[Mage] 2.4 TC, aka Guys in Blue Dresses
 

• 
  For basic knowledge about Specs, Stats, Rotations and more, your thread is Vontre's Mage Information Thread, or Mage: Help me

List of Mage Links:
WWS analyzes, PVP thread, Mage help (speccs, items)
The 3 old TC threads: TBC Mage TC, How can Arcane dmg work, 2.3 TC
Spreadsheets and Calculators: Vontre's Mage DPS spreadsheet, Lhivera's Calc, http://elitistjerks.com/f31/t17454-mage_dps_simulator/
Other Threads: http://elitistjerks.com/f31/t13305-m...gic_mechanics/, http://elitistjerks.com/f31/t15980-m...iously_broken/
The Rawr Project



->Ultimate Mage Video (Big Blue Dress, youtube) <-


-Mage Changes in 2.4 PTR 2008-02-23

• Spell Haste: Spell haste now reduces the global cooldown on spells, down to a minimum of 1 second. This change does not apply to melee and ranged abilities.
• Arcane Explosion: The damage cap for area of effect damage on this spell has been increased by approximately 50%.
• Power Infusion no longer stacks with Bloodlust/Heroism.
• Spellsteal will no longer overwrite a longer duration buff.
• Using low ranks of Fire Ward and Frost Ward spell will now be penalized the same way healing spells are penalized.
• Fire Ward: Using low ranks of this spell is now properly penalized like other healing spells are penalized.
• Frost Ward: Using low ranks of this spell will now be penalized the same way healing spells are penalized.
• Frostbite: When a frost spell is reflected back at a Mage, it is now possible for the Mage to suffer from the Frostibte effect.
• Ice Barrier: Using low ranks of this spell will now be penalized the same way healing spells are penalized.
• Permafrost: This talent will now correctly increase the movement slowing effect of frost Armor instead of the attack speed slow.
• The casting bar will now display properly when casting Blizzard or Rain of Fire with spell haste.
• Arcane Fortitude now increases your armor by an amount equal to 100% of your Intellect, up from 50%.
• Blink, Slow, and Spellsteal have all had their mana cost reduced.
• Improved FireWard has become Molten Shields.
• New Talent: Molten Shields will cause your Fire Ward to have a 10/20% chance to reflect Fire spells while active. In addition, your Molten Armor has a 50/100% chance to affect ranged and spell attacks.
• New Improved Blink: For 2/4 (Rank 1/2) seconds after casting Blink, your chance to be hit by all attacks and spells is reduced by 25%. (Screenshot)
• Icy Veins has been changed, instead of increasing the chance your chilling effects will Freeze the target by 10%, it will now give you 100% chance to avoid interruption caused by damage while casting. (duration remains the same: 20 sec).
  ---
• Players can now only carry 80 Conjured Manna Biscuits at a time.
  
  -Other mage-related changes
  
  --Other Mage related changes:
  * Spirit-Based Mana Regeneration: This system has been adjusted so that as your intellect rises, you will regenerate more mana per point of spirit.
  * Power Infusion is now 20% haste instead of 20% damage
  * Mystical Skyfire diamond is now 320 haste for 4 seconds on a 45 second internal cooldown
  * New epic gems (Notably: +10 spell haste, +6 damage/5 haste, +5 haste/+7 stamina)

General results of the 2.3 TC thread and unresolved topics of discussion


1) Skull not stacking with other triggerable trinkets has left some debate over what is the best trinket combination between Skull, Serpent-Braid (due to its continued unique ability to work as a 2nd trinket at the cost of flame cap), Hex shrunken head, and Darkmoon Card: Crusade, and Sorc's Alchemist stone
2) 2 leading dps specs, 40/0/21 or 0+/47+/11
3) Continued uselessness of spirit despite the transmutated set and our new peices of t6 having it.
4) Ignite continuing to work "weird"

Should probably add that the reason there is no clear winner when it comes to spec is that they depend on a lot of factors, mostly fight duration VS mana available, where arcane would clearly win on extremely short fights and fire clearly win on extremely long fights.

And speaking of videos, considering most of the "my DPS is low what am I doing wrong" are simply caused by not casting enough and/or wasting too much time on movement and sometimes with a little not using cooldowns properly as an extra, actually making a low-medium quality video showing a couple raid fights with fully visible UI could be useful to avoid handling those continuously posting the same questions. And while we're at preventing old questions repeating themselves, you should probably write the first line in the post in bigger letters ;)

You skipped over a few points, namely

Arcane 'rotations' are out for good. This is an indirect effect or the int/spi changes allowing for very high AB dps% (in relation to other spells). Arcane new playstyle is mostly a reverse rotation; you cast AB to refresh the AB debuff to avoid having it fall down. Then you cast fireball/frostbolt in order to generate mana via spriest/JOW/mana regen as a filler spell whenever you can't sustain your AB spammage habits.

There are a few points of contention as far as TC goes, namely

for arcane spec (40/0/21)
-> I made an open challenge to anyone to work out the math comparing whether you are better off unstacking IV and bloodlust for the express purpose of being able to spam AB during hightened DPS/DPM times. You can frostbolt with bloodlust and IV stacked, but you fall below the 1s GCD, which is why I proposed someone to check out the math, so that you could continue AB spammage during both bloodlust and IV provided that they are not stacked. Personally my gut feeling says you're better off not stacking them, but its nothing scientific by any means. The math behind it is quite involved to say the least.

for fire spec
-> TC does not handle 'proper' ignite management and assumes pretty much that every fireball will deliver its ignite dmg. In other words, it ignores the possibility of the boss dying before your ignites have had the time to run their full course. As a contention point there are many strategies for managing that, which are being somewhat contested. The possible options are roughly AB spam, AM, scorch and fireblast for the final 3-5s of a fight (in a context where you do not know how much time is left on a boss and yet have a huge buildup ignite). You know its somewhere between 0s left before boss dies and 5s. This also deals with travel time due to an indirect effect, basically, you want to avoid casts with travel time at that point in time because you want to be sure to maximize how much dmg you can deal given an unknown future. I believe only 1.5s casts should be used for that context, but again, this is debatable. Personally I finish fights with scorch/fireblast spam as a mechanism to lower this, but the practice is very much debatable.

-> as you hinted in your post, there is a contention point concerning how to stack cooldowns now given that skull does not stack with other trinkets (other than SCB). I personally believe that stacking everything minus [combustion / hex shrunken head] is the optimum choice. This implies that right after the 20s (of your cooldowns being active), you follow it up with [combustion/hex shrunken head] since you're still under the effects of bloodlust/flame cap/drums of battle.

Is this accurate? I thought this change was reverted before 2.4 went live.

Correct. Personal haste buffs will still stack with party buffs like Drums of Battle and Heroism.

To actually figure out the strategy for end-of-fight you'd have to define how the end-of-fight works, first. That means you'd have to figure out the probability density function for the fight's remaining duration as well as the stacked ignite damage and when is it going to tick off. Without that data there is no basis for any calculations of proper action.

Anyway isn't the effect of these last few seconds rather small? You'd need a massive rolling ignite into the end of the fight which we already know requires cooldowns and even then still has a low chance to happen and doesn't last long. With you getting at least 1 ignite tick between fireballs the most ignite damage you could ever lose would be 1.5x fireball worth of ignites (0.9 average non-crit fireball damage lost) plus 2s worth of casting (2/3 additional fireball). Losing a whole fireball and only a single ignite tick (if boss died as your last fireball almost hit him and ignite from the fireball before was still waiting to tick) would probably be an actually bigger loss. Overall looks like the maximum damage you can lose, if you're just spamming fireballs at the end, is quite small. And when you think of the alternatives (that I doubt really have a higher success rate if you're not sure of exact fight's remaining duration) it just doesn't look like there's much point.

Am I missing something and you actually have some huge ignite rolling when the boss dies or something? Or is this just trying to reduce that tiny DPS loss that's probably not possible to reduce by much if you don't know the exact fight's remaining duration?


For stacking vs not stacking IV with bloodlust being worth it, I think it's too personal to make a generalization and just use something like rawr.mage to figure it out for your gear and fight.

The debate is subtler than that. I define end of fight as 'a point in time where you know theres approximately, at most, 5s left before the boss dies'. It could be 0.1s before the boss death. Dodging away lost ignite ticks is a minor detail for the most part, but it can have impact bigger on other specs, such as arcane/fire, where you want to finish with AB spam (which you would probably end up doing anyway, but point being - you want to avoid if possible losing ignite dmg. The real point is more generic and subtle. Namely, given the context, 'are we better off switching to 1.5s- spells at that point in time in order to cast more spells before the boss death, even if those shorter-cast-time spells have lesser DPS than their longer cast counterparts?' Ultimately, it depends on how much of a dps difference there is between your spells. I don't know if we could design a guideline that could give a clear answer to that. As I have hinted, I believe it is worthwhile for firespec to switch to fireblast/scorch at the tail end of a fight, but its nothing really scientific.

Correct me if I'm wrong, but it essentially amounts to I don't want a fireball in the air when the boss falls over because I could have scorched and done more total damage.

Dumbed down obviously.

It really depends how sure you are of the remaing duration. If you think there are 0-5s left and it's evenly distributed, leaving ignites alone for a sec just for the sake of the example, casting a fireball would have 2/5 chance to land while casting a scorch would have 7/10 chance to land the first, 4/10 to land 2 and 1/10 to land 3. On average you would land 1.2 scorches VS 0.4 fireballs. Taking ignites into account and/or changing the duration range and distribution in any way can significantly affect that result. The sure thing is that the lower your certainity that the fireball will not land the more worthwhile it is to simply continue your rotation.

Another thing to note is that even if changing spell rotation is optimal, it'll actually increase your variance in DPS (assuming you're still not certain of the time remaining), which may or may not matter. On fights where the wiping factor is an enrage timer you should probably just assume the fight ends when you die rather than it ending when the boss dies, since if the boss dies early you win anyway but if he doesn't you want to do your best possible. Then again even enrage timers aren't accurate enough since they usually don't instagib the raid.

Just a quick note... I was 41/0/20 before and after 2.4, and Arcane explosion is rocking now. With Quag's eye, the 50% higher cap on AE, the improved mana regen from spirit/int, I am able to sustain Arcane Explosion spam on Hyjal trash much longer now for more damage, and if I get a haste proc or heroism things get even better. Before seed of corruption warlocks were crushing me on the AoE total output, and now I am competitive.

AB spamming with 2pcT5 just got better too. Icy veins prevents interruption which offsets one of the big issues with AB...spell pushback. So I can pop all cooldowns in a DPS window and really pump it out without fear of interruption. Mana regen lets you go longer. I agree with Manly that the new mode is to AB spam til dry then recover with efficient frostbolt spam. I usually lack a shadowpriest and JoW pally, so I just burn down to 20% mana and spam frostbolt until a mana gem or pot comes back, and repeat.

Updated, will do a few more revisions when i get back

Any chance you can add Rawr.Mage to the list of calculators?

Updated with rawr.mage, and i have a few points to disagree with

Manly: If the object of theorycrafting min/max is to, in the end, use that theorycraft to assist you in killing the boss, then why not take that into consideration?

Rather, what's the point of discussing letting ignite tick off on a fight like brutallus? Where he's enraged at 3%? Twins? Felmyst? My point is that if it really came down to it, and you're the last dps alive and you can cast one spell to finish him off at 1%, and you have a huge ignite ticking, why not Ice block? Blink away? And if you cant and that's your only spell you can cast, wouldn't fireball be the spell of choice since it does more damage even ignoring ignite?

The window of saying it's better to cast arcane blast if im the last mage alive (if we're talking about letting ignite tick as fiery veins) if and only if A) I'm the last person alive, so letting ignite tick might be the only way to do enough damage needed to kill him, B) That the damage lost arcane blasting instead of fireballing is enough to kill it, and C) you cannot Ice block, blink, BoP, or any other spell to buy the time to let the ignite tick seems like an awful lot of "If's", with little gain.

Anyways ignoring the tangent Serpent-braid and hex is the hands down best combo for 40/0/21, especially since it lets you use all the cooldowns at once without going under global.

Would like to raise another discussion point. Because of arcane's new style of play, which is AB spam as much as it is humanely possible. Arcane fares best for fights ranging from 1 min to 4 min or so. (You can comfortably AB with good shammy & SP support plus evoc plus gems+pots up to 4 min).

But stacking int/spirit onto gems may lengthen that time at the cost of DPS. Based on Rawr, int/spirit is best. But I question that. I feel that it is all dependent on the fight. For shorter fights where even when gemmed on damage, you won't run oom spamming AB. Then it doesnt make sense to sacrifice DPS by gemming for int/spirit.

However, for longer fights (which may or may not be that common). Then, gemming for int/spirit allows you to AB that much longer, and it does equate to overal increased DPS when measured over the whole fight. So, for arcane mages raiding in post 2.4. If there is the luxury, they may want to consider having an altarnate gear set that is gemmed for int/spirit. While having a more standard DPS set for farming/instance and most standard raid boss fights.

So, for that one or two odd fights in the game which lasts for 6 minutes or longer, throw on that int/spit gemmed set. Of course, not every one has that kind of luxury. But some mages who have been farming BT for months probably do.

Because while gemming for int/spirit may seem to be best for DPS based on Rawr. This is only true if its a long enough fight such that you need that extra mana regen and mana to finish it spamming AB. If you can finish the raid boss fight spamming AB without needing to gem your entire suit with int/spirit. Then, you will achieve higher DPS by gemming for damage.

So for example. For a 4 minute fight. Perhaps a full out dmg gemmed set will let an arcane mage achieve 2400 DPS spammin AB. For a 6 or 7 minute fight, the same arcane mage will switch to a spirit/int set that allows him to do 2200 DPS by able to AB for most or all of the duration.

Of course, if your raid is willing to throw all their innervates at you, then just gem for damage cos you won't ever run oom. :)

We must consider that eventually, Sunwell will be a 6 boss fight. Its not going to be just Brutullus. And most BT & Mt Hyjal fights are short. So, it doesn't make as much sense gemming all of your gear for less damage just because you want to be able to AB spam for that one or two long fight. I know most mages are using Brutullas as a benchmark for theorycraft right now. But eventually, when all gates are open, Sunwell won't be defined by just that one fight.

I'll contest this view. From what analysis I have done int is bettter than damage no matter the length of fight. Unless you're in a position that you don't even have to use evocation int will always be better for arcane (or until you have it stacked to disproportionate values).