Even if Burnout and Hot Streak were improved, you'd have ~50 point is fire.
And the other ~21 points?
Precision, Icy Veins, Cold Snap, Focus Magic are nice. But they are pretty poor returns for ~21 points.


50% Burnout would push 0/50/21 FFB specs that leech Winter's Chill about 5% above Barrage cycles.
That's attractive enough for the 2nd mage in the raid who doesn't mind respeccing when the WC mage is not there.

FrB is a slightly better filler than AB in that case, but requires Imp. Scorch from another mage. It's in my post.
It does however require some amount of crit. At lower gear, AB will better.

Quick group reply here:

Regarding requests to run numbers, I do plan to look at other specs, but I have not coded any of this stuff yet; I'm doing it by hand. It's going to have to come as I have the time to do it. I'd also like to give some of the other theorycrafters a chance to look over my numbers and suggest changes to assumptions or methods if they thing I'm way off on something.

Arazan: "MBAM"? I know what you mean, but it's still funny. It's actually pretty appropriate, kind of carries the implication of boomage.

Darkstar: Yes, that's correct, although it's not like I've compared a lot of different rotations, those were just two that people had been talking about so they're the ones I looked at.


I'm right there with you. It is literally impossible to save Fire with only changes to Burnout, without making it ridiculously potent. Fire DPS increases need to come from multiple talents at this point -- and frankly, what it really needs now is a damage% increasing debuff coming from some other class. Whereas previously, Fire was the only Mage school with a school-specific debuff multiplier (Winter's Chill merely compensated for all the crit increases Fire got on the casting end), now Fire is the only Mage school without one.

If the actual intent is to make Fire be the unstoppable king of AOE damage, then there are two possibilities:

1) It's worthless.

2) AOE is so utterly prevalent in level 80 encounter design that you actually want a Fire Mage specifically for that purpose...at which point it perhaps becomes mandatory, which is as bad as being worthless.

I'm too lazy to constantly type "and arcane missiles off of missile barrage procs". "MBAM" just sounds so much better. I hope it catches on


Could you link to the post where you listed your assumptions? I can just g'head and do the math myself but I'd like to start with the numbers you used for continuity's sake.

Aha, christ I feel bad for asking about builds now XD. And yeah, thanks for the clarification.

Correct, i'm a prat

Edit - MBAM is awesome

Hello;

While following the discussion about the recent talent changes, I just had two thoughts about the fire tree which I wanted to post before I forget them.

At first, "Firestarter" appears to be really weak and confusing as other posters already mentioned. Wouldn't it probably a good change to make it proc off Flamestrike, too? When spamming Flamestrike, about every second Flamestrike would be instant and thus only take 1,5 sec in the spell rotation rather than 3. This would also mean improving Flamestrike without running against the AoE-caps.
If Fire is going to be the AoE-tree, then Flamestrike, as the trees only ranged AoE-spell, does need some improvements even behind my suggestion to be effective. I guess that would be a step in the right direction (even though I still hope Fire becomes single-target-DPS-competetive).

The second thing, Blastwave now having a knockback: Why not simply make NPCs immune to that effect? I don't think this would be a serious nerf for single-performing, but this would disable the strange effect of shattering AoE-pulls, which is usually the last thing you want to do after you managed to get them all to one spot.
Any NPC that would be hard to kill alone would probably be immune to the knockback anyways. Also, even though I play since release, I can't remember seeing a NPC falling down somewhere. Do NPCs take falling damage? Could this be exploited with knockback mechanics? If so, the combination of Blastwave and a cliff could kill anything that is not knockback-immune.


Sorry if I broke into another disussion, but this thread is growing faster than I could react :>

Arcane Missiles was blowing one charge per tick (5 per cast) when I tested it this morning.

Long time reader, first time poster...

I'm surprised that those two are so close -- I thought weaving more ABs in would be a good way to trade mana for additional DPS, shows how much eyeballing things is worth (nothing). Arcane Barrage effectively becomes a way to let the Arcane Blast debuff drop to save mana while not cutting into your DPS.

What I'm asking myself is, is it preferable to MBAM as soon as it procs, rather than finish casting the three blasts before doing it? MBAM's DPS is considerably higher than Arcane Blast, at any debuff level, so it may be better to cast it as soon as it procs and then build up the debuff again hoping to get another proc ASAP -- this also increases the odds of casting a high-DPS Arcane Barrage earlier. Your excellent writeup gives me all I need to know to do the math, I'll throw the numbers in my spreadsheet when I get the chance and figure it out.

Another thing I'm wondering about MBAM is whether it may be better to cast Barrage-Missiles-Barrage-Blast when it procs, rather than just Missiles-Barrage-Blast.

This is also quite surprising. Deep Arcane with reasonably comparable mana efficiency to Frost? That's siginificant, especially since "% of base mana" spells cost a smaller portion of an Arcane Mage's mana bar thanks to their larger mana pools from Arcane Mind.

Was it mentioned if that was tallented for emp. AM?

if not, then 686-663 = 23,

23/80 = 0.2875

or 28.75% CEPS,

143.75 co-efficient per cast, which I beleive is the normal amount for base AM. So, AM scaling may well be unchanged. Furthermore, that would imply that FM is treated as additional base spell damage.

Oh Lhiv nevermind about listing your assumptions. Found the post... I was trying to look further back than that. Will do the math for ABar/Frostbolt/MBAM rotations. Be back with that in a bit.


EDIT:

Why did you not assume Earth and Moon moonkin debuff? That's a +10% arcane damage damage buff which makes it considerably nicer in comparison to frost... you counting on moonkins continuing to be basically non-viable?

The 3 second AB stack is either terrible design, or telling us not to bother with AB except as "execute" spam. While the passive haste to weave in instants will be reasonably achievable with gear and 25 raid buffs, it leaves little room latency and any MB proc that is used will drop the AB stack. The AB debuff should be increased to 5 seconds to allow better use of the spec in 5 man instances, heroics, 10 mans (where additional passive haste buffs may not be as available), and entry level raiding where there isn't the haste to support the rotation.

The last thing I want to see for mages is the affliction->destruction jump that warlocks had in TBC. I'd rather see a gradual rise in DPS with gear as I saw with both fire and frost in TBC, than play something suboptimal until just before endgame.

I don't think he listed a 3 stack, I will using his numbers.

listed:

Arcane Blast (1-stack)
Avg dmg: (985 + 1609) * 1.15 * 0.99 * 1.2475 * 1.395371463486 = 5140.83
Avg mana: round(586 * 0.97) * 0.9 - 2.1897 * 75.88 = 345.05
DPS: 5140.83 / 2.1897 = 2347.73
DPM: 5140.83 / 345.05 = 14.90

Arcane Blast (2-stack)
Avg dmg: (985 + 1609) * 1.3 * 0.99 * 1.2475 * 1.395371463486 = 5811.38
Avg mana: round(921 * 0.97) * 0.9 - 2.1897 * 75.88 = 637.55
DPS: 5811.38 / 2.1897 = 2653.96
DPM: 5811.38 / 637.55 = 9.12


Continuation !!!!!!!!!!!!!!!!!!!!!!!!!

Arcane Blast (3-stack)
Avg dmg: (985 + 1609) * 1.45 * 0.99 * 1.2475 * 1.395371463486 = 6481.92
DPS: 6481.92 / 2.1897 = 2960.19

Still lower than AB and AM on procs like you say, but more dps than frost nonetheless. All they need to do to fix things is add 1-2 seconds to the proc on AB, depending on how much haste they think you should have to be able to alternate the spells.

That looks about right. I used 1.5*1.09, thinking that EO/CSD are additive.
The difference is too small to be measurable or matter.

Yeah, those favour Frost over Arcane, making FrB a better 2.5s filler than AB once you have high crit.

But they also slightly favour your Fireball fillers in Barrage cycles (which do half the damage) over your Frostbolt in deep frost. And they also favour Arc/Fire and Frostfire over deep Frost.

You're right though, those modifier don't seem to have a massive impact on difference.
It's mainly the scaling with crit and damage that has the major impact here.