no the tooltip never changes from 2.46 seconds with my 29 haste regardless of the whether I have 2 or 3 debuffs on me.

I was testing this just the last few days. As I've been the only mage lately, I've been specced arcane for CoS. I'm positive the tooltip does show below 2.5 seconds, as I remember using the tooltip (spellbook one if that matters) when I was fully stacked to make sure I was taking enough haste gear off to avoid the GCD block.

It is clear as mud now sorry for not thinking straight.

Yes it does show below 2.5, but it doesnt change on subsequent cast while applying debuff. My tooltip shows 2.46 right now, after I apply a debuff the tooltip still shows 2.46.


hmm, I am using drdamage and have it on bar, Ill open spell book see if it changes.

ok, your right. my numbers. I was holding the mouse over the tootip, you have to pull mouse off of tooltip then back onto it before it updates.

start 2.46
1 debuff 2.13
2 debuff 1.8
3 debuff 1.47

This leads me to believe that each debuff takes .33 seconds off regardless of where you start. Hence, stacking enough haste to get it initially to 2 seconds will allow for a 1 second cast after 3 debuffs are applied. BTW this also tells me that the debuff ignores GCD, because I can actually take it below 1.5 now on live. Basically the gcd is irrevelant when it comes to arcane blast. I didnt understand it intiially becuase I was was thinking it took off .5 seconds per cast. But, 3 (.33)=.99

You can't go below the GCD even with AB. This is the reason why 3 stack AB is useless with heroism/bloodlust and arcane mages avoid haste. This changes with 2.4 where with haste you will be able to drop the GCD to 1 second.

I think you believe you going below the GCD when in fact you are not. With only a .03 off the cast time I do not think it would be noticible with even the best possible latency. It would be much easier to show if you either tried a 3 stacked AB with heroism/bloodlust, or if you are specced for icey veins, to pop that and try a 3 stacked AB. You will notice that you will hit the GCD.

I was just going by the tooltip, Its says that my cast is 1.47. Whether it actually breaks the gcd, I dont know. I would think if the tootip says it does, it should. But, who knows with blizzard.

You are wrong sir, you CAN go below anything you like. The cast time WILL be any time interval from 0 upwards. The fact that the GCD prevents you from Starting another spell is irrelevant to the point he's trying to understand, which is "does the AB debuff give a flat -0.334sec per application, or is it a % increase".

Am I glad your here... I have another question. in 2.4 if I stack enough haste to get to a 2 second cast. I Can get my arcane blast to 1 second. By doing so will my gcd be 1 second as well for that spell? or do I have to stack more haste to lower the gcd, or does arcane blast just truely ignore the gcd?


I do understand now that its a flat .334 now. thanks for all who helped me understand this.

I believe you would need 50% passive haste to lower your gcd to 1sec....which is 787 haste.

With bloodlust...you would only need 236 haste.

With IV...you would only need 393 haste.

If they let IV + Bloodlust stack as in the current build...0 haste.

However your goal shouldn't be to lower your GCD to 1 if you're going to be casting AB. This only worsens your DPM. The reasons AB is being looked at again is the changes to the mp5 calculations and that haste is no longer a wasted stat for AB...but not saying it should really be stacked.

Spell haste has absolutely no impact on any spell DPM. Although, every other stats will increase your DPM.

Thank you for saying this. It's something I see forgotten often. DPM >doesn't change with haste<. It's damage per mana. You do the same damage for the same mana. You just do it faster.

Or you could just refer to it as MPS.

mps doesn't take into account the amount of damage done.

MPS is problematic in that it's haste-dependent. DPM is an indicator of efficiency, not consumption. MPS is the opposite.

I'm not sure it's accurate to say AB ignores the GCD, it's a little odd because AB gains a haste-like benefit from both specifically haste and it's own -cast time effect. I haven't tried but nobody has mentioned yet that any kind of GCD has ever limited anything over 1.0sec so it's safe to assume the stacking of percentile haste (from talents/spells) and flat cast reduction (from talents/AB debuff) and haste bonus from gear is working fine.

It's safe to assume you will not need additional haste to move the GCD.

If you're going to take DPM into consideration, you have to know which DPS consumeables and which mana consumeables you're already using to go through the fight (assuming fire, as arcane obviously uses all mana consumeables as the extra AB spam will increase DPS more than any dps consumeable). The more mana you have (as a fire mage), the less effective more mana is, becuase you'd be dropping better mana consumeables for not-as-good damage consumeables (say, if you're very low on mana, getting more will reduce evocation time and be a big dps increase. If you're already not evocating and using flame caps but still mana potions, more mana will just allow destruction potions which aren't as big of a boost).

Overall if you're just calculating average fireball spam DPS and get 12 dmg ~= 10 haste, then obviously 12 damage is better. Wether 10 haste is quite above or under 12 damage, though, has quite a bit of dependancy on your gear, since they're rather close. Hopefully updated versions of spreadsheets/simulatros can/will properly take into account the more mana->more dps factor.

On a side note, it would be nice if on the simulator/spreadsheet you could define "no dps" periods to take movement into account, where regen from buffs (and spirit if applicible) are still taking effect as well as VT+SWP regen. I've noticed from MC all the way to BT that the fact you're not realy fighting for the whole X minutes the fight lasts has major effects on your mana sustainability - and it gets more significant the more regen you have. For example I bet on the simulator in a situation where 2.4 arcane and fire are equal, if you adjust it to 90% DPS time arcane will win all of a sudden - as not only you had an effectively 90% fight duration instead of 100% (and we know shorter fights favor arcane), you also had mana regen of 100% (a little less since the shadow priest also isn't attacking 100% but does have DoTs ticking) which favors arcane even more. My main point here is that in a real X minute fight you will have a much "harder" time to use all your mana than in a simulated "stand-and-shoot" X minute fight. And since regen is calculated on the full duration rather than the "modified" duration it makes a difference for theorycrafting the DPS of different gear/spec/setup.

Over any given amount of time you're going to regen the same amount of mana regardless of whether or not you run yourself oom (assuming you stay within the fsr). You end up with the same amount of extra spells through regen either way.

Haste allows you to dump your mana faster with AB resulting in better use of cooldowns (ap, trinkets, etc) so if anything haste slightly increases your DPM while increasing your MPS.

Why would AP inscrease your DPM? Your DPM shouldn't change when using AP since DPM is dmg/mana and while using AP you multiply both by 1.3 which means the ratio dmg/mana stay the same.

due to how things stack, AP has a bigger relative DPS increase than a relative mana cost increase. AB with 3 stacks is already +225% I think (verify plz, didn't really check), adding another 30% to that is nothing compared to adding 30% to your dps, even if that 30% is additive to other multipliers (which would make the relative increase slightly weaker but nothing close to how small 0.3X is compared to 3.25X).

At the risk of getting flamed for "useless post", I have a question: What would you guys think about Blizz making Spirit count towards Spell hit in a similar fashion to the way Int counts toward crit rating now? It would make Spirit a more useful stat and would allow for more diversity in gemming/enchanting of items the way I see it. Thoughts?

It's an unsound idea, Tucker. The highest-spirit classes are druids and priests, who both are irrelevant to it. Particularly for mages it's also irrelevant as most likely the only spec that will go nuts over spirit will be arcane and there's no hit lack in arcane for sure. On the other hand, warlocks whose eyes glaze over at it's mentions will never have spirit on their gear due to the nature of their regen.

Besides there's a massive spirit rework with WotLK we're told. Who knows what twisted madness they'll come up with.

Sambjo: You are correct but the benefit isn't flat it's spikey; That means you'll only gain value from it during AP/IV whenever it grants you one additional cast. Say it takes 140haste (randomly picked, should be close enough) to get 13 fully-ramped ABs into an AP (with IV) cast, then you'll gain approximately nill from having 139. In fact, it's arguable you'll lose, because haste is counter-productive to OO5SR rotations like (AB/AM)*3/Sc, though those, as well, are odd in that as long as AM+(fastest)AB>6.1sec. That's 6 because it takes 5sec to go OO5SR but tics are processed every 2sec.

From my understanding DPM mean Damage per Mana and AP increase the damage you do by 30% after every stacking effect you can add (trinket/destro/crit bonus/etc).

It looks like simple math unless i'm missing something
If you say
X = damage after every stacking effect without AP
Y = cost of the spell

then DPM = X/Y

AP increases your damage by 30% and increase the cost of the spell by 30% (Arcane Power - Spells - World of Warcraft)

DPM with AP = (X * 1.3) / (Y * 1.3) = X/Y = DPM

This is true only if AP increase your damage after all stacking effect which is what believe it does.

Thx for response. That makes sense. Just would be nice to find a use for all the spirit Blizz likes to put on mage items, but hopefully that will be addressed in WotLK.