Pintofbrew

I disagree with most your points. Firstly, not wanting to risk turning this into a PvP thread, suffice to say that Frost Armor is not clearly better than Molten. It has been proven that 5% crit mitigation is comparable to any armor benefit granted by Frost Armor, Slow on the armor is next to useless when every melee class snare is more than our armor's, and whether 10% stun is superior to 15% Frostbite is entirely a prefferential matter. Note that Frostbite has diminishing returns, does -not- save you any damage until at best 1s later (when you've gotten enough distance, while at 50% move speed), does not prevent the melee from doing anything against you (eg CloS, Vanish, Imp-Sprint, Deadly Throw, Shapechange or gaining life/rage from Second Wind) and only gives you one shatter-target for one icelance, which given in Arena we're not particularly heavy on damage dealing and damage generally only goes to specific targets we're focusing, doesn't help much. Not much point in your RMP to get a Frostbite on a rogue that's targeting you when you're focusing on the opposing Priest. However, I do not want to get into "this armor clearly PvP" because I'm not a PvPer. I merely want to display that what seems obvious to you may not necessarily be so.

Conversely, I use Frost armor in PvE when AoE is particularly tight. If some tankadins go down, Frost Armor is the first thing that goes up for me. It clearly will do nothing versus Abominations but hell will it help versus non-elites or smaller mobs.

You see the benefits of Mage Armor for PvP yet you do not see the PvE benefits? Half the duration on any DoT cast on you from a Boss looks to me like one hell of an armor... From Doomfire DoT to Mark of Kaz'rogal I don't disagree that it's great for PvP but please, don't try and confuse "great for PvP" with "should be PvP alone". The warlock change was as far as I can see clearly done for three reasons:

1) To limit drain-tanking/ lock-druid efficiency in arena by removing 24% healing from the lock, or forcing him to drop +120dmg
2) Because the HP5 on Demon Armor was utterly useless, and consequently so was Demon Armor.
3) Because it makes sense to have lock-tanks with Metamorphosis, which we know does 360% armor, combined with increased healing intake to synergize for tanking/getting healed.

Locks had one shit shield and one too good. We have three Meh ones. All that's changing is one which was most useless for all but one spec/occasion is gaining more of what it should have done all along;

It is far, far too early and frivolous to say "Arcane is clearly oriented for PvP". For BG's 2 years ago at lvl 60, it looks perfect, yes. For PvP in WotLK? I think you're making assumptions about talents we don't know, effects we don't have data on, and encounters we haven't seen. Arcane is not at all clear in how it functions or how we can produce DPS from it, and it won't be until we see how everything scales in WotLK, spells in particular.

 

Deedre

Pintofbrew, just read your remarkable post about post 3.0 pre WotLK deep frost raiding in the 2.4 TC topic. You ought to move that one over here.

Executive summary for those who didn't catch it: with 61 talent points and WotLK talents, deep frost is looking astonishingly good for dps. Lots of assumptions here, but it's all spelled out in Pint's post and the logic seems sound to me.

 

Pintofbrew

edit: Reposted from different thread. No idea why I thought 2.4 would be a better choice for this.

In the vein of Roywyns hypothetical 3.0 new talent scenario I ran a sim of my own comparing frost specs.

Two sims I ran, one with fairly optimal T6 gear and one with perfect Sunwell gear. Here's how I simmed the events.

Firstly, find WE damage from sim by subtracting it, then add 67% to it. This produced a DPS increase of 120dps for both T6 and SWP gearing. This is probably because there is little changing in terms of +damage gear. This also assumes one won't gain WE chain-cast overlap; at 3 total WE, this is impossible on a 6m fight, so it's not an issue. It will however become apparent when one attempts to have 2 WE in a short period, or after the first Cold Snap is done, how soon the 2nd one can come up. Either way it's irrelevant to the sim.

Secondly, assume Winter's Grasp only buffs the next Frostbolt, it results in "20% of your frostbolt give you 50%crit" This can quite badly be assumed as a global 10% crit buff, in the same vein that Frostbite is a 7.5% crit buff, when solo, obviously. I am aware this is a bad simulation, and it's assuming it only changes one spell rather than the next bolt-and-lance is faulty, but it's acceptable to underestimate it for this preliminary testing. I established what 10% extra crit is like in two ways: One, subtract ToW (while elemental is not specced and hitcapped regardless) and multiplied by three, and then compared that result to the DPS benefit of having Winter's Chill. Results were very coherent, with one coming up worth 150DPS for 9% crit and the other 165DPS for 10% crit. A fairly solid indication of convergence. This held for both T6 and SWP gear.

Thirdly, I assumed Chilled to the Bone would have roughly the same effect as Arctic Winds, only a little less. 98DPS/103DPS was to be had from said talent, depending on gearing.

Ultimately:

BC talents, frost spec, T6: 2088DPS
BC talents, frost spec, SWP: 2443DPS

WotLK talents, frost spec, T6: 2519DPS
WotLK talents, frost spec, SWP: 2813DPS

Notes:

1) This is assuming WG has a much smaller effect than it seems to have currently. It is hard to model and inaccurate to "botch" using icelance, or indeed the talent affecting the next two frostbolts, as it's likely to do. This is doubly true of the SWP gearing, as "optimal" in this case means 370haste and a skull. This would likely result in 3 frostbolts and one lance per WG, or one frostbolt and four lances per WG as it is now. A different sim needs to be written to establish exactly how much can be had from WG.

2) SP, ToW, WoA, Mana Spring all on. I am not aware whether Rawr takes into account any of those for the WE. Note that ToW is a massive benefit to the little fucker, as he's deadly low on hit% and that WoA is gained 1.3* over because he also gains 30% of your own benefit from it.

3) I am aware that in the previous Rawrs WE would not be stacked with Trinks/CDs/destropots.

4) This spec can be specced at 70. Unlike 33/38/0 which I'm sure we'd all love to try out asap, full frost can be specced tomorrow morning. Sadly, the best one can do in terms of talent economy is 65 points (and that's at 4/5 WC and no Ice Floes) for War Pirate :: Talent tree Mage . This is clearly 5 short of either Clearcasting or Magic Absorbption. The former being an obvious economy talent the latter combining economy with some inadverted pushback protection by providing resists. At a push, one could spec out of Arctic Winds to make 10 talent points for either talent butthat'd represent 100dps worth of drop. Worth it? you decide. I'm looking forward to 10/0/61 with Mage Armor as much as I am 33/38/0 if I'll be totally honest. At 120+MotW all resist and a 3k base ice-barrier, that's all the interrupt resist I need as frost. Molten's 3% crit can bite my ass.

Conclusions:

Assuming 3.0 goes forward, and burnout is the same as 1/2 spellpower, as it seems to be, speccing Frost for the remainder of the days until WotLK is upon us seems to harbor most benefits. This however, as always, assumes the pet will indeed survive and the ever-present frost problem of interrupts will not prove to be worse than the advantage of 61frost over 0/50/11 fire -plus- the 4% hit bonus to rogues/fury-warr/enh-shamans.

This also assumes 61-arcane does not work better still; 10% more regen, 3% more crit, and possibly Netherwind Procs checking each AM pulse combined with that unknown quantity: Arcane Barrage make for exceptionally hard to sim output.

 

Lileith

The problem is that currently you cannot include impact in your pvp build (unless as I said 0/5/56 without improved cs).

In a RMP mirror match you won't focus the opposing priest , it's all about rogue1vsmage2 rogue2vsmage1 , a frostbite may allow you to cast a sheep on the enemy priest forcing it to trinket or the other mage to IB.

All those spells aren't magic effect (as they aren't dispellable) and probably won't be affected by the Mage Armor. Having half duration on dispellable effect in pve has close to 0 benefit.
Of course if Mage Armor affect all "spells" effect , the same way cloack of shadow does (poison curse ect...) It will be so overpowered that Mage Armor indeed doesn't need any more love.
But I highly doubt because It would cause a lot of issues with other's pve boss abilities (Geddon/Solarian bomb exploding faster , stun on Zul'jin p2 procing sooner(this one is dispellable though so...) or this kind of side effect) which would be very hard to balance.

I though it was the point of this post! We have so few information we are like paleontologist trying to explain the dinosaur extinction, no one is right or wrong we are just sharing ideas: if it works that way? what would be better in this case? ect.
My initial point was, now that warlocks have a clear pve and clear pvp armor, what if it was the same for us?

 

Rounced

What changes to Arcane Mind? According to that wiki site it is still giving 15% more Intellect, did I miss something that someone else found?

As for the range boost on Magic Attunement, do you really think it will increase the radius of Arcane Explosion? The wording of the talented range addition is exactly the same as Flame Throwing and we all know that Flame Throwing doesn't do anything to the radius of Flamestrike.

 

Pintofbrew

I won't get into a PvP discussion, aside from one point. Until recently, 5 or 8 points fire has honestly been prefered over 17 arcane in all but 2v2 mage, unless I understand wrongly. The point being 15 fillers arcane were not justified for i-CS unless you were against a druid. But again: Let's not get into pvp semantics.

You may be correct about undispellable effects not being affected, and it indeed may be so. If this is the case, your point of "not as good for PvE" stands.

Yes, assuming is one of the points of the post, passing decided judgement in the face of dramatically lacking evidence is what I'm criticising. Clearly arcane, which in no way at all is showing us how it's destined to play out, is -not- oriented to pvp entirely, as you state. I see talents which are much more pvp-oriented in Fiery Payback and Deep Freeze than I do in anything arcane. The only thing I could agree -looks- like a pvp talent in the tree is Incanter's Abs. and even that is more a BG gimmick and a trick to abuse incoming raid damage via the new shatter armor to my eyes, rather than a PvP tool. You said Arcane was clearly pvp oriented, I opposed it.

I don't think Demon armor so much "became the PvP armor", as much as it became the -tanking- armor. Some times tanking shares characteristics with pvp and that's why it looks like a PvE/P distinction but as far as I can see the change was only to nerf Fel Armor and to bring a utility purpose to Demon Armor.


Typo. Meant meditation. As in change from 15% to 30% regen. Corrected on original post. Who can tell about Attunement? AE used to have 6y more than it does now provided you jumped, which I certainly did, and I hardly think it was a dramatic improvement over the current one. It's hard to establish a correlation between the two; does Flame Throwing increase the max distance flamestrike can be cast perhaps? if so we may be in luck. Even if it doesn't, comparing a targetted aoe with a PBAoE may not necessarily make for identical results. Either way, I won't sniff at 6y range extra on poly, it'll certainly make trash-pull poly more comfortable and easier.

 

Roywyn

Stats for a deep fire spec were:
692 int, 290 spi, 33.69% crit, 13.67% hit, 31.48% haste, 1603 fire damage (1607 for frost with Soulfrost).
Not that hit/cri are generic, without school specific talents.

Those numbers are fully raid buffed including totems and consumables.
241 mp5 from mana spring/BoW/shadow priest.
Add Sliver 320 dmg/15s/90s and Skull 175 haste/20s/120s procs.

Level 73 mob, 360s fight, 300 (150) mp5 shadow priest.
50 ms casting delay.
1 Heroism, no drums (not sure how ratations are handled, etc.). Using DPots/FCaps if they are better than mana pots/gems.

The Blue Bar and you - the complete Fire Mage 2.4 mana compendium: http://elitistjerks.com/658230-post3191.html

DPS spec and class comparison in Naxxramas gear: http://code.google.com/p/simulationc...i/SampleOutput
This assumes Rune of Razorice and Slow for Torment unless noted as "No TtW".
More detailed at http://elitistjerks.com/923110-post2775.html

And [Timbal's Focusing Crystal] doesn't proc on AM.

 

Lileith

I ran a simulation which doesn't try to estimate an average uptime, I will try to explain how it works:

I use an array to include all datas and I refresh the values each loop:

For 3 mages It looks like

# Mage1 Mage2 Mage3

sp [1500 1500 1500]
spellpower

fbct [2.2 2.2 2.2]
frostbolt actual cast time, can be lowered mid fight using skull IV , bloodlust

ilct [1.3 1.3 1.3]
icelance global cooldown , can be lowered the same way

cooldown [2.2 2 0.1]
Each loop this is lowered by 0.1 , when it reachs 0 the mage(i) decides what to do next frostbolt or IL, If the mage had been casting a frostbolt during the previous time (see "iscasting") the damages are calculated now.
The starting values of the "cooldowns" are random so the different mages aren't synchronised. If the mage(i) decides to cast il/fb the cooldown is set to fbct(i) or ilct(i)

iscasting [1 1 0]
In this exemple Mage1 and Mage2 are casting a frostbolt , when cooldown(1) cooldown(2) will reach 0 ,I will add the frostbolt damage to their total damage (damages of course depend If wg is up or not).

damage [x y z]


I use a global variable "wg" and "wgduration" which are refreshed every loop (-0.1 on duration , wg=0 if duration reach 0) and will be changed in case of proc (any damage done I rand a number to check if WG procs or not)

I also add the use of Icy veins and skull of guldan whenever they can, and have an option for bloodlust.

Concerning the choice between Icelance or Frostbolt if follows the rules:

If (wgduration < fbct) & (wg=1) (if wg will fade before the casting time of the frostbolt ends)
then cast Icelance
else cast Frostbolt

The pet dps is an offset value (I could take into account the 4% hit from WG,and bloodlust but It would be a lot more complicated)

Using the best sunwell stuff available + 1 elemental shaman It gave me those numbers:
(NOTE: this simulation doesn't take mana into account at all)

1 Mage 2812,8 DPS
2 Mages 3031,0 DPS
3 Mages 3157,0 DPS
4 Mages 3255,1 DPS
5 Mages 3304,4 DPS

It seems we have the same numbers for 1 Mage only, and shows that the synergy between multiple frost mages (in the case the debuff affect the whole raid of course) is insane.

 

Roywyn

Can your programme extract the uptime of Winter's Grasp (just as% duration of the fight) in the different cases of 1-5 mages?

The Blue Bar and you - the complete Fire Mage 2.4 mana compendium: http://elitistjerks.com/658230-post3191.html

DPS spec and class comparison in Naxxramas gear: http://code.google.com/p/simulationc...i/SampleOutput
This assumes Rune of Razorice and Slow for Torment unless noted as "No TtW".
More detailed at http://elitistjerks.com/923110-post2775.html

And [Timbal's Focusing Crystal] doesn't proc on AM.

 

Lhivera

There's also the matter of the attack speed reduction, however -- essentially -25% haste. I'm pretty sure that's worth a great deal more than -5% crit.

Regardless, Blizzard's view of Mage armor buffs is clearly that they should be multipurpose:

- Ice Armor: high mitigation vs. melee, escapes
- Mage Armor: high mitigation vs. casters, PvE endurance
- Molten Armor: low mitigation vs. all, escapes, PvE DPS

...frankly, if they'd just add a +Frost damage when Ice Armor is active feature to Frozen Core, each school would have its very own appropriate armor for both PvE and PvP.

 

Lhivera

Do you have information saying that Imp. WE is +10/20/30 secs, and not +5/10/15 secs? I thought the latest information was pointing toward the latter.

What I've been doing is calculating a rotation:

1) WG allows one Shatter Combo: 8x Frostbolt at X crit, 1x Frostbolt at X+50 crit, 1x Ice Lance at X+45 crit
2) WG allows two Frostbolts, 1 Ice Lance: 7x Frostbolt at X crit, 2x Frostbolt at X+50 crit, 1x Ice Lance at X+45 crit

 

Lileith

Those numbers are accurate with +/-1% due to the random number used

1 Mage : 37.5% uptime
2 Mages : 63.6% uptime
3 Mages : 78.2% uptime
4 Mages : 86.3% uptime
5 Mages : 91.7% uptime

 

Roywyn

Hm, which version of Winter's Grace did you use?
Your numbers look like the 20% proc/4 second one to me.

There was a link with a 10% proc/5 second WG debuff from the talent tree in game.
And I think that was was also corrected on war-tools.com before the trees were taken down.
Any chance you could redo the simulation with those changes? Sorry for asking a lot ...

http://img530.imageshack.us/img530/3691/frostfn1.png
http://img164.imageshack.us/img164/7783/requestzb9.png
http://img530.imageshack.us/img530/5...earcaneda8.png

Some screenies that indicate/suggest a +5/10/15s uptime for WE and a 10%/5sec WG.
I don't know if those are more accurate than the others though.


Thanks for the awesome work, Lileith!

The Blue Bar and you - the complete Fire Mage 2.4 mana compendium: http://elitistjerks.com/658230-post3191.html

DPS spec and class comparison in Naxxramas gear: http://code.google.com/p/simulationc...i/SampleOutput
This assumes Rune of Razorice and Slow for Torment unless noted as "No TtW".
More detailed at http://elitistjerks.com/923110-post2775.html

And [Timbal's Focusing Crystal] doesn't proc on AM.

 

Lileith

Yeah I used the 20%/4sec one.

With 10%/5sec the numbers are:

1 Mage 2725 DPS 25,7% uptime
2 Mages 2882 DPS 45% uptime
3 Mages 3022 DPS 56% uptime
4 Mages 3091 DPS 68.5% uptime
5 Mages 3153 DPS 75% uptime

Which is of course a lot lower than with the 20%/4sec version. Actually we could even consider WG "balanced" :p

Tomorrow I'll add the new Improved Moonkin aura in my simulation to check If it'll be better than an elemental shaman.

 

◊ Vontre

I already wrote a simulation method several days ago for the new ice mage. It includes a number of ice lances against frozen targets based on proc uptime as well. Here are some results with Manly's gear (edited to be spec-neutral):

Frost: 2123 dps (+6513 mana)
WGR: 16.03% uptime

Arcane 48/0/13: 1931 dps (-643 mana)

Fire/Cold Snap: 2146 dps (-4963 mana)

Arcane/Fire: 2175 dps (+7741 mana)

I don't have it open right now though, since I don't want to tempt fate (or Blizzard, what have you).

www.magegraf.com
www.magegraf.com/deathknight

Raiding is full of challenge. Sometimes there is fire. You have to not be in the fire.

 

Pintofbrew

I assumed last data was 3 points for +10 sec each. For half that, subtract 60dps aprox.

In order to not have to either start manipulating rawr.mage or introduce maths which may or may not include faults with rawr's other workings, I assumed WG only affected the next FrB. This, as I said, is over-simplified to give a "worst-case scenario" of it's effect. It turned out to be around 150dps worth. Shall we say a 10%-15% buff over that to include Bolt-lance? Or perhaps 25% to include bolt-bolt-lance, because the latest gearset (SWP) also included a whopping 370 haste.

Bottom line, if we get WotLK talents I'm seeing it exceptionally hard to not spec frost.

As for the mage-stacking concept, we already assumed a few pages back that it's rather rediculous if it goes through as is, and as such I assumed that one's WG will only affect self with respect to the shatter. Note, this says nothing about the uptime of the hit chance. I also assumed it'd be only melee hit, as I believe it's inconsistent with the notion of spellhit, and in all honesty rather idiotic, as practically the only things that benefit from spellhit are pets. Anyone who'se raiding and not capped is quite patently mad.

 

Lileith

I'm not sure it would be that rediculous, anyone tried to estimate the dps a Warlock would get with the exact same stats and with the new wotlk destruction talent (+10% damage on crit) and not having to refresh Immolation? I doubt it'll be that far from 3 stacked frost mage dps.

 

Rounced

That's actually what they should do to make Burnout worth speccing. Just add "any fire spell critical has a 20/40/60/80/100% chance to make the target vulnerable to fire damage" at the end of the talent and that should make it worth speccing. That way a crit with Scorch (with 3/3 Improved Scorch) would give 2 stacks of Fire Vulnerability instead of just one and any crit Fireball during spam would refresh the stack as well. Would also be nice in PvP since then the debuff would occur along with ignite and make it a little harder for ignite to simply be clensed off a target.

 

Pintofbrew