Pintofbrew

It depends on the gear, it depends on the talent spec, it depends on the raid setup and it depends on the debuffs on the mob.

AB may have 13% from CoS while in certain guilds fire may not have even 10% from CoE.

AB gains nothing from Bloodlust, while Fireball gains a helluvalot when in molten fury range

Fball has molten fury, which combined with flamecap/destropot/trink/combustion can eke out massive burst, while AB only has AP.

AB does not have any interupt resist.

Molten Fury still affects AB, though in order to have MF you need to forego deep arcane talents.

It's not easy to model these circumstances. Asking "Does AB out-dps" is a question that can't be answered in one line. Which car is faster? It depends on the tires, race-track, conditions, driver... Horsepower won't give you aclear answer

 

• Arawethion

On the relative value of stats (the math here is borrowed from a thread I made a little while back, but not too complicated. Just replicated so you know what I'm talking about).

-------------------

The following are fixed parameters:
A: Base damage of a spell, with 0 +dmg.
B: Damage coefficient.
C: Critical bonus.

The following are variables:
d: +dmg
h: chance to hit, i.e. 1-(chance to miss)
c: chance to crit
s: haste

Expectation damage per unit time is:
E = (A+Bd)*(h+Cc)*(1+s)

Taking partial derivatives:
dE/dd = B*(h+Cc)*(1+s)
dE/dh = (A+Bd)*(1+s)
dE/dc = (A+Bd)*C*(1+s)
dE/ds = (A+Bd)*(h+Cc)

-------------------

Working the numbers for my current build/character (armory in my profile):

+1100 to Frost (totem and raid buffs), 0 haste, capped hit, and 20 crit on the char sheet.

A = 620
B = 0.914 (damage tax has been repealed)
C = 1.09 (Chaotic Skyfire Diamond)

d = 1100
h = 0.99
c = 0.35 (Emp. Frostbolt, WC)
s = 0

dE/dd = 0.914*(0.99+1.09*0.35) = 1.253
dE/dh = (620+0.914*1100) = 1625
dE/dc = (620+0.914*1100)*1.09 = 1771
dE/ds = (620+0.914*1100)*(0.99+1.09*0.35) = 2229

Factoring by itemization costs:
+all damage: 1.067/0.855 = 1.47
+Frost damage: 1.067/0.7 = 1.79
Hit rating: 1625/1262 = 1.288
Crit rating: 1771/2208 = 0.802
Haste rating: 2229/1577 = 1.413

----------------------

So, point-for-point, +dmg and hit are nearly equal, countering the notion that HR is far better than all other stats. When itemization costs are taken into account, +dmg is more efficient to gear for by a good margin.

Taking the WE into account makes +dmg better than hit even 1-for-1.

So, what math are people relying on for the generally accepted conclusion that hit, when available, is better than all other stats?

Answers to Moonkin questions:
0) Read the TTT/use the spreadsheet: http://elitistjerks.com/f47/t66856-moonkin_pve_dps/
1) Maintain high DoT uptime. Use WiseEclipse.
2) Nothing beats 2T8.
3) Yes, sometimes you cast many Wraths and no Eclipse procs. Deal with it.

 

◊ Vontre

Yes.

www.magegraf.com

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

"We agree with Communism." - Greg Street 2009

 

Copernicus

Deep fire with a higher base +dmg, I think. Also a two-roll system makes hit more valuable, and I believe work has been done to prove that system is true... except I haven't seen the actual math or proof.


I'll try to work up a chart to show the various values of dmg/hit/crit/haste at different gear levels and specs (using Vontre or Lhivera's theorycraft sheets). The various values of the ratings change over time in relationship to +dmg (although not compared to themselves).

 

Muphrid

The problem is with your model.

Neglecting the need for a time (t) term of some kind, your model is a one-roll model, when accepted theory (open to dispute, but more or less accepted theory all the same) is two-roll.

That is, the rest of us are mostly using...

This makes the relevant deltas...

I use deltas instead of partials or derivatives only for generality; it is the case that the ∆E's are only first-order for these stats, but not for all stats in all situations.

To greatly simplify how to determine which stat is point-for-point better, it's useful to look at ∆E/E.

This eliminates the need for considering the other stats in a pairwise comparison.

Thus, for your stats, we get...

With a simple correction to +damage for item costs:

So we can see that 1 item point in Frost damage is better allocated than 1 item point in hit...for you. However, dabbling in benefit per item point is only of interest towards balancing of stats in a general sense: in terms of weighing items, it's not terribly relevant.

At any rate, you can see that 1 +hit is better than 1 +damage (by 42%, in fact) point for point. The one-roll model made hit's value deflate as crit increased.

 

• Arawethion

Seriously? Where have I been?

Is there a thread discussing this somewhere?

Answers to Moonkin questions:
0) Read the TTT/use the spreadsheet: http://elitistjerks.com/f47/t66856-moonkin_pve_dps/
1) Maintain high DoT uptime. Use WiseEclipse.
2) Nothing beats 2T8.
3) Yes, sometimes you cast many Wraths and no Eclipse procs. Deal with it.

 

Roywyn

I have no clue where you've been, but casters most likely use a two-roll system for spells, just like physical classes use for special attacks (yellows). Normal attacks (whites) use a one-roll system. Hunter autoshots were yellow and now are white and can't glance, I have no clue how they behave.

Quick question about 1 roll or 2 roll system for casters

This hints at a two-roll and an odd crit-depression against high-level targets.

 

Kir

Curious, we're starting to work on Illidan now and I'm going to spec frost, does the ghost +hit for frost still seem exist? I'm going to keep a close eye on my resist rates in recount if I itemize for it, but wanted to know as it would make a small difference between picking up certain items (heroic badges frost offhand with +hit vs the haste offhand) beforehand.

 

JasonX

Yes, you still benefit from the ghost +hit.

 

Ediction

Wasn't sure where to post this and I don't have permission to create new threads yet...

With the pending changes to the Frost tree in 2.3.2, is the new raid spec going to be 9/41/11 with Icy Veins in place of Cold Snap? The current tooltip for Icy Veins says "Icy Veins decreases casting time of ALL spells by 20%". IV+trinkets+combustion every three minutes sounds attractive... Add in well timed BL's and it's a nice chunk of damage - It's going to look like you're casting scorches when you're bombing fireballs. If you don't need the aoe abilities I'd think it was something along these lines: Talent Calculator - World of Warcraft

If this was posted elsewhere, let me know, but I certainly hadn't seen anything related yet.

 

Dustwhisper

2/48/11

Drop clearcasting for IV.

 

Ediction

Ah, that makes sense. I'm kind of surprised IV isn't just for frost spells...

 

Pintofbrew

Since BC came out Blizz has intently tried to integrate class schools into each other, specifically to enhance synergy. It is no surprise it is continuing along this route. Other examples include:
1) Shatter changed to affect all schools
2) Master of Elements proccing off frost as well as fire
3) Play With Fire affecting all schools (Give Back Beta PwF: +2%out/+1%in!)
4) Spell Power affecting all schools' crits.

 

Ghalt

Sorry if that's the way it came across. I'm of the school of thought that trash doesn't really matter. But the whole 'scorch the tail' theory which someone else brought up didn't really seem to apply to Bosses, which confused me. I wasn't sure if someone was advocating using it just to squeeze in one more spell on trash, or if the Molten Fury combined with crit chance somehow mathematically made Scorch Spam over a long period of time better DPS than Fireball Spam. (Which would have surprised me, and since I asked has been answered.)

 

Vodorod

I'm just at the Kara level and I picked Frozen Shadoweave tailor because I'd rather not give up Girdle of Ruination this early on. Also I don't see much prospect of decent shoulders any time soon. I'd welcome informed opinions on that choice, but more specifically, if I build a 40/0/21 spec, is there more of a raiding +dmg boost from 5/5 into AF or AC? That's assuming that AF's resistance of target reduction equals a significant +hit%; it's easy to see how clearcasting helps in a long bossfight. Thanks

 

Ghalt

Wait...

...unless I'm missing something, IV would give about 2 extra Fireballs every 3 minutes.

Is that DPS increase really worth giving up Clearcasting, even with the new Mana Gems?

Or, do we do 6/44/11 which gives us a 2% Clearcasting, IV, and all the Fire Dmg that matters (losing Pyro, BlastWave, etc...things not used in most boss fights)?

 

Copernicus

From people who have done the math, the new mana gems are equivalent to clearcasting.

And for 90% of the deep fire mages out there, our mana pool isn't a large concern. I can sustain a five-minute fight without a shadow priest. With even 50% uptime on a shadow priest (rotating groups with a warlock) mana is no concern at all and I can destruction potion/flame cap.

 

andastra

It still depends on the fight. Having to use evocation versus not having to, even if it's only a tick or two, could wipe out the dps gain from IV. Of course, I'm still going for IV, since it's an on-demand burst.

 

◊ Vontre

I typically end fights in Hyjal/BT with over half a mana bar and no pots or evo used....

Your results may vary.

And yeah, spells are almost definitely 2-roll. You have to hit to crit. This makes hit absurdly good for offensive casters.

www.magegraf.com

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

"We agree with Communism." - Greg Street 2009

 

Gohei27

This post is more of a retaliation for arcane specs than anything.

I disagree with this point of view completely, I honestly believe you haven't been able to see the full scope of arcane. I'm really disappointed by this because I think people come here for useful information, and they take your opinion very seriously.

I've raided all content as arcane, i specced arcane near the end of 2.1, in fact are whole mage team switched to arcane shortly after seeing the results. To suceed as arcane you do however need to wear two pieces of tier 5, i prefer the pants and shoulders compared to tier 6. However you choose is fine. You also need a shadow priest, a very gewd one. Our usual group make-up would be 3 mages (all arcane), 1 shadow priest, +1 (shaman, hunter, warlock, healer, depends on the encounter mainly)

There are a lot of perks to being arcane, I think it gives you complete control over your dps. With fire and frost you are stuck with a set cycle, set time, set dps. Arcane does have a rotation, I use ABx3, Blast/Lance, AM, repeat. However I rarely stick to it, I actually AB until I proc clearcast, cast AM, cast Blast/Lance, repeat. Experience is the big winner with arcane, just knowing your fights, how long they are, what interuptions you'll have, etc. The idea is arcane blast as much as possible, eat pots and gems as early and frequently as possible, get those evos in. I find slow to be useful on Illidan for the demons, striders in Vashj, and once you get used to having this spell, whenever you get aggro. Primatic cloak might not seem like a viable talent, but mage tanking isn't out of the ordinary, bloodboil randomly picks his tank, stamina is a raiding stat, prismatic cloak is essentially stamina.

Arcane's biggest draw back is mana usage, you don't get to use destro pots, instead always having to use mana pots. You could use destros on Akama or RoS, but you might have to use a rotation than constantly spam AB. You need a shadow priest in order to compete and not be a gimp. Lots of mages don't like having to be with in 30 yards to attack, but this becomes your new home quickly. You have to get up front on Illidan if you want to do damage, you have to blink out of rain of fires on azlagor (it was a boring fight anyway), and if doomfires are a problem for you at 41 yards, then you might take some time adjusting to 30, but it isn't a big deal at all.

I also want to point out the 2.3 buffs to fire and frost, I honestly believe these buffs were done to make them more competive with arcane. At the same time they are trying to help arcane where they think it needs help, which is why we now have 15% more regen while casting, 60% total. On top of the new incoming gem buff, not only do you get 3 gems, but for the first time the gem gives back more than the mana cost. Meaning I can do even more damage during longer fights since recasting them is viable.

I try to end every fight oom. Here is a link to one our Illidan kills. There is also room for improvement, so it can do better than preformed.

http://armory.mmo-champion.com.nyud....37170zyBei.png

 

• frmorrison

Yah, there was a thread that this guy attacked Dr. Doom 3 levels below his level, and did it with 10% to hit and then 0% to hit, and you could see the loss in crit between the two attempts.


The Warlock spreadsheet accounts for the two roll system.

 

Zephriel

This is false. AB's debuff and associated mana cost increase is additive, not multiplicative. Without the 2-piece T5 bonus, AB costs 195*(1+(0.75(3))) ~= 633 mana fully stacked, not 195*(1.75^3). Furthermore, it's worth noting that the 2-piece bonus ignores AB debuffs when calculating the 20% increase, meaning it's a flat cost increase of 195*.2 = 39 mana per cast. AP works the same way when applying its 30% cost increase, adding only ~59 mana to each cast. AB, even with 2-piece T5 and AP active, has a maximum cost of ~731 mana per cast. That's still no small price to pay in a boss fight, but it's not nearly as unmanageable as you implied.


I, like Gohei, don't believe arcane is getting a fair shake in this thread. Arcane DPS doesn't involve simply repeating ABx3->AM->Scorch ad nauseum. Unlike deep fire and its FBx8->Scorch rotation, arcane relies on a combination of two rotations to maximize DPS:

AB (spammed)
ABx3->AM->Scorch

Deep fire's FBx8->Scorch falls in between those two arcane rotations in terms of DPS. Plugging my gear and stats (including my 2-piece T5) into Lhivera's theorycrafting calculator, and trading hit/crit as needed to achieve the spell hit caps for both deep arcane (with mage armor) and deep fire (with molten armor), I get the following results:

2621.26 DPS -- (deep arcane) AB spam w/AP, 2-piece T5
2016.35 DPS -- (deep arcane) AB spam, 2-piece T5
1484.33 DPS -- (deep fire) FBx8->Scorch, Molten fury averaged
1376.62 DPS -- (deep arcane) ABx3->AM->Scorch, AP averaged, 2-piece T5

That was expected. Deep fire's basic rotation outperforms deep arcane's by a little over 9%. But it also implies that if AB spam can be sustained for a certain percentage of the fight, deep arcane will overtake deep fire.

Let's stick with the worst-case scenario for deep arcane AB spam, and not include AP.

1376.62(1-x) + 2016.35x = 1484.33
639.73x = 107.71
x ~= 0.168

So, at worst, I'd need to sustain AB spam for 16.8% of my total DPS time. That's not far-fetched, especially in shorter boss fights or fights with plenty of casting interruptions.

To be fair, let's add a 5% damage bonus to Fireball (4-piece T6 bonus) and recalculate.

1554.93 DPS -- (deep fire) FBx8->Scorch, Molten fury averaged, 4-piece T6

1376.62(1-x) + 2016.35x = 1554.93
639.73x = 178.31
x = ~=0.279

Now AB spam would need to be applied for 27.9% of the total DPS time for deep arcane to come out on top.

The point I'm trying to make is this: While deep arcane doesn't match deep fire at the very top level of gear, it's a legitimate (arguably still superior) DPS choice for mages at the T5 and early Hyjal/BT level of raid progression and gear, if its rotations are used properly.

 

◊ Vontre

I've been completely fair to arcane specs in my posting. I listed its obvious advantages, and, if in your judgement those advantages make it a worthwhile spec I'm not here to tell you to change your mind. However, this guide is intended to be a quick reference for mages looking for general info and I'm not going to go telling them to go spec arcane because some people like it. Most mages are going to do more dps as fire, even if they are T5 geared. Arcane is at best difficult to play, and at worst behind on single target dps. Not only is it less damaging on paper, it has less range than a fire spec and less pushback protection. Don't forget I was one of the first people on these boards to really seriously investigate arcane cycles; I am well aware of its advantages as well as its limitations. The potential was interesting and I played with it in raids for a long time, but with 2.3 its time is pretty much over.

I encourage mages to play with talent and gear selection as much as they please. However, my personal recommendation is based on theorycraft, and is not going to change.

www.magegraf.com

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

"We agree with Communism." - Greg Street 2009

 

Kavan

I think the assessment of arcane in the post is decent, but covers only the absolute basics and misses some of the most important aspects. I definitely wouldn't call the spec outdated. As Zephriel pointed out above proper mixing of AB spam and AB rotation is the key to proper use of the spec. I understand that the recommendations are made based on theorycrafting, but proper modeling of arcane is difficult to say the least.

Mana management is the most important element in arcane and most theorycrafting tools don't take that into account. There are two aspects to this. First is five second rule and how it interacts with Arcane Missiles. The other is realizing that in most encounters you won't have 100% dps time and that mana regen in off-time is very important. Take for example our first Kael'thas kill Wow Web Stats. You'll see that I only have 72% dps time, that means about 5 minutes of full spirit regen. Obviously different encounters have different specifics, but Arcane can take full use of down time and drastically increase dps when casting while other specs are mostly limited to fixed dps output.

The other aspect is spell selection. While your spreadsheet does a decent job at this it still has many problems compared to true LP solutions. For example if I enter in my gear it will suggest that ABx3+AM+Sc is not possible compared to target consumption rate and will use ABx2+AM+Sc for AM+Sc filler, but for AMx2 filler it will decide that ABx3+AMx2 is possible and say this is actually the best cycle when it obviously doesn't make any sense. A true LP solver would realize that the actual solution is a combination of ABx3+AM+Sc and ABx2+AM+Sc (given the cycles that are tested).

So what I'm saying is that basing recommendations on theorycrafting is good, but one has to realize the limitations of theorycrafting tools when making such recommendations.

 

Gohei27

I can't argue your math or your theorycraft, I do however speak from experience. I can't argue how I loved speccing fire, I would of done it earlier in AQ if I didn't rely on IB at the time for Huhu and Anubisaths. However I did change towards the end since you could not beat the dps. In turn fights like Huhu and trash Defenders ended much quicker. In Naxx fire was king, no class could compare to mage dps, it was completely broken.

Illidan however isn't a fair fight to gauge fire vs arcane, however in that WWS we killed him in 15 minutes, not only was the kill fast I maintained my mana for 15 minutes with no problem. This isn't a 60 second spec, you are wrong. If you seriously tested out arcane heavily and still think fire is better. Then you have a seriously flaw in your test. It's biggest draw back is being dependant on someone else in your raid to rely on steady and high dps. If you notice our shadow priest gave us 32k mana and I used every drop, I ate several gems and potions and used every drop, you end fights with half a mana bar and evo still up?...

Fire loses to threat and time vs arcane, the longer the fight the better chance you have, however I can power up, blow it all, evo, then start a cycle and threat was never an issue. Our tank would have to get parried and dodged too many times in order to pull aggro. 40% vs 10% is just insane. The ability to control your dps is where I think arcane ultimately wins. You can keep a steady cycle just like fire, or you can use every last bit of mana left to get every bit of dps in.

With the changes to 2.3 fire and frost are now viable, before hand they didn't even have a chance. Recently we have mages testing fire now, seeing how it adds up. It does compete now I won't argue that, but I do not believe for one moment that these changes are making arcane obsolete, even with four piece t6.

Your two points about push back and range have no effect on arcane btw in high end raiding. 11 yards isn't an extra fireball when doomfire is coming, in fact I bet I can get my quick AB off and run while you would only get your spell off if it was mid cast. On Azgalor the silence is actually long enough to drop the debuff (so just spam ab, clearcasting missile of course), so even though you have range I'll get more dps in during those short silences, if I get a resist on silence gogo AP. Range has no advantage on Winterchill, Anetheron, Kaz'rogol, Nej'entus, Akama, Teron, Bloodboil, RoS, Mother or Council. The one fight fire's range really beats arcane is Supremus, enjoy your partial resists.

Push back only effects me on one fight which would be the second phase of RoS, the push back also helps me. I can keep spamming ab, the push back helps on the mana use, then I missile during the start of every deaden. Given the reset of mana every phase and I get to use AP every phase, fire doesn't even come close to me on this fight, it is too short. So pushback has almost zero effect on my dps and I use it to my advantage when it does.

The only fight I've lost to fire to so far has been Council which I died on and to be completely honest, I probably would of still lost if I had lived.

I know it might seem like I'm taking this too far, but people take your opinion very seriously. To say that it isn't viable and that no serious high end raider would spec this way is absurd.

When I read this sentence I honestly think you just didn't like the spec and that personal preference dictates your opinion more than facts. It might be difficult at first and maybe in the sense you have to pay attention but I think this spec is the most fun to play. Going from spamming frostbolt and managing your timers, then to maintaining scorch and timers and spamming fireball, to spamming ab, missiling on clearcast, managing timers and mana. I however think that this spec is best suited for high end raiders, since they would probably pay the most attention and be better skilled.

http://armory.mmo-champion.com.nyud....37170zyBei.png