Allow me to explain the 'hating'. This is a thread about theorycraft, and by and large the people here are very much interested in scientific and mathematical analysis of different specs. When people come in here all passionate about one spec or another and give little evidence to support their claims other than "well it works great for me" or "I beat the other mages in my guild", the community around here is going to be fairly dismissive. In fact, if you even dare bring up a subject other than single-target uninterrupted nuke fests, you will probably be ridiculed. AoE usefulness? BAH! Shoe me more Teron Gorefiend parses!

I think we all agree that player skill and luck are a large part of each player's effectiveness in a raid, so anecdotal evidence or a WWS parse here and there is just not going to convince anyone of anything. I'm firmly convinced that deep fire + icy veins has unbeatable potential in the largest number of situations. Would you beat me on a meter if the fight was 40 seconds long? I would assume so. Those aren't terribly common though.

Now as much as I think my spec is preferable to yours (otherwise I wouldn't be using it) I recognize the merits of it, and I don't think you're being a crappy mage for using it. Just because my average dps output might be 2% greater than yours doesn't mean your spec worthless and that you should be banned from playing a mage. Spec choice is really a drop in the bucket in the long run, and while I think it is our nature as competitive people to want to prove to the world that our way of doing things is right, don't be surprised when someone else with just as much gusto as yourself comes along and poo-poo's your ideas.

Spell power also adds quite a lot by giving 175% vs. 150% crits with AE.

edit- Clear casting also helps tremendously with AE. It is the one place that I miss it most as 2/48/11.

Manly, i disagree, no one's yet to accurately model 40/0/21's advantage of having consistently paired cooldowns, unlike fiery veins where you have combustion, icy veins, and trinkets all on different cooldowns

Last I checked, Lhivera's calculator uses On-Use trinkets and cooldowns as soon as possible, which would pair the cooldowns. However, what the issue is (as far as I know) is calculating them combined (the extra spellpower the additional frostbolt would get during IV +trinket).

Lhivera's script actually tries to pair cooldowns if it makes a difference. It accurately models the interaction of spell power, trinket use and IV/AP.

Well, the point was just that it works around the AOE cap more than anything else.

And yes searix, cooldown stacking has been checked. The spec just does not fares well. Even if another mage stacks winter's chill.

Well, using my gear for reference, which I know is not fully min/maxed, but should give a good idea. I put the same spell damage for both frost and fire, ignoring the minor difference from hellfire encased pendant being replaced. And swapping belt of blasting to anetheron's noose won't really have any real impact on dps.

And yes I know this is beta stuff, but I am fairly sure it shows you a good idea of how your dps goes as the fight progresses.

In case you ask, the rules are fairly simple for now, all cooldowns are popped asap (threat be damned), with the only exception of lust being popped near 20% for deep-fire.

3 min fight, full consumables (drums applied on your own)
Fire: 2218 dps
Frost: 2204 dps
Arcane/Frost: 2009 dps (4pct6)
Arcane/Frost: 2075 dps (2pct5 (dps would be lower if I adjusted the stats loss from this change) (3x AB/3xFB rotations)

(edit: last one is destro lock)

I don't get the reference, where are those graphs from? Do they model the ghost 3% hit? The fact that AP can be popped ~10-20% more often than combustion? That you can mana dump AB spam at the end of the fight to raise your dps?

Yes, ghost hit is accounted for, it knows the cooldowns of every activable cooldowns (I mean c'mon now...). It even models semi-correctly > 3min for combustion. Yes, it pops your AP/IV/snap/IV during lust. Also pops WE when appropriate for deep frost. In fact, the only thing it does improperly is not stacking very optimally firespec cooldowns because it actually uses it at the start rather than the end (and flamecap), but regardless it does give a rather solid idea.

Can you clarify your graphs with labels and such (Axes and full graph label)?


I find this sort of thing absolutely fascinating (like the triple point diagram for fire specs a while back).

The 40/0/21 build does not compete with either accepted fire builds for single target dps.

It does have many raid advantages and should not be overlooked.

A list of several of them are as follows:

ILVL:

According to my math, 2406 +spell dmg is needed for 10 crit rating to be the same dps upgrade as 12 spell damage for single target frostbolt dps with a benchmark 35% spell crit. For ilvl cost comparison I used 10 crit rating = 12 spell dmg (based off of gems).

Math:
10 spell crit rating = .45% crit
assume current crit% as 35%, (base crit% varies greatly on gear/buffs)
x = +spell damage worn
(661+x)(3/3.5)(.95)(1-.35)+(661+x)(3/3.5)(.95)(.35)(2.25)= average frostbolt dmg
(661+x+12)(3/3.5)(.95)(.65)+(661+x+12)(3/3.5)(.95)(.35)(2.25) = (661+x)(3/3.5)(.95)(.6455)+(661+x)(3/3.5)(.95)
(.3545)(2.25) - solve for x
x = 2406 spell dmg

How much +spell dmg is needed for 10 crit rating to be the same dps upgrade as 12 spell damage for single target fireball in a fire spec with a benchmark 45% spell crit?

Math:
10 spell crit rating = .45% crit
assume current crit % as 45%, (base crit% varies greatly on gear/buffs)
x = +spell damage worn
(809+x)(1.15)(1-.45)+(809+x)(1.15)(.45)(1.5)(1.4)= average fireball dmg
(809+x+12)(1.15)(.55)+(809+x+12)(1.15)(.45)(1.5)(1.4) = (809+x)(1.15)(.5455)+(809+x)(1.15)(.4545)(1.5)(1.4)
(821+x)(.6325)+(821+x)(1.08675) = (809+x)(.627325)+(809+x)(1.0976175)
x = 7213 spell dmg

What do these numbers mean?
The ratio of effectiveness for +crit for arc/frost is 7213/2406, or extremely close to 3:1. That means for every piece of gear you have which gives you +crit, the +crit helps an arc/frost build with that ilvl by 3x as much as a fire build. Likewise, if you want to gem your gear with +crit (due to being an AE spammer) you will not be gimping your frostbolt dps as much as if you were a fire spec.

Added to this are the talents Arcane Mind and Mind Mastery, which increase the bonus from int spent on ilvl by leaps and bounds. In a full t6 set with endgame BT/pvp in every other slot as well the bonus to dmg from mind mastery and bonus to crit from arcane mind provide a substantial increase in dps vs. fire spec.

The current +3% ghost hit from frostbolt also frees up substantial spell hit ilvl to substitute +crit/+sta/+dmg as well.

TALENTS:

Every talent spent for the 40/0/21 build is useful. This may not be important to many people due to a 'bottom line' approach to damage meters - but it does mean that all 61 points are spent effectively and like that concept appeals to some people.

Most importantly there is freedom to move points from arcane focus to magic absorbtion and from elemental precision to arctic reach if you have large +spell hit on your gear. Both of these talents are extremely good. The 2/48/11 build has no talent freedoms like this.

PLAY STYLE:

Due to the extremely large number of by-choice dps enhancement spells and abilities the arc/frost spec gives you the most ability to burst dps and to utillize freedom in boss encounters. Very few fights let you sit there and dps without interruption, a consistancy fire spec thrives on. Because you can time your dps bursts in arc/frost you can maximize utility of boss timers and your own resists as well as group timers like drums and heroism.

Besides single target dps, there are times when AoE dps is important. Although Blizzard has pretty much neutered our dps output in this category compared to warlocks, we can provide almost as good AoE dps with the arc/frost build. Fire spec has much worse AoE dps.

PoM sheep can save raid wipes.

If you don't like spending money respeccing to pvp, the arc/frost build is a much better pvp build for battlegrounds compared to the fire specs.

DISADVANTAGES:

No Blastwave.
No Dragonbreath.
Spell haste hits the GCD with frostbolt much easier - so haste gear has a much quicker cap to help dps.
No dot component to spells, so on mobile fights you do not get near as much dps during motion.

FURTHER THEORYCRAFTING:

My question is this:
In a 3 mage raid setup on a single target boss can/would a 2 arc/frost mage 1 deep frost mage setup outdps a 3 fire mage setup assuming full t6 gear/ilvl (bonuses from arcane mind and mind mastery has largest influence on dps)?

If so, a 2 huge AoE/burst mage and 1 utility/survivability mage seems much more usable than 3 fire mages on trash for a raid leader.

Sorry guys, but i think we've had a huge amount of posts and TC done, why/where and when arcane is proper and that all in all fire (and even Ice) is the better joice.

Arcane is nice, i also love to play it - but in 98% of fights it's not equal.

I'd love to see the discussion going back to manly's post concerning the COE thematics and - maybe the future outlook of arcane as an util alternative raidspecc.

... and just again only my 2 Cents

I recently wondered weather to focus on a Frost build or a Fire/IV build and clearly Fire/IV does more damage in theory.
However my guild progress currently is MH complete and Bt up to Gorefiend, and I came to find that I perform closer to my theoretical dps on most fights with frost than with fire. More importantly I perform better with Frost than with Fire dpswise on certain encounters I did not expect it.

Part of the explanation I came up with is that the some part of the bosses/our strategies requires me to stop dps:
Winterchill -> need to move out of AoE or am entombed
Anetheron -> I typically am on add duty here. Short times of no dps, while I wait for the new add to positioned.
Azgalor -> silence periods interupts casting. I tend to restack scorch 8 seconds before it runs out to be safe. I had bad results with less careful aproaches.
Supremus -> gets out of range a lot

What these fights have in common that on every occasion that I loose casting time, it gets subtracted from my fireball time, because maintaining a firm scorch debuff is mandatory. Fireball time is however better dps than scorch time! Hence: not dpsing x% of the boss encounter time results in worse fire performance than just saying I loose x% of overall damage. Frost does not have this problem, because debuff practically never drops and is reapplied by the main and only nuke spell.

An often used argument against frost is that you need to ensure the water elementals survival and spell pushback which lets it perform worse in theory than in practice.
However I wondered if this is an occasion where fire performs worse than expected, and more importantly: Did I overlook approaches to model this? (because I didnt recall the issue)

Your thoughts?

9% Dark moon Faire buff as 40/0/21
Wow Web Stats

I haven't had the chance to test fire due to my gear being a lot better suited for Arcane, however I don't really find the difference to be big enough to justify specing fire. Even if I could maintain the extra 100 DPS, there are a lot of extra small advantages to Arcane in the raiding environment. My gear is almost flawless for arcane sub a couple small minor upgrades for extra haste pieces from a belt and a necklace I think, where as I still don't have an Archimonde Sword for Fire Spec.


Also, one thing to note is I don't cast frostbolt, that really doesn't make a lot of sense to me since the point of 40/0/21 is really to double stack your haste on AM, where as you would just clip your frostbolts with this much haste.

Oh, the downsides of fire are quite simple. For the most part, this is what matters most:

disadvantages -- fire
**************************
-ramp-up time (scorch stacking)
-ignites falling off particularly on low hp targets (ie: trash mobs)
-3% more incoming damage (this really really should not affect your playstyle or dps, so it is rather moot)
-relies on proper timing of cooldowns (yes, bloodlust on execute range with everything else stacked will make a tremendous difference)
-inconsistent ignite damage (sometimes you win due to rolling ignites, sometimes you lose due to them falling off)

disadvantages -- frost
**************************
-ramp-up time (wc build-up time)
-no pushback protection
-range
-slightly worse scaling than fire
-absolutely requires proper timing of WE before bloodlust
-your WE absolutely cannot die. if it does so will your dps.
-your WE needs a SP to be able to cast the full distance, particularly with bloodlust (again, very minor detail)

disadvantages -- arcane/frost
**************************
-ramp-up time (you absolutely require another mage to stack WC for you on any fight duration that is 'long' (or if you prefer, any fight where most of your damage will not be AB) if you want to have any kind of hope to show on DMs)
-reliance on both COS and COE
-no pushback protection (both AB and frostbolt)
-you cannot afford to screw up a rotation. the dps loss from having to begin a rotation with a 2.5s AB is huge. your rotations must begin with a 1.5s AB.
-range
-bad DPM (this is arguably not a big deal)
(and debatably reliance on 2pct5, but that's somewhat minor given that you can swap em in and out depending on fight duration)

An important part of a mage's dps is to minimize the downtimes. Your examples are good :

- Blink is really your biggest friend on Winterchill, allowing you to quickly escape a d&d or break an ice tomb (yes it does work). Managing this simple spell will considerably increase your dps (and survival).
- Just wait for the inferno to be positionned correctly before dpsing him. You can in most cases afford to lose 7-8 seconds dps worth on the inferno. If placed well, you are in range of both Anetheron and the inferno, which means you barely loose any time switching target. The only loss as a fire mage is the time stacking scorches.
- Equip your entire RS set on Azgalor. Yes you will lose quite a lot of dps, but very much less than being silenced 5 seconds every 15 seconds. On yesterday's Azgalor, 341 RS (no mage armor), and only 1 silence on the whole encounter. I would have been top 3 dps if he didn't doom me at 10%. RS is the way.
- Again, blink is your best friend on Supremus, and using fireball means more range. I am the first surprised to almost top dps on this dude (first usually being a hunter) as a fire mage with only 48 SP, but anyway, mobility is the key here.

I've been frost spec a few days while trying and beginning to farm Illidan, and therefore have been able to test it against all T6 bosses. The surprise is the really high sustained dps you can get as a deep frost (basically p1 Illidan), but what really kills it is the lack of any pushback protection or worthwile instant you can cast while moving (you end up casting a fire blast, which as a frost spec is not nearly as good as it can be). Being lucky and having both a paladin and an earth shield on Teron allowed me to get like 1600 dps, once. Even though this sounds solid, it just doesn't compete with what I can get with less effort as a fire mage. Another difficult boss is RoS, where p2 basically kills the frost spec.

The fact is that a spreadsheet will give you a Patchwerk dps score, but the frost spec is much more vulnerable to any disrupting environnement, and therefore will finally do a lot less than the maths would suggest. The WE is a huge boost, but I guess anyone here experienced on Illidan p2 the "ok I'm popipng the ele now, use bloodlust... no wait he's been instagibbed".

If you want to know what is the best suited for you, just try both specs, the spreadhseet will allow you to better choose your gear and tell you what to expect in a perfect situation, it's up to you to see which spec gives you the best dps in raiding situation. If you don't take environnement into account, the fire vs frost debate will never end.

I'm not sure why that may have happened. I was just using it to show my DPS using just arcane missiles however as I did not get ghosted.

I wish you could quote sigs, or perhaps you should just make yours size 72 font Manly. PSA: Teron parses with more than 5 million damage are bugged and count Ghosts as part of RDPS. Please do not use those. Remove 4750-5000 RDPS to get a good idea of the real RDPS.

I admit, I never viewed 40/0/21 as an AM spam build, mostly because you end up lacking emp. arcane missiles. But yeah, in any case, you're using a best-case scenario for AM spam (lots of pushback with 100% pushback prevention, short fight duration allowing doubly icy veins to really jump the numbers by a big number).

I hope you won't take it the wrong way, but I think your parse just illustrates how poorly AM spam scales. I'll try and ignore the fact that the log is merged and assume this did not impact your results.

40/0/21 - AM spam Wow Web Stats
************
-> 2'15'' duration, 2264 dps
-> 13% COS
-> +9% dmg from darkmoon faire buff
-> 4 drums of battle
-> 100% pushback prevention

vs

2/48/11 Wow Web Stats
************
-> 2'46'' duration, 2189 dps
-> no COE
-> no darkmoon faire buff
-> 2 drums of battle (both of which I activated on my own, so 2 lost GCD there)
-> 70% pushback prevention

AM spam was much more effective back when there was aggro caps, the swap in my gear right now is such a big leep that jumping to a heavy Fire build simply wouldn't increase my DPS. I'd also have to consider losing some fire talents and go 0/40/21 as not having imp blizzard in a raiding environment would simple create more effort on our raid team for hyjal/illidan.

Pushing 2k without COE is very impressive though.


Also this is the build I go to get emp AM. Talent Calculator - World of Warcraft

Well Mojache, pushing 2k without COE was mostly unheard of pre 2.3.2. However, the point I wanted to make was just how much of a DPS boost icy veins turned out to be for fire spec as well. It really affected tremendously the results.

Also Mojache, now that you edited your post to show your build, I must say I am quite surprised to see your choices. I do liken the idea behind it; you did not really lose much for that AM spam. 2x arcane potency, which is not a big deal, and only 1 mind mastery. I think it is a very fair tradeoff for cold snap, to say the least.

We will typically run 2 mages and 3 warlocks, getting a COE is possible for us it just needs to be justified. I have no real rough number of a DPS curse, but even if it's only 100DPS for a geared warlock, that still means we need to do at least 50DPS each more then an arcane build to be worth it. It doesn't seem like a lot but back before this patch it was compared to spaming AM and putting out 1800 on every single fight.

I planned on testing fire this week but we weren't fully setup for it. I'll try it next week, although I'm still missing quite a few pieces for an optimal build as well as needing a second T6 helm to regem for 3% crit meta.